)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"90f473e65d15cb569e3aa13a43af5ec1f8a5fc0b","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Add specification for the non-persistent PKI token provider and"},{"line_number":10,"context_line":"associated work."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"Change-Id: I6b72466397823967d8183f01e57d46d717003b25"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":9,"id":"1ae5cdf2_50a419b9","line":11,"updated":"2014-06-05 21:06:24.000000000","message":"bp non-persistent-tokens","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"}],"specs/juno/non-persistent-tokens.rst":[{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"512240c571a8b17eed7fee9f35a89ecd4a33c428","unresolved":false,"context_lines":[{"line_number":82,"context_line":""},{"line_number":83,"context_line":"All cases of PKI tokens will be decoded (in a similar fashion to auth_token"},{"line_number":84,"context_line":"middleware) in memory instead of making calls to the DB (limited cases) instead"},{"line_number":85,"context_line":"of referencing data stored in the token DB table."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Notifications Impact"}],"source_content_type":"text/x-rst","patch_set":2,"id":"3ae8d1ca_d73fe623","line":85,"updated":"2014-05-28 14:20:41.000000000","message":"This last little bit seems wordy. Is \u0027instead of making calls to the DB\u0027 and \u0027instead of referencing data stored in the token DB table\u0027 referring to the same operation?","commit_id":"59cd4e6c3ae1fe0f3fe8c9db999b9c6d8e2c65b0"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"094b3cb7a084551a123de6b1dc68b0341ffe7e9c","unresolved":false,"context_lines":[{"line_number":82,"context_line":""},{"line_number":83,"context_line":"All cases of PKI tokens will be decoded (in a similar fashion to auth_token"},{"line_number":84,"context_line":"middleware) in memory instead of making calls to the DB (limited cases) instead"},{"line_number":85,"context_line":"of referencing data stored in the token DB table."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Notifications Impact"}],"source_content_type":"text/x-rst","patch_set":2,"id":"3ae8d1ca_0072d448","line":85,"in_reply_to":"3ae8d1ca_d73fe623","updated":"2014-05-29 00:35:19.000000000","message":"Hah! Good catch, duplicated wording. Yes, that should be less wordy.","commit_id":"59cd4e6c3ae1fe0f3fe8c9db999b9c6d8e2c65b0"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":10,"context_line":""},{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"},{"line_number":12,"context_line":"implement a token provider that does not require enumeration of the tokens"},{"line_number":13,"context_line":"for revocation purposes. This change will eliminate the need for persisting"},{"line_number":14,"context_line":"token information to a backend (SQL, Memcached, etc)."},{"line_number":15,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_00eaf60e","line":12,"updated":"2014-05-31 06:34:20.000000000","message":"The phrasing \"does not require enumeration of the tokens\" is a little nuanced for someone not intimate with Keystone.  What you are trying to say is that with the addition of the revocation events system Keystone no longer has to rely on simply sending services a list of token identifiers for revoked tokens and these services do not have to check the list to determine if a token is still valid.  I would be much more verbose about how this used to actually work.","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"347df28ec9f9adfe6cba2f30ccf072c35feec1b6","unresolved":false,"context_lines":[{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":10,"context_line":""},{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"},{"line_number":12,"context_line":"implement a token provider that does not require enumeration of the tokens"},{"line_number":13,"context_line":"for revocation purposes. This change will eliminate the need for persisting"},{"line_number":14,"context_line":"token information to a backend (SQL, Memcached, etc)."},{"line_number":15,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_5373f1e8","line":12,"in_reply_to":"1ae5cdf2_00eaf60e","updated":"2014-05-31 21:57:30.000000000","message":"Done in next patch","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"},{"line_number":12,"context_line":"implement a token provider that does not require enumeration of the tokens"},{"line_number":13,"context_line":"for revocation purposes. This change will eliminate the need for persisting"},{"line_number":14,"context_line":"token information to a backend (SQL, Memcached, etc)."},{"line_number":15,"context_line":""},{"line_number":16,"context_line":"Existing Launchpad BP::"},{"line_number":17,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_40423e11","line":14,"updated":"2014-05-31 06:34:20.000000000","message":"I would add another sentence explaining why revocation events enable us to no longer have to persist token information.  \"Since we now have revocation events, we no longer have to persist token information because...\"","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"347df28ec9f9adfe6cba2f30ccf072c35feec1b6","unresolved":false,"context_lines":[{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"},{"line_number":12,"context_line":"implement a token provider that does not require enumeration of the tokens"},{"line_number":13,"context_line":"for revocation purposes. This change will eliminate the need for persisting"},{"line_number":14,"context_line":"token information to a backend (SQL, Memcached, etc)."},{"line_number":15,"context_line":""},{"line_number":16,"context_line":"Existing Launchpad BP::"},{"line_number":17,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_73770dd5","line":14,"in_reply_to":"1ae5cdf2_40423e11","updated":"2014-05-31 21:57:30.000000000","message":"Done in next patch","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":21,"context_line":"Problem Description"},{"line_number":22,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":23,"context_line":""},{"line_number":24,"context_line":"Currently there is significant overhead in managing the persistence backend for"},{"line_number":25,"context_line":"the Token data. This administrative and performance overhead manifests in"},{"line_number":26,"context_line":"a number of ways:"},{"line_number":27,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_005f56e6","line":24,"updated":"2014-05-31 06:34:20.000000000","message":"Great problem description","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":39,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":42,"context_line":"stable storage. When using the new non-persistent token provider, UUID will"},{"line_number":43,"context_line":"not be available."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_8072867a","line":42,"updated":"2014-05-31 06:34:20.000000000","message":"Should make it clear that this does not imply we are deprecating UUID. Just that that token model cannot benefit from this optimization","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"347df28ec9f9adfe6cba2f30ccf072c35feec1b6","unresolved":false,"context_lines":[{"line_number":39,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":42,"context_line":"stable storage. When using the new non-persistent token provider, UUID will"},{"line_number":43,"context_line":"not be available."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_937499d0","line":42,"in_reply_to":"1ae5cdf2_8072867a","updated":"2014-05-31 21:57:30.000000000","message":"Done in next patch","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":56,"context_line":"Data Model Impact"},{"line_number":57,"context_line":"-----------------"},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":60,"context_line":"object (ensuring consistent model extracted from the versioned-token-format)."},{"line_number":61,"context_line":"The token db table will be maintained to accommodate backwards compatibility"},{"line_number":62,"context_line":"and use of the providers (notably UUID) that require persistence."}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_80f9a6ca","line":59,"updated":"2014-05-31 06:34:20.000000000","message":"More detail on what you mean by a uniform object would be helpful.  perhaps a simple example of the uniform object?","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"347df28ec9f9adfe6cba2f30ccf072c35feec1b6","unresolved":false,"context_lines":[{"line_number":56,"context_line":"Data Model Impact"},{"line_number":57,"context_line":"-----------------"},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":60,"context_line":"object (ensuring consistent model extracted from the versioned-token-format)."},{"line_number":61,"context_line":"The token db table will be maintained to accommodate backwards compatibility"},{"line_number":62,"context_line":"and use of the providers (notably UUID) that require persistence."}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_6ecac450","line":59,"in_reply_to":"1ae5cdf2_80f9a6ca","updated":"2014-05-31 21:57:30.000000000","message":"I\u0027ll just be more explicit that until now each token version was drastically different (dict formats)","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":64,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":65,"context_line":"object format."},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"JSON Schema for each token version will need to be developed for complete"},{"line_number":68,"context_line":"validation as part of the conversion from the versioned-token-format to the"},{"line_number":69,"context_line":"internal token object."},{"line_number":70,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_203292a7","line":67,"updated":"2014-05-31 06:34:20.000000000","message":"We didn\u0027t already have JSON schemas for each token version?","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"68f008959381dd76b12b369c8bdf25a454b35e29","unresolved":false,"context_lines":[{"line_number":64,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":65,"context_line":"object format."},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"JSON Schema for each token version will need to be developed for complete"},{"line_number":68,"context_line":"validation as part of the conversion from the versioned-token-format to the"},{"line_number":69,"context_line":"internal token object."},{"line_number":70,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_6ed1e4fa","line":67,"in_reply_to":"1ae5cdf2_203292a7","updated":"2014-05-31 21:58:16.000000000","message":"Nope. Nothing to validate with specifically. We ... don\u0027t even really document token format well as is.","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"bd025028dcbed6e1284960ee918da92922938285","unresolved":false,"context_lines":[{"line_number":81,"context_line":"use and format of the token should not be impacted."},{"line_number":82,"context_line":""},{"line_number":83,"context_line":"All cases of PKI tokens will be decoded (in a similar fashion to auth_token"},{"line_number":84,"context_line":"middleware) in memory instead of referencing data stored in the token DB "},{"line_number":85,"context_line":"persistence store."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_e017caeb","line":84,"updated":"2014-05-31 06:34:20.000000000","message":"remove white space (red char)","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"347df28ec9f9adfe6cba2f30ccf072c35feec1b6","unresolved":false,"context_lines":[{"line_number":81,"context_line":"use and format of the token should not be impacted."},{"line_number":82,"context_line":""},{"line_number":83,"context_line":"All cases of PKI tokens will be decoded (in a similar fashion to auth_token"},{"line_number":84,"context_line":"middleware) in memory instead of referencing data stored in the token DB "},{"line_number":85,"context_line":"persistence store."