)]}'
{"attic/v3/identity-api-v3.rst":[{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":2046,"context_line":""},{"line_number":2047,"context_line":"::"},{"line_number":2048,"context_line":""},{"line_number":2049,"context_line":"    GET /v3/auth/tokens?allow_expired\u003d1"},{"line_number":2050,"context_line":""},{"line_number":2051,"context_line":"Relationship:"},{"line_number":2052,"context_line":"``http://docs.openstack.org/api/openstack-identity/3/rel/auth_tokens``"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_0c715aa8","line":2049,"updated":"2016-10-07 16:03:12.000000000","message":"What if this was just: \n\n  Get /v3/auth/tokens?allow_expired\n\n`allow_expired` already portrays the meaning and we wouldn\u0027t need to assign it a value. Would `allow_expires\u003d0` be the default behavior? I think the absence of the query string would validate a token with expiry taken into consideration. Assigning a value to the query string allows for more permutations of the query string - which can lead to end user confusion (i.e. allow_expires\u003d0 and the absence of allow_expired essentially being the same thing).\n\nI\u0027d vote to just have the presence of `allow_expired` as a query string trigger the feature.","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":2046,"context_line":""},{"line_number":2047,"context_line":"::"},{"line_number":2048,"context_line":""},{"line_number":2049,"context_line":"    GET /v3/auth/tokens?allow_expired\u003d1"},{"line_number":2050,"context_line":""},{"line_number":2051,"context_line":"Relationship:"},{"line_number":2052,"context_line":"``http://docs.openstack.org/api/openstack-identity/3/rel/auth_tokens``"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_d1ce9f59","line":2049,"in_reply_to":"1a95cdbc_0c715aa8","updated":"2016-10-11 16:47:09.000000000","message":"I think the problem would be if you add a query parameter like this to the API, inevitably you\u0027ll get a client trying to set it `allow_expired\u003d0`. So I think we\u0027d still want to check it.","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"}],"specs/keystone/ocata/allow-expired.rst":[{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":31,"context_line":"continuous user authentication was to have the service reuse the same token"},{"line_number":32,"context_line":"when communicating with others. This works most of the time, however there is"},{"line_number":33,"context_line":"an edge case that if the token expiry period is crossed between when the user"},{"line_number":34,"context_line":"submits a token and the service passes the token on then token authentication"},{"line_number":35,"context_line":"will fail. At scale this edge case becomes a serious problem and is a negative"},{"line_number":36,"context_line":"experience for users because the operation may have been running a significant"},{"line_number":37,"context_line":"time before failure occurs. It also presents a difficult rollback operation for"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_d39c0a6c","line":34,"updated":"2016-10-04 19:19:46.000000000","message":"nit: token validation","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":32,"context_line":"when communicating with others. This works most of the time, however there is"},{"line_number":33,"context_line":"an edge case that if the token expiry period is crossed between when the user"},{"line_number":34,"context_line":"submits a token and the service passes the token on then token authentication"},{"line_number":35,"context_line":"will fail. At scale this edge case becomes a serious problem and is a negative"},{"line_number":36,"context_line":"experience for users because the operation may have been running a significant"},{"line_number":37,"context_line":"time before failure occurs. It also presents a difficult rollback operation for"},{"line_number":38,"context_line":"services because initial service to service requests may succeed which then"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_7395f63e","line":35,"updated":"2016-10-04 19:19:46.000000000","message":"nit: At scale,","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":45,"context_line":"A problem with this is because inter-service communication happens over public"},{"line_number":46,"context_line":"APIs there is no way to know when a user\u0027s request entered a deployment. Glance"},{"line_number":47,"context_line":"is unaware of whether a user wanted to list images or if nova is performing the"},{"line_number":48,"context_line":"operation on behalf of a user."},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"Proposed Change"},{"line_number":51,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_539e3a57","line":48,"updated":"2016-10-04 19:19:46.000000000","message":"+1 for the whole problem description","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":58,"context_line":"it along with an \"X-Service-Token\" with a service role (to distinguish it from"},{"line_number":59,"context_line":"any old user trying to use an expired token) we should take that to mean the"},{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_93c7e244","line":61,"updated":"2016-10-04 19:19:46.000000000","message":"+1","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configured on the server to prevent infinite reuse."