},{"line_number":86,"context_line":""},{"line_number":87,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1ae5cdf2_336e6502","line":84,"in_reply_to":"1ae5cdf2_e017caeb","updated":"2014-05-31 21:57:30.000000000","message":"*grr* white space!","commit_id":"b0066d4136dcd453239fb9c42215910315e5e2c3"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":5,"context_line":" http://creativecommons.org/licenses/by/3.0/legalcode"},{"line_number":6,"context_line":""},{"line_number":7,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":8,"context_line":"Non Persistent Tokens"},{"line_number":9,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":10,"context_line":""},{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_20e97fb8","line":8,"updated":"2014-06-02 20:57:00.000000000","message":"Non-Persistent","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":10,"context_line":""},{"line_number":11,"context_line":"With the addition of the revocation events system it is now possible to"},{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"id based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID, etc) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token ids generated by"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_a0dcaf13","line":13,"updated":"2014-06-02 20:57:00.000000000","message":"ID","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":13,"context_line":"id based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID, etc) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token ids generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token"},{"line_number":18,"context_line":"middleware`` on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token ids that are revoked."}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_c0d7fbeb","line":16,"updated":"2014-06-02 20:57:00.000000000","message":"IDs","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID, etc) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token ids generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token"},{"line_number":18,"context_line":"middleware`` on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token ids that are revoked."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"`current bp ephemeral-pki-tokens"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_20009f66","line":18,"updated":"2014-06-02 20:57:00.000000000","message":"``auth_token`` middleware","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token ids generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token"},{"line_number":18,"context_line":"middleware`` on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token ids that are revoked."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"`current bp ephemeral-pki-tokens"},{"line_number":22,"context_line":"\u003chttps://blueprints.launchpad.net/keystone/+spec/ephemeral-pki-tokens\u003e`_"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_4003eb72","line":19,"updated":"2014-06-02 20:57:00.000000000","message":"IDs","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":35,"context_line":"  the required enumeration of each user\u0027s tokens to handle revocations under"},{"line_number":36,"context_line":"  the Token Revocation List. Even with the revocation events, the overhead"},{"line_number":37,"context_line":"  in running the backing store can be significant while providing little"},{"line_number":38,"context_line":"  added value"},{"line_number":39,"context_line":""},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"Proposed Change"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_c0ee1bb4","line":38,"updated":"2014-06-02 20:57:00.000000000","message":"period.","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":69,"context_line":"change will allow for better control of data provided within a token and"},{"line_number":70,"context_line":"matching against data within a token (regardless of version)."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"The token db table will be maintained to accommodate backwards compatibility"},{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_80697327","line":72,"updated":"2014-06-02 20:57:00.000000000","message":"DB","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":93,"context_line":""},{"line_number":94,"context_line":"All cases of PKI tokens will be decoded (in a similar fashion to auth_token"},{"line_number":95,"context_line":"middleware) in memory instead of referencing data stored in the token DB"},{"line_number":96,"context_line":"persistence store."},{"line_number":97,"context_line":""},{"line_number":98,"context_line":""},{"line_number":99,"context_line":"Notifications Impact"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_40884be7","line":96,"updated":"2014-06-02 20:57:00.000000000","message":"the ability to use the token backend for auditing will be lost, and we\u0027ll have to rely on the new notifications support","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":106,"context_line":""},{"line_number":107,"context_line":"When the new provider is used (non-persistent-token-provider) only PKI tokens"},{"line_number":108,"context_line":"will be issued, and the short-hash version of the token cannot be used to"},{"line_number":109,"context_line":"interact with Keystone."},{"line_number":110,"context_line":""},{"line_number":111,"context_line":"Performance Impact"},{"line_number":112,"context_line":"------------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_6055474f","line":109,"updated":"2014-06-02 20:57:00.000000000","message":"that seems like API impact, sort of","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":117,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":118,"context_line":"stored outside of the PKI signed token data."},{"line_number":119,"context_line":""},{"line_number":120,"context_line":"Other Deployer Impact"},{"line_number":121,"context_line":"---------------------"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"Deployers will need to leverage the Revocation Events system instead of the"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_c0a61b53","line":120,"updated":"2014-06-02 20:57:00.000000000","message":"i think this section should enumerate new config options","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":121,"context_line":"---------------------"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"Deployers will need to leverage the Revocation Events system instead of the"},{"line_number":124,"context_line":"Token Revocation List to utilize the non-persistent-token-provider."},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"Developer Impact"},{"line_number":127,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_40d68bc6","line":124,"updated":"2014-06-02 20:57:00.000000000","message":"-why-the-extra-hyphens?","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":126,"context_line":"Developer Impact"},{"line_number":127,"context_line":"----------------"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"All references to token_api.get_token() (excluding the UUID token provider)"},{"line_number":130,"context_line":"will cease to be available. Direct lookup of token data without decoding"},{"line_number":131,"context_line":"the PKI token (when PKI tokens are used) will not be available."},{"line_number":132,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_a0e66fb2","line":129,"updated":"2014-06-02 20:57:00.000000000","message":"``backtick all the things``","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"0a2ac023f9fe991b14630bb4013f714a848aa2f0","unresolved":false,"context_lines":[{"line_number":126,"context_line":"Developer Impact"},{"line_number":127,"context_line":"----------------"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"All references to token_api.get_token() (excluding the UUID token provider)"},{"line_number":130,"context_line":"will cease to be available. Direct lookup of token data without decoding"},{"line_number":131,"context_line":"the PKI token (when PKI tokens are used) will not be available."},{"line_number":132,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_16f5a9ed","line":129,"in_reply_to":"1ae5cdf2_a0e66fb2","updated":"2014-06-02 21:55:36.000000000","message":"``All`` ``references`` ``to`` ``token_api.get_token`` ``(````excluding`` ``the`` ``UUID`` ``token provider````)`` ... ? :P","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"30afb4dbfbe35b4de884a79f0c061494613d2658","unresolved":false,"context_lines":[{"line_number":196,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":197,"context_line":""},{"line_number":198,"context_line":"* Under \"Token Revocation\"::"},{"line_number":199,"context_line":"  http://dolphm.com/openstack-icehouse-design-summit-outcomes-for-keystone/"},{"line_number":200,"context_line":""},{"line_number":201,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/reuse-token"},{"line_number":202,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"1ae5cdf2_206fbffe","line":199,"updated":"2014-06-02 20:57:00.000000000","message":"i made a useful thing!","commit_id":"e408e2ef79b8222c9b808b2cbe592a14807527b1"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"b560b4458daacd464171067fdaba7e3fd34588fd","unresolved":false,"context_lines":[{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token IDs that are revoked."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"`current bp non-persistent-tokens"},{"line_number":22,"context_line":"\u003chttps://blueprints.launchpad.net/keystone/+spec/non-persistent-tokens\u003e`_"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_e99f7bb2","line":19,"updated":"2014-06-03 03:05:06.000000000","message":"Great description!","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"6a192e7dd148f26dd5aa16e4b5ff9cd0d825f069","unresolved":false,"context_lines":[{"line_number":29,"context_line":"a number of ways:"},{"line_number":30,"context_line":""},{"line_number":31,"context_line":"* SQL Backend sees significant bloat in the DB causing poor performance,"},{"line_number":32,"context_line":"  significant consumption of storage resources."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"* Memcached (and largely all Key-Value-Store backends) show poor handling of"},{"line_number":35,"context_line":"  the required enumeration of each user\u0027s tokens to handle revocations under"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_9d30ede9","line":32,"updated":"2014-06-03 06:30:39.000000000","message":"and processing resources (just sayin\u0027)","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"c93be4a6960ac9e49f7911980b2f5482c45800d6","unresolved":false,"context_lines":[{"line_number":29,"context_line":"a number of ways:"},{"line_number":30,"context_line":""},{"line_number":31,"context_line":"* SQL Backend sees significant bloat in the DB causing poor performance,"},{"line_number":32,"context_line":"  significant consumption of storage resources."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"* Memcached (and largely all Key-Value-Store backends) show poor handling of"},{"line_number":35,"context_line":"  the required enumeration of each user\u0027s tokens to handle revocations under"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_d7a23ffc","line":32,"in_reply_to":"1ae5cdf2_9d30ede9","updated":"2014-06-03 15:56:11.000000000","message":"I\u0027ll add that little verbiage.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"6a192e7dd148f26dd5aa16e4b5ff9cd0d825f069","unresolved":false,"context_lines":[{"line_number":62,"context_line":"-----------------"},{"line_number":63,"context_line":""},{"line_number":64,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":65,"context_line":"object moving away from a dictionary with different layout per token version."},{"line_number":66,"context_line":"This change to a uniform model will eliminate the complex conditional code"},{"line_number":67,"context_line":"to lookup the same data across each of the token versions (currently v2 and"},{"line_number":68,"context_line":"v3 tokens have drastically different schemas for the same information). This"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_9d470d4b","line":65,"updated":"2014-06-03 06:30:39.000000000","message":"nit: with *a* different","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"c93be4a6960ac9e49f7911980b2f5482c45800d6","unresolved":false,"context_lines":[{"line_number":62,"context_line":"-----------------"},{"line_number":63,"context_line":""},{"line_number":64,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":65,"context_line":"object moving away from a dictionary with different layout per token version."},{"line_number":66,"context_line":"This change to a uniform model will eliminate the complex conditional code"},{"line_number":67,"context_line":"to lookup the same data across each of the token versions (currently v2 and"},{"line_number":68,"context_line":"v3 tokens have drastically different schemas for the same information). This"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_b79db3be","line":65,"in_reply_to":"1ae5cdf2_9d470d4b","updated":"2014-06-03 15:56:11.000000000","message":"Done in next patch.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"6a192e7dd148f26dd5aa16e4b5ff9cd0d825f069","unresolved":false,"context_lines":[{"line_number":67,"context_line":"to lookup the same data across each of the token versions (currently v2 and"},{"line_number":68,"context_line":"v3 tokens have drastically different schemas for the same information). This"},{"line_number":69,"context_line":"change will allow for better control of data provided within a token and"},{"line_number":70,"context_line":"matching against data within a token (regardless of version)."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"The token DB table will be maintained to accommodate backwards compatibility"},{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_5d64a5e4","line":70,"updated":"2014-06-03 06:30:39.000000000","message":"+++++","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"b560b4458daacd464171067fdaba7e3fd34588fd","unresolved":false,"context_lines":[{"line_number":70,"context_line":"matching against data within a token (regardless of version)."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"The token DB table will be maintained to accommodate backwards compatibility"},{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":76,"context_line":"object format."}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_097e676b","line":73,"updated":"2014-06-03 03:05:06.000000000","message":"Nice clarification","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"6a192e7dd148f26dd5aa16e4b5ff9cd0d825f069","unresolved":false,"context_lines":[{"line_number":103,"context_line":"--------------------"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"None"},{"line_number":106,"context_line":""},{"line_number":107,"context_line":"Other End User Impact"},{"line_number":108,"context_line":"---------------------"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_bdea090d","line":106,"updated":"2014-06-03 06:30:39.000000000","message":"nit: needs a blank line here (I think we need 2 blank line between sections)","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"c93be4a6960ac9e49f7911980b2f5482c45800d6","unresolved":false,"context_lines":[{"line_number":103,"context_line":"--------------------"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"None"},{"line_number":106,"context_line":""},{"line_number":107,"context_line":"Other End User Impact"},{"line_number":108,"context_line":"---------------------"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_57670fe9","line":106,"in_reply_to":"1ae5cdf2_bdea090d","updated":"2014-06-03 15:56:11.000000000","message":"Done in next patch.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"b560b4458daacd464171067fdaba7e3fd34588fd","unresolved":false,"context_lines":[{"line_number":134,"context_line":"----------------"},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"All references to ``token_api.get_token`` (excluding use in the"},{"line_number":137,"context_line":"``UUID token provider``) will cease to be removed. Direct lookup of token"},{"line_number":138,"context_line":"data without decoding the PKI token (when PKI tokens are used) will not be"},{"line_number":139,"context_line":"available."},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_29f883e4","line":137,"updated":"2014-06-03 03:05:06.000000000","message":"cease to be removed sounds funny. Is there a typo here?  Perhaps you meant \"will need to be removed?\"","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"c93be4a6960ac9e49f7911980b2f5482c45800d6","unresolved":false,"context_lines":[{"line_number":134,"context_line":"----------------"},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"All references to ``token_api.get_token`` (excluding use in the"},{"line_number":137,"context_line":"``UUID token provider``) will cease to be removed. Direct lookup of token"},{"line_number":138,"context_line":"data without decoding the PKI token (when PKI tokens are used) will not be"},{"line_number":139,"context_line":"available."},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_f76c9bd2","line":137,"in_reply_to":"1ae5cdf2_29f883e4","updated":"2014-06-03 15:56:11.000000000","message":"Good catch.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"6a192e7dd148f26dd5aa16e4b5ff9cd0d825f069","unresolved":false,"context_lines":[{"line_number":174,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"* This spec depends (for validation code path) on"},{"line_number":177,"context_line":"  https://blueprints.launchpad.net/keystone/+spec/keystone-api-validation"},{"line_number":178,"context_line":""},{"line_number":179,"context_line":"* This spec depends on full revocation event support (auth_token middleware)"},{"line_number":180,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_9de38d32","line":177,"updated":"2014-06-03 06:30:39.000000000","message":"i believe the url was change to just /api-validation","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"5c957f1e584f42ba3732f2b6a467d9ebd1613543","unresolved":false,"context_lines":[{"line_number":174,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"* This spec depends (for validation code path) on"},{"line_number":177,"context_line":"  https://blueprints.launchpad.net/keystone/+spec/keystone-api-validation"},{"line_number":178,"context_line":""},{"line_number":179,"context_line":"* This spec depends on full revocation event support (auth_token middleware)"},{"line_number":180,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_f4e01efc","line":177,"in_reply_to":"1ae5cdf2_9de38d32","updated":"2014-06-03 13:24:16.000000000","message":"++, thanks for catching this.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"c93be4a6960ac9e49f7911980b2f5482c45800d6","unresolved":false,"context_lines":[{"line_number":174,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":175,"context_line":""},{"line_number":176,"context_line":"* This spec depends (for validation code path) on"},{"line_number":177,"context_line":"  https://blueprints.launchpad.net/keystone/+spec/keystone-api-validation"},{"line_number":178,"context_line":""},{"line_number":179,"context_line":"* This spec depends on full revocation event support (auth_token middleware)"},{"line_number":180,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_9746f74d","line":177,"in_reply_to":"1ae5cdf2_f4e01efc","updated":"2014-06-03 15:56:11.000000000","message":"Done in next patch.","commit_id":"cf6f3e92a3cdfb59ddda096766923b539b9d7485"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"75d78bc004cdd47ec15748b4dcdc5f9aedddcb1a","unresolved":false,"context_lines":[{"line_number":209,"context_line":""},{"line_number":210,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/reuse-token"},{"line_number":211,"context_line":""},{"line_number":212,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/ephemeral-pki-tokens"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1ae5cdf2_63d5ff1a","line":212,"updated":"2014-06-04 19:10:43.000000000","message":"The URL to the \"ephemeral-pki-tokens\" spec is not correct. Has it moved?","commit_id":"efb792766423fb0d6ede819858548129fda8e663"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"22302ec1a60c31381526ab5b3896e8f6bf6d9f33","unresolved":false,"context_lines":[{"line_number":209,"context_line":""},{"line_number":210,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/reuse-token"},{"line_number":211,"context_line":""},{"line_number":212,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/ephemeral-pki-tokens"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1ae5cdf2_3e2a2a90","line":212,"in_reply_to":"1ae5cdf2_63d5ff1a","updated":"2014-06-04 20:03:14.000000000","message":"Ah Good catch, yes.","commit_id":"efb792766423fb0d6ede819858548129fda8e663"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"ID based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID, etc) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_55eca7f3","line":15,"updated":"2014-06-10 01:17:51.000000000","message":"remove \"etc\" since it\u0027s already a list of examples.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"ID based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID, etc) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_1ca0ed8e","line":15,"in_reply_to":"1ae5cdf2_55eca7f3","updated":"2014-06-13 17:55:47.000000000","message":"Done","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token IDs that are revoked."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"`current bp non-persistent-tokens"},{"line_number":22,"context_line":"\u003chttps://blueprints.launchpad.net/keystone/+spec/non-persistent-tokens\u003e`_"},{"line_number":23,"context_line":""},{"line_number":24,"context_line":"Problem Description"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_95d62f26","line":21,"updated":"2014-06-10 01:17:51.000000000","message":"what does \"current\" mean here? suggest remove it.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"},{"line_number":19,"context_line":"of token IDs that are revoked."},{"line_number":20,"context_line":""},{"line_number":21,"context_line":"`current bp non-persistent-tokens"},{"line_number":22,"context_line":"\u003chttps://blueprints.launchpad.net/keystone/+spec/non-persistent-tokens\u003e`_"},{"line_number":23,"context_line":""},{"line_number":24,"context_line":"Problem Description"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_fca421a7","line":21,"in_reply_to":"1ae5cdf2_95d62f26","updated":"2014-06-13 17:55:47.000000000","message":"Done","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":44,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":45,"context_line":"stable storage. When using the new non-persistent token provider, UUID and"},{"line_number":46,"context_line":"short-hashed versions of PKI tokens will not be supported. The UUID and"},{"line_number":47,"context_line":"standard PKI token providers will not be deprecated in the ``Juno`` development"},{"line_number":48,"context_line":"cycle."},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_f594d345","line":47,"updated":"2014-06-10 01:17:51.000000000","message":"why not?","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":44,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":45,"context_line":"stable storage. When using the new non-persistent token provider, UUID and"},{"line_number":46,"context_line":"short-hashed versions of PKI tokens will not be supported. The UUID and"},{"line_number":47,"context_line":"standard PKI token providers will not be deprecated in the ``Juno`` development"},{"line_number":48,"context_line":"cycle."},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"Alternatives"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_dc9a0561","line":47,"in_reply_to":"1ae5cdf2_f594d345","updated":"2014-06-13 17:55:47.000000000","message":"Added more verbiage for this to add clarity","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":61,"context_line":"Data Model Impact"},{"line_number":62,"context_line":"-----------------"},{"line_number":63,"context_line":""},{"line_number":64,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":65,"context_line":"object moving away from a dictionary with a different layout per token version."},{"line_number":66,"context_line":"This change to a uniform model will eliminate the complex conditional code"},{"line_number":67,"context_line":"to lookup the same data across each of the token versions (currently v2 and"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_7588c322","line":64,"updated":"2014-06-10 01:17:51.000000000","message":"what\u0027s a \"uniform object\"? is it different from an \"object\"?","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":61,"context_line":"Data Model Impact"},{"line_number":62,"context_line":"-----------------"},{"line_number":63,"context_line":""},{"line_number":64,"context_line":"The general data model for the token (in-memory) will change to be a uniform"},{"line_number":65,"context_line":"object moving away from a dictionary with a different layout per token version."},{"line_number":66,"context_line":"This change to a uniform model will eliminate the complex conditional code"},{"line_number":67,"context_line":"to lookup the same data across each of the token versions (currently v2 and"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_5c6fd55f","line":64,"in_reply_to":"1ae5cdf2_7588c322","updated":"2014-06-13 17:55:47.000000000","message":"reworded to be more clear.