}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_f303c674","line":63,"updated":"2016-10-04 19:19:46.000000000","message":"nit: you don\u0027t really need a value on this query string. ``?allow_expired`` is totally sufficient.","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":7191,"name":"Jamie Lennox","email":"jamielennox@gmail.com","username":"jamielennox"},"change_message_id":"09332b227f573e2c4cec8dc1548fec10b6a378eb","unresolved":false,"context_lines":[{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configured on the server to prevent infinite reuse."}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_aaa0d891","line":63,"in_reply_to":"1a95cdbc_f303c674","updated":"2016-10-04 23:40:16.000000000","message":"I\u0027ve never really liked this syntax. At one point we tried to come to a consensus - i don\u0027t remember what it was.","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configured on the server to prevent infinite reuse."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"In auth_token middleware when a valid (must not be expired) service token is"},{"line_number":69,"context_line":"presented with a request auth_token will make a validation request that may"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_933c4231","line":66,"updated":"2016-10-04 19:19:46.000000000","message":"nit: configurable* ; is there a use case for preventing \"infinite reuse\" or can this configuration default to null/\"forever\"?","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":7191,"name":"Jamie Lennox","email":"jamielennox@gmail.com","username":"jamielennox"},"change_message_id":"09332b227f573e2c4cec8dc1548fec10b6a378eb","unresolved":false,"context_lines":[{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configured on the server to prevent infinite reuse."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"In auth_token middleware when a valid (must not be expired) service token is"},{"line_number":69,"context_line":"presented with a request auth_token will make a validation request that may"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_6ab5f0cb","line":66,"in_reply_to":"1a95cdbc_933c4231","updated":"2016-10-04 23:40:16.000000000","message":"I\u0027m not sure, i figured it would be something people would want to control but we can always start it out without this option.","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configured on the server to prevent infinite reuse."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"In auth_token middleware when a valid (must not be expired) service token is"},{"line_number":69,"context_line":"presented with a request auth_token will make a validation request that may"},{"line_number":70,"context_line":"return an expired token and fill this data for use by the service."},{"line_number":71,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_33836ece","line":68,"updated":"2016-10-04 19:19:46.000000000","message":"nit: ``auth_token``","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":73,"context_line":"must also provide their current service token with a request to skip user token"},{"line_number":74,"context_line":"expiry checking."},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"Keystone vs auth_token middleware"},{"line_number":77,"context_line":"---------------------------------"},{"line_number":78,"context_line":""},{"line_number":79,"context_line":"There has been a few opinions on whether auth_token middleware should be"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_f35f4673","line":76,"updated":"2016-10-04 19:19:46.000000000","message":"This could be in the alternatives section.","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":82,"context_line":"validity of the service token."},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"In my opinion the addition of the ``allow_expired`` flag is not a security risk"},{"line_number":85,"context_line":"so long as it is explicit opt-in behaviour. Thus so long as auth_token"},{"line_number":86,"context_line":"middleware is only setting the flag in the appropriate place there is no"},{"line_number":87,"context_line":"additional need to have keystone make this determination."},{"line_number":88,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_335a8e84","line":85,"updated":"2016-10-04 19:19:46.000000000","message":"+1","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":92,"context_line":"keystone it would simply be \"validate_token: service_role:service\" for allow"},{"line_number":93,"context_line":"expired). We will therefore either need to add oslo.policy to auth_token"},{"line_number":94,"context_line":"middleware or provide a simple ``service_token_roles`` config option for"},{"line_number":95,"context_line":"validating the service token in middleware."