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":70,"context_line":"matching against data within a token (regardless of version)."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"The token DB table will be maintained to accommodate backwards compatibility"},{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":76,"context_line":"object format."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_f5a21399","line":73,"updated":"2014-06-10 01:17:51.000000000","message":"This looks like it\u0027s describing a non-impact, not an impact.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":70,"context_line":"matching against data within a token (regardless of version)."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"The token DB table will be maintained to accommodate backwards compatibility"},{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":76,"context_line":"object format."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_1c79cd2e","line":73,"in_reply_to":"1ae5cdf2_f5a21399","updated":"2014-06-13 17:55:47.000000000","message":"Reworded.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":76,"context_line":"object format."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"JSON Schema for each token version will need to be developed for complete"},{"line_number":79,"context_line":"validation as part of the conversion from the versioned-token-format to the"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_d5ea77ab","line":76,"updated":"2014-06-10 01:17:51.000000000","message":"this is scary just due to the size of the token table. How about supporting the old formats? (Suggest add to alternatives section.)\n\nDoes this only happen if ephemeral tokens are wanted? (If not, maybe should have a separate spec for consistent token format?)","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":73,"context_line":"and use of the providers (notably UUID) that require persistence."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The token data table will be migrated to support the new consistent token"},{"line_number":76,"context_line":"object format."},{"line_number":77,"context_line":""},{"line_number":78,"context_line":"JSON Schema for each token version will need to be developed for complete"},{"line_number":79,"context_line":"validation as part of the conversion from the versioned-token-format to the"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_fc7d011d","line":76,"in_reply_to":"1ae5cdf2_d5ea77ab","updated":"2014-06-13 17:55:47.000000000","message":"Changed this part of the spec to serialize into the current schema.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":84,"context_line":""},{"line_number":85,"context_line":"When the new provider is used (non-persistent-token-provider) only PKI tokens"},{"line_number":86,"context_line":"will be issued, and the short-hash version of the token cannot be used to"},{"line_number":87,"context_line":"interact with Keystone."},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Security Impact"},{"line_number":90,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_b5fecb69","line":87,"updated":"2014-06-10 01:17:51.000000000","message":"what will the server respond with if you use the token hash?","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":84,"context_line":""},{"line_number":85,"context_line":"When the new provider is used (non-persistent-token-provider) only PKI tokens"},{"line_number":86,"context_line":"will be issued, and the short-hash version of the token cannot be used to"},{"line_number":87,"context_line":"interact with Keystone."},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"Security Impact"},{"line_number":90,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_7c8ad11b","line":87,"in_reply_to":"1ae5cdf2_b5fecb69","updated":"2014-06-13 17:55:47.000000000","message":"Done","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":11333,"name":"David J Hu","email":"david.hu@hpe.com","username":"dhu"},"change_message_id":"dde8f7e2213be8199e5c77308e7b3dec1033daac","unresolved":false,"context_lines":[{"line_number":96,"context_line":"middleware) in memory instead of referencing data stored in the token DB"},{"line_number":97,"context_line":"persistence store. The ability to use the token backend for auditing will no"},{"line_number":98,"context_line":"longer be available. Instead audit will rely exclusively on the ``CADF``"},{"line_number":99,"context_line":"authentication notifications."},{"line_number":100,"context_line":""},{"line_number":101,"context_line":""},{"line_number":102,"context_line":"Notifications Impact"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_31efcf1f","line":99,"updated":"2014-06-12 04:32:08.000000000","message":"Does this present a security risk if memory dump attack had occurred?","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":96,"context_line":"middleware) in memory instead of referencing data stored in the token DB"},{"line_number":97,"context_line":"persistence store. The ability to use the token backend for auditing will no"},{"line_number":98,"context_line":"longer be available. Instead audit will rely exclusively on the ``CADF``"},{"line_number":99,"context_line":"authentication notifications."},{"line_number":100,"context_line":""},{"line_number":101,"context_line":""},{"line_number":102,"context_line":"Notifications Impact"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_1c92ad56","line":99,"in_reply_to":"1ae5cdf2_31efcf1f","updated":"2014-06-13 17:55:47.000000000","message":"No more risk than what is currently implemented.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":114,"context_line":"------------------"},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_95154f9d","line":117,"updated":"2014-06-10 01:17:51.000000000","message":"change \"Token\" to \"token\" (lowercase)","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":114,"context_line":"------------------"},{"line_number":115,"context_line":""},{"line_number":116,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_dcacc5a1","line":117,"in_reply_to":"1ae5cdf2_95154f9d","updated":"2014-06-13 17:55:47.000000000","message":"Done","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":116,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."},{"line_number":121,"context_line":""},{"line_number":122,"context_line":"Other Deployer Impact"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_15213ffc","line":119,"updated":"2014-06-10 01:17:51.000000000","message":"is this new or did it work before? I assume it worked before (with v3 anyways)","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":116,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."},{"line_number":121,"context_line":""},{"line_number":122,"context_line":"Other Deployer Impact"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_bca9b991","line":119,"in_reply_to":"1ae5cdf2_15213ffc","updated":"2014-06-13 17:55:47.000000000","message":"This works now, it\u0027s just a requirement to do the full decode instead of asking the DB. It\u0027s a change in where the data is sourced, no change in functionality.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":220,"name":"Haneef Ali","email":"haneef.ali@hp.com","username":"haneef"},"change_message_id":"9a4258d508eefd6510a17b5a33a707cc6fbd6917","unresolved":false,"context_lines":[{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."},{"line_number":121,"context_line":""},{"line_number":122,"context_line":"Other Deployer Impact"},{"line_number":123,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_d848952f","line":120,"updated":"2014-06-11 22:38:20.000000000","message":"How about all identity api calls? keystone any how needs to decode them. There is no middleware involved here","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":117,"context_line":"consumption by not requiring persistence of Token data. Tokens will need to be"},{"line_number":118,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":119,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":120,"context_line":"stored outside of the PKI signed token data."},{"line_number":121,"context_line":""},{"line_number":122,"context_line":"Other Deployer Impact"},{"line_number":123,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_7cbc91dc","line":120,"in_reply_to":"1ae5cdf2_d848952f","updated":"2014-06-13 17:55:47.000000000","message":"AuthContextMiddleware is the internal-keystone middleware to handle this. See under \"work items\"","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":124,"context_line":""},{"line_number":125,"context_line":"To leverage the non-persistent tokens deployers will need to do the following:"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_d586f7da","line":127,"updated":"2014-06-10 01:17:51.000000000","message":"Is Keystone going to raise an error if it\u0027s not configured correctly? i.e., with both revocation list support and ephemeral token provider?\n\nAlso, should be able to document how this is done.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":124,"context_line":""},{"line_number":125,"context_line":"To leverage the non-persistent tokens deployers will need to do the following:"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_3cabe975","line":127,"in_reply_to":"1ae5cdf2_d586f7da","updated":"2014-06-13 17:55:47.000000000","message":"Added more info.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":220,"name":"Haneef Ali","email":"haneef.ali@hp.com","username":"haneef"},"change_message_id":"9a4258d508eefd6510a17b5a33a707cc6fbd6917","unresolved":false,"context_lines":[{"line_number":125,"context_line":"To leverage the non-persistent tokens deployers will need to do the following:"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""},{"line_number":131,"context_line":"No new configuration options will be added."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_78230158","line":128,"updated":"2014-06-11 22:38:20.000000000","message":"is the token revocation event  really feasible in production?\nI don\u0027t think we have system bus where every service listens. Unless such an infrastructure is required, we can\u0027t rely on events.\n\nEven if we rely on events, those events are not signed and it has security issues","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":125,"context_line":"To leverage the non-persistent tokens deployers will need to do the following:"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""},{"line_number":131,"context_line":"No new configuration options will be added."}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_1c6f0d93","line":128,"in_reply_to":"1ae5cdf2_78230158","updated":"2014-06-13 17:55:47.000000000","message":"This is a concern that comes from using the revocation events, not really in scope for this proposal. The revocation events are handled by polling for events from keystone (now) and will be including notifications. Signed notifications are another issue all together (much larger issue) that will need to be addressed by KDC (kite under barbican)","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""},{"line_number":131,"context_line":"No new configuration options will be added."},{"line_number":132,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_95caaf18","line":129,"updated":"2014-06-10 01:17:51.000000000","message":"What\u0027s the impact on auth_token? Is it going to automatically work by discovering if the server is configured to use ephemeral tokens?\n\nAlso, should be able to document how this is done.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":126,"context_line":""},{"line_number":127,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":128,"context_line":""},{"line_number":129,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":130,"context_line":""},{"line_number":131,"context_line":"No new configuration options will be added."},{"line_number":132,"context_line":""}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_9cbb1dc3","line":129,"in_reply_to":"1ae5cdf2_95caaf18","updated":"2014-06-13 17:55:47.000000000","message":"Added more info. This does depend on auth_token having revocation event support. (dependencies)","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"1c29b970cc8468d54bb16ea049abe7e9d1a8f9f9","unresolved":false,"context_lines":[{"line_number":204,"context_line":"References"},{"line_number":205,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":206,"context_line":""},{"line_number":207,"context_line":"* Under \"Token Revocation\"::"},{"line_number":208,"context_line":"  http://dolphm.com/openstack-icehouse-design-summit-outcomes-for-keystone/"},{"line_number":209,"context_line":""},{"line_number":210,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/reuse-token"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_95bf0f60","line":207,"updated":"2014-06-10 01:17:51.000000000","message":"does the :: without a blank line afterwards render properly?","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"bda04859daec5c9b9ec178c52524d2070ec58019","unresolved":false,"context_lines":[{"line_number":204,"context_line":"References"},{"line_number":205,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":206,"context_line":""},{"line_number":207,"context_line":"* Under \"Token Revocation\"::"},{"line_number":208,"context_line":"  http://dolphm.com/openstack-icehouse-design-summit-outcomes-for-keystone/"},{"line_number":209,"context_line":""},{"line_number":210,"context_line":"* https://blueprints.launchpad.net/keystone/+spec/reuse-token"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1ae5cdf2_3cc20934","line":207,"in_reply_to":"1ae5cdf2_95bf0f60","updated":"2014-06-13 17:55:47.000000000","message":"The :: should have been 1 : it\u0027s not meant to be a code block. Fixed.","commit_id":"8b4a20e89c12316eaab0c48289a98838abdb539a"},{"author":{"_account_id":220,"name":"Haneef Ali","email":"haneef.ali@hp.com","username":"haneef"},"change_message_id":"914bb4aa78d2c17a28cc757b8d456e4b870f89ef","unresolved":false,"context_lines":[{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"ID based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_72d96255","line":15,"updated":"2014-06-19 01:29:13.000000000","message":"What if the user has 2 tokens and one is revoked.?  In this case how will you identify the token?  Won\u0027t you rely on ID.  Who keeps the ID of the token that is revoked when I call DELETE token","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1916,"name":"Guang Yee","email":"gyee@suse.com","username":"guang-yee"},"change_message_id":"d31563b93ed457c7283e587b00c40f8979ae83f5","unresolved":false,"context_lines":[{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"ID based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_7b6f76c4","line":15,"in_reply_to":"1ae5cdf2_72d96255","updated":"2014-06-19 07:15:56.000000000","message":"If I am reading the sentence correctly, seem like revoke by token ID is not supported? Only revoke by class?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":12,"context_line":"implement a token provider that does not require the ability to lookup a token"},{"line_number":13,"context_line":"ID based upon data contained within the token (e.g. revoking all tokens issued"},{"line_number":14,"context_line":"to a specific User ID). The new revocation event uses data values from a class"},{"line_number":15,"context_line":"of tokens (e.g. User ID, Domain ID, Project ID) to determine if a token"},{"line_number":16,"context_line":"is revoked instead of relying on an explicit list of token IDs generated by"},{"line_number":17,"context_line":"Keystone. The new revocation matching is performed within the ``auth_token``"},{"line_number":18,"context_line":"middleware on each token instead of within Keystone to generate the list"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_ee5e1912","line":15,"in_reply_to":"1ae5cdf2_7b6f76c4","updated":"2014-06-22 02:59:54.000000000","message":"Revocation events never reference the token ID, they can revoke (in theory) a token by the expiration time (unique to parent/child tokens). There really is almost no reason to revoke by ID, almost every event we actually want to revoke for would be by class.\n\nRevocation events (iirc) already has support for revoke by expiration.\n\n@Haneef, please look below at my other (big) comment about how revocation events occur. Let me know and we can discuss in IRC (with adam young) the Revocation Events (merged) and how they are to work.\n\nIn short, the TRL (token revocation list) has a lot of short comings and issues. Revocation events aims to solve it. Revocation event specific logic is *out of scope* of this specification.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"63066e63d3b64cdbbd8ba8914a9feb05d331113d","unresolved":false,"context_lines":[{"line_number":24,"context_line":"Problem Description"},{"line_number":25,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":26,"context_line":""},{"line_number":27,"context_line":"Currently there is significant overhead in managing the persistence backend for"},{"line_number":28,"context_line":"the Token data. This administrative and performance overhead manifests in"},{"line_number":29,"context_line":"a number of ways:"},{"line_number":30,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_2d70d072","line":27,"updated":"2014-06-18 20:43:59.000000000","message":"Wrap text at 79 chars","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"1204d6d190e72c1ae5bfeab7da26b06176b1d2fc","unresolved":false,"context_lines":[{"line_number":24,"context_line":"Problem Description"},{"line_number":25,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":26,"context_line":""},{"line_number":27,"context_line":"Currently there is significant overhead in managing the persistence backend for"},{"line_number":28,"context_line":"the Token data. This administrative and performance overhead manifests in"},{"line_number":29,"context_line":"a number of ways:"},{"line_number":30,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_c97d36b4","line":27,"in_reply_to":"1ae5cdf2_2d70d072","updated":"2014-06-18 21:16:23.000000000","message":"The keystone template (https://github.com/openstack/keystone-specs/blob/master/specs/template.rst) says wrap text at 79 chars but I think it should be 80 chars right?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"640f65d830e32148b76d3d6980627d9671596180","unresolved":false,"context_lines":[{"line_number":24,"context_line":"Problem Description"},{"line_number":25,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":26,"context_line":""},{"line_number":27,"context_line":"Currently there is significant overhead in managing the persistence backend for"},{"line_number":28,"context_line":"the Token data. This administrative and performance overhead manifests in"},{"line_number":29,"context_line":"a number of ways:"},{"line_number":30,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_58b18eff","line":27,"in_reply_to":"1ae5cdf2_c97d36b4","updated":"2014-06-22 16:42:09.000000000","message":"This is already wrapping at 79 characters. The \u0027Newline\u0027 is the 80th character. The statement of 79 characters is meant to indicate 79 characters + newline (how the hacking check for normal code works).","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"63066e63d3b64cdbbd8ba8914a9feb05d331113d","unresolved":false,"context_lines":[{"line_number":44,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":45,"context_line":"stable storage. When using the new non-persistent token provider, UUID and"},{"line_number":46,"context_line":"short-hashed versions of PKI tokens will not be supported. The UUID and"},{"line_number":47,"context_line":"standard PKI token providers will not be deprecated in the ``Juno`` development"},{"line_number":48,"context_line":"cycle. Deprecation of the original PKI token providers (PKI and PKIZ) will"},{"line_number":49,"context_line":"occur as early as the K cycle; this delay in deprecation allows for operators"},{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_cd084cdc","line":47,"updated":"2014-06-18 20:43:59.000000000","message":"Wrap text at 79 chars","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"640f65d830e32148b76d3d6980627d9671596180","unresolved":false,"context_lines":[{"line_number":44,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":45,"context_line":"stable storage. When using the new non-persistent token provider, UUID and"},{"line_number":46,"context_line":"short-hashed versions of PKI tokens will not be supported. The UUID and"},{"line_number":47,"context_line":"standard PKI token providers will not be deprecated in the ``Juno`` development"},{"line_number":48,"context_line":"cycle. Deprecation of the original PKI token providers (PKI and PKIZ) will"},{"line_number":49,"context_line":"occur as early as the K cycle; this delay in deprecation allows for operators"},{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_78b48af0","line":47,"in_reply_to":"1ae5cdf2_895ece43","updated":"2014-06-22 16:42:09.000000000","message":"Same as above.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"1204d6d190e72c1ae5bfeab7da26b06176b1d2fc","unresolved":false,"context_lines":[{"line_number":44,"context_line":"The change will be to eliminate the need to persist tokens to any form of"},{"line_number":45,"context_line":"stable storage. When using the new non-persistent token provider, UUID and"},{"line_number":46,"context_line":"short-hashed versions of PKI tokens will not be supported. The UUID and"},{"line_number":47,"context_line":"standard PKI token providers will not be deprecated in the ``Juno`` development"},{"line_number":48,"context_line":"cycle. Deprecation of the original PKI token providers (PKI and PKIZ) will"},{"line_number":49,"context_line":"occur as early as the K cycle; this delay in deprecation allows for operators"},{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_895ece43","line":47,"in_reply_to":"1ae5cdf2_cd084cdc","updated":"2014-06-18 21:16:23.000000000","message":"The keystone template (https://github.com/openstack/keystone-specs/blob/master/specs/template.rst) says wrap text at 79 chars but I think it should be 80 chars right?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1091,"name":"Arun Kant","email":"arukant@cisco.com","username":"arunkant-uws"},"change_message_id":"5758dbbe4b4e515a84f05e8e31454135c12f870d","unresolved":false,"context_lines":[{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"},{"line_number":51,"context_line":"default deployment model. The UUID token provider will be evaluated for"},{"line_number":52,"context_line":"deprecation based upon how many deployments continue to utilize UUID over PKI"},{"line_number":53,"context_line":"tokens."},{"line_number":54,"context_line":""},{"line_number":55,"context_line":"Alternatives"},{"line_number":56,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_f563a117","line":53,"updated":"2014-06-18 23:18:33.000000000","message":"Is this bp saying that both tokens and revocation event list would not be stored in DB on keystone side? If yes, then \n\nHow openstack service(s) will know about revocation events? \n   -- Are we expecting all services to subscribe to keystone notifications (revocation events) which means creating their own queues per service (may be per service node, az etc)? These can be too many queues.\n   -- What happens if new service node is added, how it will become aware of recent revocation events?\n   -- Is each openstack service will have its notification listener implementation or is it something going to provided by keystone?\n   -- Are these services going to consume those events and maintain those events lists internally in memory? If yes, then how nodes get recent list of revocation events after its restart?\n   -- If revocation event is not delivered and handled successfully at service side (listener), it can result in unauthorized access issue. So keystone notification reliability, prompt delivery becomes quite critical from transport provider (Rabbitmq, qpid etc.) perspective.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":220,"name":"Haneef Ali","email":"haneef.ali@hp.com","username":"haneef"},"change_message_id":"914bb4aa78d2c17a28cc757b8d456e4b870f89ef","unresolved":false,"context_lines":[{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"},{"line_number":51,"context_line":"default deployment model. The UUID token provider will be evaluated for"},{"line_number":52,"context_line":"deprecation based upon how many deployments continue to utilize UUID over PKI"},{"line_number":53,"context_line":"tokens."},{"line_number":54,"context_line":""},{"line_number":55,"context_line":"Alternatives"},{"line_number":56,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_b2c92a5c","line":53,"in_reply_to":"1ae5cdf2_afdaa5e6","updated":"2014-06-19 01:29:13.000000000","message":"What do you mean by active event?  