},{"line_number":96,"context_line":""},{"line_number":97,"context_line":"Alternatives"},{"line_number":98,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_73541654","line":95,"updated":"2016-10-04 19:19:46.000000000","message":"Good point, I hadn\u0027t considered this.","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":102,"context_line":"Security Impact"},{"line_number":103,"context_line":"---------------"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"There is a definite"},{"line_number":106,"context_line":"This changes the security profile of OpenStack, it is actually to make"},{"line_number":107,"context_line":"it more secure.  A token validator needs to opt in to this scheme, and thus"},{"line_number":108,"context_line":"it cannot be used to unrevoke tokens etc.  The expection is that when one"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_73db36ee","line":105,"updated":"2016-10-04 19:19:46.000000000","message":"105Gdd","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":113,"context_line":""},{"line_number":114,"context_line":"The most important benefit will be from the shortening of the expiry"},{"line_number":115,"context_line":"times of tokens.  Tokens used as a proxy for authentication should be"},{"line_number":116,"context_line":"used within a few minutes of issuance.  THis change will allow that to happen."},{"line_number":117,"context_line":""},{"line_number":118,"context_line":"Notifications Impact"},{"line_number":119,"context_line":"--------------------"}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_d3e24a9f","line":116,"updated":"2016-10-04 19:19:46.000000000","message":"nit: This*","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":124,"context_line":"Other End User Impact"},{"line_number":125,"context_line":"---------------------"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"Directly hte change will be only in the validation API called by"},{"line_number":128,"context_line":"Middleware.  Indirectly, token expiration times will drop, and other"},{"line_number":129,"context_line":"services will need to react accordingly.  Horizon especially will have"},{"line_number":130,"context_line":"to re-authenticate far more often than it does now."}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_938602b8","line":127,"updated":"2016-10-04 19:19:46.000000000","message":"nit: the*","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":4,"name":"Dolph Mathews","email":"dolph.mathews@gmail.com","username":"dolph"},"change_message_id":"d038d8587e595440486f149f7b1388e847206e47","unresolved":false,"context_lines":[{"line_number":125,"context_line":"---------------------"},{"line_number":126,"context_line":""},{"line_number":127,"context_line":"Directly hte change will be only in the validation API called by"},{"line_number":128,"context_line":"Middleware.  Indirectly, token expiration times will drop, and other"},{"line_number":129,"context_line":"services will need to react accordingly.  Horizon especially will have"},{"line_number":130,"context_line":"to re-authenticate far more often than it does now."},{"line_number":131,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"1a95cdbc_f3bbc673","line":128,"updated":"2016-10-04 19:19:46.000000000","message":"nit: s/will/can/","commit_id":"f8e496827336c9ed074845a33985b836f8d0533a"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":34,"context_line":"submits a token and the service passes the token on then token validation"},{"line_number":35,"context_line":"will fail. At scale, this edge case becomes a serious problem and is a negative"},{"line_number":36,"context_line":"experience for users because the operation may have been running a significant"},{"line_number":37,"context_line":"time before failure occurs. It also presents a difficult rollback operation for"},{"line_number":38,"context_line":"services because initial service to service requests may succeed which then"},{"line_number":39,"context_line":"cannot be rolled back with the expired token."},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"We essentially need a way to have auth_token middleware give information for"},{"line_number":42,"context_line":"expired tokens to the services in situations where the token has already been"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_ac0c7043","line":39,"range":{"start_line":37,"start_character":28,"end_line":39,"end_character":45},"updated":"2016-10-11 16:47:09.000000000","message":"nit: An example of this type of rollback would be good.","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":38,"context_line":"services because initial service to service requests may succeed which then"},{"line_number":39,"context_line":"cannot be rolled back with the expired token."},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"We essentially need a way to have auth_token middleware give information for"},{"line_number":42,"context_line":"expired tokens to the services in situations where the token has already been"},{"line_number":43,"context_line":"validated by an existing trusted service."