If there exists a token which expires in next 4 hours and it is revoked.  What constitutes an active event  if the client calls in the next hour","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1916,"name":"Guang Yee","email":"gyee@suse.com","username":"guang-yee"},"change_message_id":"d31563b93ed457c7283e587b00c40f8979ae83f5","unresolved":false,"context_lines":[{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"},{"line_number":51,"context_line":"default deployment model. The UUID token provider will be evaluated for"},{"line_number":52,"context_line":"deprecation based upon how many deployments continue to utilize UUID over PKI"},{"line_number":53,"context_line":"tokens."},{"line_number":54,"context_line":""},{"line_number":55,"context_line":"Alternatives"},{"line_number":56,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_bbe3be03","line":53,"in_reply_to":"1ae5cdf2_b2c92a5c","updated":"2014-06-19 07:15:56.000000000","message":"Well, we just eliminated *real time validation* option with this approach, unless we can consume the events real time by implementing a lightweight event listener.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"},{"line_number":51,"context_line":"default deployment model. The UUID token provider will be evaluated for"},{"line_number":52,"context_line":"deprecation based upon how many deployments continue to utilize UUID over PKI"},{"line_number":53,"context_line":"tokens."},{"line_number":54,"context_line":""},{"line_number":55,"context_line":"Alternatives"},{"line_number":56,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_ce3cfdc1","line":53,"in_reply_to":"1ae5cdf2_bbe3be03","updated":"2014-06-22 02:59:54.000000000","message":"@Guang, we don\u0027t really do \"real time validation\" with PKI as is. we poll for the TRL. The listener would be beyond this scope and would be an improvement that the revocation events has over the revocation list. We effectively eliminated real time validation when we use the full PKI token ids (using the short hashes is bad for a number of reasons, see recent django_openstack_auth issues)\n\n@Haneef,\n\nThe Keystone server has a list of currently active revocation events (you need to go look at the revocation event system to get a full understanding), but in short it handles revocations based upon elements within the Token data (e.g. user_id, domain_id, etc). Instead of enumerating the tokens like we do with persistent backends, we use the events.\n\nThe Events are requested in the same manner as the current revocation list would be. The middleware says \"hey what are the revocation events\" and keystone responds with a list (there are more complex actions such as \"give me revocations that occured since XXX time\" etc). The middleware then uses logic to determine if the revocation event matches a token, if the token matches *any* revocation event, it is revoked.\n\nThe fundamental difference here is that instead of a *list* of token ids that are revoked, we have a list of events.\n\nThe long term goal is to not only have the poll for current revocation events (events only need to last for the  maximum expiration time of the token + small window), but also listen for events via a listener and consuming notifications. An \"active\" event is simply an event that has not expired (was issued more recently than *now - maximum_token_life*. The initial implementation will poll for events as follows:\n\n1. When the middleware is started (e.g. thread initialized), the complete list of events is fetched from keystone (Same as the revocation list today).\n\n2. On a fixed interval the middleware will update it\u0027s cached copy of the revocation events (same as it updates the revocation list today).\n\n3. (future) A notification of a revocation event will be consumed, updating the list for closer to real time revocations (revocation list cannot and will not do this, the volume of notifications would be overwhelming in many cases)","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":50,"context_line":"and deployers to adopt the new non-persistent tokens before it becomes the"},{"line_number":51,"context_line":"default deployment model. The UUID token provider will be evaluated for"},{"line_number":52,"context_line":"deprecation based upon how many deployments continue to utilize UUID over PKI"},{"line_number":53,"context_line":"tokens."},{"line_number":54,"context_line":""},{"line_number":55,"context_line":"Alternatives"},{"line_number":56,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_afdaa5e6","line":53,"in_reply_to":"1ae5cdf2_f563a117","updated":"2014-06-19 01:14:58.000000000","message":"Events are stored server side (in keystone), tokens are not, nor will there be an enumerated list of the tokens.\n\nThe revocation events are polled for on an interval and will (eventually) emit notifications as well. If a service doesn\u0027t subscribe for notifications or misses one, the standard poll interval (same exposure as the enumerated token revocation list) will get that service caught up on the revocations that are active.\n\nAt any startup, the service will ask keystone (same as it does for the revocation list today) for all active events to ensure it is up-to-date.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1916,"name":"Guang Yee","email":"gyee@suse.com","username":"guang-yee"},"change_message_id":"d31563b93ed457c7283e587b00c40f8979ae83f5","unresolved":false,"context_lines":[{"line_number":61,"context_line":"any persistence of the token. Long term, these incremental improvements would"},{"line_number":62,"context_line":"drive towards non-persistent tokens, as that becomes the simplest/most"},{"line_number":63,"context_line":"maintainable solution (no risk of data loss or data corruption in the"},{"line_number":64,"context_line":"backing store, no long-term memory/storage overhead, etc)."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"Data Model Impact"},{"line_number":67,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_bbcade6b","line":64,"updated":"2014-06-19 07:15:56.000000000","message":"How about token reuse? Why do we have to issue a different token every time?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":61,"context_line":"any persistence of the token. Long term, these incremental improvements would"},{"line_number":62,"context_line":"drive towards non-persistent tokens, as that becomes the simplest/most"},{"line_number":63,"context_line":"maintainable solution (no risk of data loss or data corruption in the"},{"line_number":64,"context_line":"backing store, no long-term memory/storage overhead, etc)."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"Data Model Impact"},{"line_number":67,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_33194a88","line":64,"in_reply_to":"1ae5cdf2_bbcade6b","updated":"2014-06-22 02:59:54.000000000","message":"We advocate token-reuse now. It doesn\u0027t get done often. I don\u0027t see it as an alternative.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":5707,"name":"Henry Nash","email":"henryn@linux.vnet.ibm.com","username":"henry-nash"},"change_message_id":"2ceac66627a81823ecc95d76032b5d6bf0f7ac98","unresolved":false,"context_lines":[{"line_number":66,"context_line":"Data Model Impact"},{"line_number":67,"context_line":"-----------------"},{"line_number":68,"context_line":""},{"line_number":69,"context_line":"The general data model for the token (in-memory) will change to be a token"},{"line_number":70,"context_line":"version agnostic memory construct. This will be a departure from the current"},{"line_number":71,"context_line":"Python dictionary that is used (and varies in structure based upon token"},{"line_number":72,"context_line":"version). This change will eliminate some of the more complex conditional code"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_69a095f4","line":69,"updated":"2014-06-19 05:06:52.000000000","message":"It\u0027s not quite cleat (at least to me) why the new data model is required....why wouldn\u0027t just restricting Keystone to issuing PKI tokens plus auth middleware always using revocation events be enough to achieve the stated goal?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":66,"context_line":"Data Model Impact"},{"line_number":67,"context_line":"-----------------"},{"line_number":68,"context_line":""},{"line_number":69,"context_line":"The general data model for the token (in-memory) will change to be a token"},{"line_number":70,"context_line":"version agnostic memory construct. This will be a departure from the current"},{"line_number":71,"context_line":"Python dictionary that is used (and varies in structure based upon token"},{"line_number":72,"context_line":"version). This change will eliminate some of the more complex conditional code"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_ae41713a","line":69,"in_reply_to":"1ae5cdf2_69a095f4","updated":"2014-06-22 02:59:54.000000000","message":"Because we need need to decode the token, cleanly within keystone. There is little reason to not include this work as the non-persistence changes will make for a ton more code. This is really to help ensure cleanliness of new code instead of having a ton more conditional \"if v2 do X, if v3 do Y\" since we _never_ lookup the tokens once this is implemented.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"6fc69819f14f23f9dffb39441e5f52011a63f0e6","unresolved":false,"context_lines":[{"line_number":69,"context_line":"The general data model for the token (in-memory) will change to be a token"},{"line_number":70,"context_line":"version agnostic memory construct. This will be a departure from the current"},{"line_number":71,"context_line":"Python dictionary that is used (and varies in structure based upon token"},{"line_number":72,"context_line":"version). This change will eliminate some of the more complex conditional code"},{"line_number":73,"context_line":"that is used to support the different token versions now."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The new in-memory token model will be serialized as closely as possible to"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_21dbaa22","line":72,"updated":"2014-06-15 22:04:41.000000000","message":"This is such a cool benefit that the  ugly conditional code will be removed","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1091,"name":"Arun Kant","email":"arukant@cisco.com","username":"arunkant-uws"},"change_message_id":"5758dbbe4b4e515a84f05e8e31454135c12f870d","unresolved":false,"context_lines":[{"line_number":70,"context_line":"version agnostic memory construct. This will be a departure from the current"},{"line_number":71,"context_line":"Python dictionary that is used (and varies in structure based upon token"},{"line_number":72,"context_line":"version). This change will eliminate some of the more complex conditional code"},{"line_number":73,"context_line":"that is used to support the different token versions now."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The new in-memory token model will be serialized as closely as possible to"},{"line_number":76,"context_line":"the SQL token backend\u0027s schema to eliminate the need for a DB migration. Other"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_70dddf6f","line":73,"updated":"2014-06-18 23:18:33.000000000","message":"Is there a sample how this new in-memory data model looks like? \n\nWhatever format/ structure is defined may get impacted with new requirements e.g. hierarchical multitenacy (unless json schema has additionalproperties as true at every level).","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":70,"context_line":"version agnostic memory construct. This will be a departure from the current"},{"line_number":71,"context_line":"Python dictionary that is used (and varies in structure based upon token"},{"line_number":72,"context_line":"version). This change will eliminate some of the more complex conditional code"},{"line_number":73,"context_line":"that is used to support the different token versions now."},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"The new in-memory token model will be serialized as closely as possible to"},{"line_number":76,"context_line":"the SQL token backend\u0027s schema to eliminate the need for a DB migration. Other"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_8f132964","line":73,"in_reply_to":"1ae5cdf2_70dddf6f","updated":"2014-06-19 01:14:58.000000000","message":"The issue with \"with additional properties at every level\" is that the tokens already suffer from bloat. Any and all forms of data being added to the token should be closely regulated. The in-memory format will be very simple object + properties with a method to emit a versioned token format.\n\nThe JSONSchema (listed below) will be used for deserializing the unpacked PKI JSON data and validating.\n\nIf we need to support the \"wild west of whatever anyone wants to put in the tokens\" we have a problem. There is no reason a token needs to contain every ounce of information we can cram in, it should provide exactly what we need and no more.  I do not think the hierarchy data will end up in the token (as proposed nor as we described at the summit).\n\nAny and all new requirements for the token data absolutely should be vetted and approved before they are allowed.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6460,"name":"Brad Topol","email":"btopol@us.ibm.com","username":"btopol"},"change_message_id":"6fc69819f14f23f9dffb39441e5f52011a63f0e6","unresolved":false,"context_lines":[{"line_number":79,"context_line":"a seamless upgrade path to non-persistent tokens."},{"line_number":80,"context_line":""},{"line_number":81,"context_line":"JSON Schema for each token version will need to be developed for complete"},{"line_number":82,"context_line":"validation as part of the conversion from the versioned-token-format to the"},{"line_number":83,"context_line":"internal token object."},{"line_number":84,"context_line":""},{"line_number":85,"context_line":"REST API Impact"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_41e0b6cc","line":82,"updated":"2014-06-15 22:04:41.000000000","message":"Another great benefit!  JSON Schemas!  This deal gets better and better!","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"2d375bdb92e4a54033514ad4810ff27ff6c0e1de","unresolved":false,"context_lines":[{"line_number":88,"context_line":"When the new provider is used (non-persistent-token-provider) only PKI tokens"},{"line_number":89,"context_line":"will be issued, and the short-hash version of the token cannot be used to"},{"line_number":90,"context_line":"interact with Keystone. The server should respond with an HTTP 404 if the"},{"line_number":91,"context_line":"short hash is used in the request raising the TokenNotFound exception."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Security Impact"},{"line_number":94,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_d08cf3d9","line":91,"updated":"2014-06-18 23:46:22.000000000","message":"For v3, the tokens are in the X-Auth-Token or X-Subject-Token header, not in the URL path, so the resource exists. 404 should be used only when the resource (as identified by the path) doesn\u0027t exist. A 403 response would be more appropriate.\n\nAlso, is the client supposed to know that the request was rejected because the server doesn\u0027t accept hashed tokens anymore? Because it doesn\u0027t look from this like the client can tell.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":88,"context_line":"When the new provider is used (non-persistent-token-provider) only PKI tokens"},{"line_number":89,"context_line":"will be issued, and the short-hash version of the token cannot be used to"},{"line_number":90,"context_line":"interact with Keystone. The server should respond with an HTTP 404 if the"},{"line_number":91,"context_line":"short hash is used in the request raising the TokenNotFound exception."},{"line_number":92,"context_line":""},{"line_number":93,"context_line":"Security Impact"},{"line_number":94,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_f2b49237","line":91,"in_reply_to":"1ae5cdf2_d08cf3d9","updated":"2014-06-19 01:14:58.000000000","message":"I will clarify and update this to match. 100% agree with the first statement.\n\nAs to the second part, we wont be issuing short hashes for anything, 404 seems correct, the resource does not exist. We could use an alternate code (NotImplented, Gone, etc) if we wanted to differentiate (easy to detect if it\u0027s a PKI or PKIZ token)","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1916,"name":"Guang Yee","email":"gyee@suse.com","username":"guang-yee"},"change_message_id":"d31563b93ed457c7283e587b00c40f8979ae83f5","unresolved":false,"context_lines":[{"line_number":115,"context_line":"None"},{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Performance Impact"},{"line_number":118,"context_line":"------------------"},{"line_number":119,"context_line":""},{"line_number":120,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":121,"context_line":"consumption by not requiring persistence of token data. Tokens will need to be"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_1b2bd2eb","line":118,"updated":"2014-06-19 07:15:56.000000000","message":"Migration Impact?\n\nCan we have both non-persistent and persistent running in the same environment? It would be a major setback if we can\u0027t do rolling (zero downtime) upgrade.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":115,"context_line":"None"},{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Performance Impact"},{"line_number":118,"context_line":"------------------"},{"line_number":119,"context_line":""},{"line_number":120,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":121,"context_line":"consumption by not requiring persistence of token data. Tokens will need to be"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_4e78ed8e","line":118,"in_reply_to":"1ae5cdf2_1b2bd2eb","updated":"2014-06-22 02:59:54.000000000","message":"You cannot run non-persistence and persistence at once. How could you? This has no impact on the middleware or non-keystone services and only changes how keystone works internally.\n\nIf you want persistent tokens don\u0027t change the provider. If you want to run revocation events and token revocation lists at the same time (*not part of this spec, it\u0027s independent of this change*) there is no reason provided the middleware can support it.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":8978,"name":"Marek Denis","email":"marek.denis+openstack@gmail.com","username":"marek-denis"},"change_message_id":"86c7632dcedbc5086911c5bf21cf84a4937046a4","unresolved":false,"context_lines":[{"line_number":119,"context_line":""},{"line_number":120,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":121,"context_line":"consumption by not requiring persistence of token data. Tokens will need to be"},{"line_number":122,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":123,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":124,"context_line":"stored outside of the PKI signed token data."},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_9602eb70","line":122,"updated":"2014-06-16 13:04:09.000000000","message":"v2.0 what..Identity API ?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":119,"context_line":""},{"line_number":120,"context_line":"Generally speaking this change should reduce load on the system and resource"},{"line_number":121,"context_line":"consumption by not requiring persistence of token data. Tokens will need to be"},{"line_number":122,"context_line":"decoded via CMS, raising the CPU cost of handling tokens via the v2.0"},{"line_number":123,"context_line":"GET and HEAD (validate and check) mechanisms because the data will not be"},{"line_number":124,"context_line":"stored outside of the PKI signed token data."},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_7278827e","line":122,"in_reply_to":"1ae5cdf2_9602eb70","updated":"2014-06-19 01:14:58.000000000","message":"Yes, corrected in next patch.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"2d375bdb92e4a54033514ad4810ff27ff6c0e1de","unresolved":false,"context_lines":[{"line_number":131,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"    * Code to detect mis-configuration will be required. If Keystone is"},{"line_number":134,"context_line":"      mis-configured, a configuration exception will be raised."},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"    * Implement a new consistent configuration error exception for Keystone."},{"line_number":137,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_f031ef0b","line":134,"updated":"2014-06-18 23:46:22.000000000","message":"I think this belongs in work items and not in deployer impacts... is the deployer supposed to write some code?\n\nsame with the next sub-bullets","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":131,"context_line":"* Use revocation events instead of token revocation list"},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"    * Code to detect mis-configuration will be required. If Keystone is"},{"line_number":134,"context_line":"      mis-configured, a configuration exception will be raised."},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"    * Implement a new consistent configuration error exception for Keystone."},{"line_number":137,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_b295aa89","line":134,"in_reply_to":"1ae5cdf2_f031ef0b","updated":"2014-06-19 01:14:58.000000000","message":"++ moved in next patch.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"2d375bdb92e4a54033514ad4810ff27ff6c0e1de","unresolved":false,"context_lines":[{"line_number":140,"context_line":""},{"line_number":141,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* ``auth_token`` middleware will need to be configured to use revocation"},{"line_number":144,"context_line":"  events. This will be added to the deployment docs for OpenStack. No other"},{"line_number":145,"context_line":"  changes to ``auth_token`` middleware will be needed."},{"line_number":146,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_f063af17","line":143,"updated":"2014-06-18 23:46:22.000000000","message":"don\u0027t need `` here since auth_token is a thing and not a constant string.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":140,"context_line":""},{"line_number":141,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* ``auth_token`` middleware will need to be configured to use revocation"},{"line_number":144,"context_line":"  events. This will be added to the deployment docs for OpenStack. No other"},{"line_number":145,"context_line":"  changes to ``auth_token`` middleware will be needed."},{"line_number":146,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_d292369e","line":143,"in_reply_to":"1ae5cdf2_f063af17","updated":"2014-06-19 01:14:58.000000000","message":"Done in next patch","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"2d375bdb92e4a54033514ad4810ff27ff6c0e1de","unresolved":false,"context_lines":[{"line_number":141,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* ``auth_token`` middleware will need to be configured to use revocation"},{"line_number":144,"context_line":"  events. This will be added to the deployment docs for OpenStack. No other"},{"line_number":145,"context_line":"  changes to ``auth_token`` middleware will be needed."},{"line_number":146,"context_line":""},{"line_number":147,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_90742bd0","line":144,"updated":"2014-06-18 23:46:22.000000000","message":"should auth_token be able to tell if the server supports revocation events and use them automatically rather than having to be configured?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"8db35bb4dd97b434236b49e43502d6317f644475","unresolved":false,"context_lines":[{"line_number":141,"context_line":"* Set the token provider to the new non-persistent token provider"},{"line_number":142,"context_line":""},{"line_number":143,"context_line":"* ``auth_token`` middleware will need to be configured to use revocation"},{"line_number":144,"context_line":"  events. This will be added to the deployment docs for OpenStack. No other"},{"line_number":145,"context_line":"  changes to ``auth_token`` middleware will be needed."},{"line_number":146,"context_line":""},{"line_number":147,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_52ee2623","line":144,"in_reply_to":"1ae5cdf2_90742bd0","updated":"2014-06-19 01:14:58.000000000","message":"Extension discovery! :) If we can detect, yes, if we can\u0027t no. I\u0027ll update saying this is conditional on the revocation events and detection code.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"63066e63d3b64cdbbd8ba8914a9feb05d331113d","unresolved":false,"context_lines":[{"line_number":156,"context_line":"data without decoding the PKI token (when PKI tokens are used) will not be"},{"line_number":157,"context_line":"available."},{"line_number":158,"context_line":""},{"line_number":159,"context_line":"Any interaction with token data will be done in a consistent manner through the"},{"line_number":160,"context_line":"new uniform token object, eliminating the need for specific code handling a"},{"line_number":161,"context_line":"versioned token dictionary."},{"line_number":162,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_ed242862","line":159,"updated":"2014-06-18 20:43:59.000000000","message":"Wrap text at 79 chars","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"1204d6d190e72c1ae5bfeab7da26b06176b1d2fc","unresolved":false,"context_lines":[{"line_number":156,"context_line":"data without decoding the PKI token (when PKI tokens are used) will not be"},{"line_number":157,"context_line":"available."},{"line_number":158,"context_line":""},{"line_number":159,"context_line":"Any interaction with token data will be done in a consistent manner through the"},{"line_number":160,"context_line":"new uniform token object, eliminating the need for specific code handling a"},{"line_number":161,"context_line":"versioned token dictionary."},{"line_number":162,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_fc4d4e73","line":159,"in_reply_to":"1ae5cdf2_ed242862","updated":"2014-06-18 21:16:23.000000000","message":"The keystone template (https://github.com/openstack/keystone-specs/blob/master/specs/template.rst) says wrap text at 79 chars but I think it should be 80 chars right?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"640f65d830e32148b76d3d6980627d9671596180","unresolved":false,"context_lines":[{"line_number":156,"context_line":"data without decoding the PKI token (when PKI tokens are used) will not be"},{"line_number":157,"context_line":"available."},{"line_number":158,"context_line":""},{"line_number":159,"context_line":"Any interaction with token data will be done in a consistent manner through the"},{"line_number":160,"context_line":"new uniform token object, eliminating the need for specific code handling a"},{"line_number":161,"context_line":"versioned token dictionary."},{"line_number":162,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_18b706d9","line":159,"in_reply_to":"1ae5cdf2_fc4d4e73","updated":"2014-06-22 16:42:09.000000000","message":"Same as above. We might want to fix thetemplate to be mroe clear.","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":1916,"name":"Guang Yee","email":"gyee@suse.com","username":"guang-yee"},"change_message_id":"d31563b93ed457c7283e587b00c40f8979ae83f5","unresolved":false,"context_lines":[{"line_number":168,"context_line":"-----------"},{"line_number":169,"context_line":""},{"line_number":170,"context_line":"Primary assignee:"},{"line_number":171,"context_line":"  Morgan Fainberg (mdrnstm)"},{"line_number":172,"context_line":""},{"line_number":173,"context_line":"Other contributors:"},{"line_number":174,"context_line":"  Adam Young (ayoung)"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_bbcd7e5f","line":171,"updated":"2014-06-19 07:15:56.000000000","message":"mdrnstm? that your other handle? :)","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"4e64d98177fb450032eed99e374d01fe9bcd2283","unresolved":false,"context_lines":[{"line_number":168,"context_line":"-----------"},{"line_number":169,"context_line":""},{"line_number":170,"context_line":"Primary assignee:"},{"line_number":171,"context_line":"  Morgan Fainberg (mdrnstm)"},{"line_number":172,"context_line":""},{"line_number":173,"context_line":"Other contributors:"},{"line_number":174,"context_line":"  Adam Young (ayoung)"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_ae6891ba","line":171,"in_reply_to":"1ae5cdf2_bbcd7e5f","updated":"2014-06-22 02:59:54.000000000","message":"It is my launchpad ID. (and twitter if you didn\u0027t notice)","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"63066e63d3b64cdbbd8ba8914a9feb05d331113d","unresolved":false,"context_lines":[{"line_number":180,"context_line":""},{"line_number":181,"context_line":"* Token Validation / Conversion to new token object"},{"line_number":182,"context_line":""},{"line_number":183,"context_line":"* Keystone AuthContextMiddleware updated to always do CMS decode of PKI tokens."},{"line_number":184,"context_line":""},{"line_number":185,"context_line":"* Elimination of all calls to token_api.get_token except when using the UUID"},{"line_number":186,"context_line":"  token provider."}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_2d1e100e","line":183,"updated":"2014-06-18 20:43:59.000000000","message":"Wrap text at 79 chars","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"1204d6d190e72c1ae5bfeab7da26b06176b1d2fc","unresolved":false,"context_lines":[{"line_number":180,"context_line":""},{"line_number":181,"context_line":"* Token Validation / Conversion to new token object"},{"line_number":182,"context_line":""},{"line_number":183,"context_line":"* Keystone AuthContextMiddleware updated to always do CMS decode of PKI tokens."},{"line_number":184,"context_line":""},{"line_number":185,"context_line":"* Elimination of all calls to token_api.get_token except when using the UUID"},{"line_number":186,"context_line":"  token provider."}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_3c5416fd","line":183,"in_reply_to":"1ae5cdf2_2d1e100e","updated":"2014-06-18 21:16:23.000000000","message":"The keystone template (https://github.com/openstack/keystone-specs/blob/master/specs/template.rst) says wrap text at 79 chars but I think it should be 80 chars right?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"63066e63d3b64cdbbd8ba8914a9feb05d331113d","unresolved":false,"context_lines":[{"line_number":208,"context_line":""},{"line_number":209,"context_line":"* Initially an alternate test scenario will be needed to validate the non-"},{"line_number":210,"context_line":"  persistent token provider. Once revocation events becomes the default"},{"line_number":211,"context_line":"  deployment method, the non-persistent provider should become default as well."},{"line_number":212,"context_line":""},{"line_number":213,"context_line":""},{"line_number":214,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_6d18181f","line":211,"updated":"2014-06-18 20:43:59.000000000","message":"Wrap text at 79 chars","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":11045,"name":"Bob Thyne","email":"bob.thyne@hp.com","username":"bobt"},"change_message_id":"1204d6d190e72c1ae5bfeab7da26b06176b1d2fc","unresolved":false,"context_lines":[{"line_number":208,"context_line":""},{"line_number":209,"context_line":"* Initially an alternate test scenario will be needed to validate the non-"},{"line_number":210,"context_line":"  persistent token provider. Once revocation events becomes the default"},{"line_number":211,"context_line":"  deployment method, the non-persistent provider should become default as well."},{"line_number":212,"context_line":""},{"line_number":213,"context_line":""},{"line_number":214,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_5c59e233","line":211,"in_reply_to":"1ae5cdf2_6d18181f","updated":"2014-06-18 21:16:23.000000000","message":"The keystone template (https://github.com/openstack/keystone-specs/blob/master/specs/template.rst) says wrap text at 79 chars but I think it should be 80 chars right?","commit_id":"56d2b0380ad1b46271b1be7072193d8c7e0dcade"},{"author":{"_account_id":7232,"name":"Subbu Allamaraju","email":"subbu@subbu.org"},"change_message_id":"cf51235f0a49e1cc491b1685d10c09c44ea3dc32","unresolved":false,"context_lines":[{"line_number":37,"context_line":"  in running the backing store can be significant while providing little"},{"line_number":38,"context_line":"  added value."},{"line_number":39,"context_line":""},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"Proposed Change"},{"line_number":42,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":43,"context_line":""}],"source_content_type":"text/x-rst","patch_set":13,"id":"baada198_c22e1aa5","line":40,"updated":"2014-07-17 15:24:36.000000000","message":"Here is one more. Consider several OpenStack deployments in different data centers. To provide a consistent user experience, it makes sense to map each deployment as a \"region\" in Keystone (\"region\" is an unfortunate choice - AZ a la AWS is a better term) and share the same keystone between all those deployments.\n\nThis leads to a multi-master setup for Keystone. With both PKI or UUID tokens, replication overhead for a multi-master setup is high. Synchronous replication with PKI or UUID also increases latency for frequently used requests like authentication. At work, we\u0027ve considered ways to skip PKI replication, but had to abandon it since revoked tokens need to be replicated.\n\nEphemeral (or non-persistent) tokens will help us ease replication/latency overhead. This BP is a must have to build large globally distributed OpenStack clouds.","commit_id":"050cf5a942f79fde8d02a402d61763aeaab4c8a7"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"ea51813a0da827d171dec6837ee5eff8925de4b9","unresolved":false,"context_lines":[{"line_number":61,"context_line":"  any persistence of the token. Long term, these incremental improvements would"},{"line_number":62,"context_line":"  drive towards non-persistent tokens, as that becomes the simplest/most"},{"line_number":63,"context_line":"  maintainable solution (no risk of data loss or data corruption in the"},{"line_number":64,"context_line":"  backing store, no long-term memory/storage overhead, etc)."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"* Expand the reuse of valid tokens. It is currently advocated that tokens"},{"line_number":67,"context_line":"  should be reused where possible, however, there are a number of cases where"}],"source_content_type":"text/x-rst","patch_set":13,"id":"1ae5cdf2_df971fea","line":64,"updated":"2014-06-23 20:44:48.000000000","message":"Morgan - I think gyee was suggesting that keystone could hand out an existing token again rather than having clients cache tokens.","commit_id":"050cf5a942f79fde8d02a402d61763aeaab4c8a7"},{"author":{"_account_id":2903,"name":"Morgan Fainberg","email":"morgan.fainberg@gmail.com","username":"mdrnstm"},"change_message_id":"89f62fad8668e597bb0c0555ed87ed65c6686d73","unresolved":false,"context_lines":[{"line_number":61,"context_line":"  any persistence of the token. Long term, these incremental improvements would"},{"line_number":62,"context_line":"  drive towards non-persistent tokens, as that becomes the simplest/most"},{"line_number":63,"context_line":"  maintainable solution (no risk of data loss or data corruption in the"},{"line_number":64,"context_line":"  backing store, no long-term memory/storage overhead, etc)."},{"line_number":65,"context_line":""},{"line_number":66,"context_line":"* Expand the reuse of valid tokens. It is currently advocated that tokens"},{"line_number":67,"context_line":"  should be reused where possible, however, there are a number of cases where"}],"source_content_type":"text/x-rst","patch_set":13,"id":"1ae5cdf2_a8e13b2c","line":64,"in_reply_to":"1ae5cdf2_df971fea","updated":"2014-06-23 21:45:50.000000000","message":"Ah, I don\u0027t think it changes much of the verbiage here if keystone reissues tokens or clients cache.","commit_id":"050cf5a942f79fde8d02a402d61763aeaab4c8a7"},{"author":{"_account_id":6486,"name":"Brant Knudson","email":"blk@acm.org","username":"blk-u"},"change_message_id":"ea51813a0da827d171dec6837ee5eff8925de4b9","unresolved":false,"context_lines":[{"line_number":81,"context_line":"than having an ever expanding set of conditional code to handle the variations"},{"line_number":82,"context_line":"on token structure across an increasing number of versions. Both v2 and v3"},{"line_number":83,"context_line":"tokens have significant structural differences that will add significant"},{"line_number":84,"context_line":"complications to the code-base when eliminating the calls to .get_token."},{"line_number":85,"context_line":""},{"line_number":86,"context_line":"The new in-memory token model will be serialized as closely as possible to"},{"line_number":87,"context_line":"the SQL token backend\u0027s schema to eliminate the need for a DB migration. Other"}],"source_content_type":"text/x-rst","patch_set":13,"id":"1ae5cdf2_bf60930e","line":84,"updated":"2014-06-23 20:44:48.000000000","message":"seems like this could be a separate spec.","commit_id":"050cf5a942f79fde8d02a402d61763aeaab4c8a7"}]}