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"A problem with this is because inter-service communication happens over public"},{"line_number":46,"context_line":"APIs there is no way to know when a user\u0027s request entered a deployment. Glance"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_073295f1","line":43,"range":{"start_line":41,"start_character":0,"end_line":43,"end_character":41},"updated":"2016-10-11 16:47:09.000000000","message":"Solution, remove from Problem Description.","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":57,"context_line":"Keystone/auth_token middleware should support that if a token is submitted to"},{"line_number":58,"context_line":"it along with an \"X-Service-Token\" with a service role (to distinguish it from"},{"line_number":59,"context_line":"any old user trying to use an expired token) we should take that to mean the"},{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_6c6ade82","line":60,"updated":"2016-10-07 16:03:12.000000000","message":"How easy would it be for an attacker to exploit this? They would need a valid service token in order to do so - right?","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":57,"context_line":"Keystone/auth_token middleware should support that if a token is submitted to"},{"line_number":58,"context_line":"it along with an \"X-Service-Token\" with a service role (to distinguish it from"},{"line_number":59,"context_line":"any old user trying to use an expired token) we should take that to mean the"},{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_c7933da2","line":60,"range":{"start_line":60,"start_character":15,"end_line":60,"end_character":35},"updated":"2016-10-11 16:47:09.000000000","message":"In terms of validation, are we only talking about token expiration? Or, does it include revocation events as well?","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configurable on the server to prevent infinite reuse."}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_6c01feae","line":63,"updated":"2016-10-07 16:03:12.000000000","message":"I\u0027d prefer this to just be ``?allow_expired`` since it\u0027s simpler and eliminates unnecessary permutations of: \n\n  ?allow_expired\u003d1\n  ?allow_expired\u003dtrue\n  ?allow_expired\u003d0\n  ?allow_expires\u003dfalse","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":60,"context_line":"token has been previously validated by a service we trust and accept the token"},{"line_number":61,"context_line":"and fill the data expect by a service."},{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configurable on the server to prevent infinite reuse."}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_c7a09d93","line":63,"range":{"start_line":63,"start_character":44,"end_line":63,"end_character":57},"updated":"2016-10-11 16:47:09.000000000","message":"Why do we need a flag at all? Why not just make this the default; let operators opt-out (I\u0027m not even sure we even need an opt-out).","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":18338,"name":"Ron De Rose","email":"ronald.de.rose@intel.com","username":"derosenet"},"change_message_id":"9ea3f3804007895cbeaae0e45fbc15f5484c6821","unresolved":false,"context_lines":[{"line_number":62,"context_line":""},{"line_number":63,"context_line":"The keystone side of this is fairly simple. We add a flag ``?allow_expired\u003d1``"},{"line_number":64,"context_line":"to the existing ``GET /auth/tokens`` validation route that allows fetching an"},{"line_number":65,"context_line":"expired token. The length of how long an expired token should be fetchable for"},{"line_number":66,"context_line":"should be configurable on the server to prevent infinite reuse."},{"line_number":67,"context_line":""},{"line_number":68,"context_line":"In ``auth_token`` middleware when a valid (must not be expired) service token"},{"line_number":69,"context_line":"is presented with a request auth_token will make a validation request that may"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_076a9580","line":66,"range":{"start_line":65,"start_character":55,"end_line":66,"end_character":22},"updated":"2016-10-11 16:47:09.000000000","message":"nit: should be fetchable for a configurable amount of time...","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":100,"context_line":"Others"},{"line_number":101,"context_line":"++++++"},{"line_number":102,"context_line":""},{"line_number":103,"context_line":"Fixup trusts or OAuth to make them more generally usable."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"Security Impact"},{"line_number":106,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_ec96ee3c","line":103,"updated":"2016-10-07 16:03:12.000000000","message":"What is missing from trusts in order to accomplish this?","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":2218,"name":"Adam Young","email":"adam@younglogic.com","username":"ayoung"},"change_message_id":"3b81ec133f59d17f6291e4d8e242b2234d454f14","unresolved":false,"context_lines":[{"line_number":100,"context_line":"Others"},{"line_number":101,"context_line":"++++++"},{"line_number":102,"context_line":""},{"line_number":103,"context_line":"Fixup trusts or OAuth to make them more generally usable."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"Security Impact"},{"line_number":106,"context_line":"---------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"fa6399be_f9ce2182","line":103,"in_reply_to":"1a95cdbc_ec96ee3c","updated":"2016-10-11 20:42:46.000000000","message":"They were the means we discussed in the past to get around having to have long lived tokens.  IT really is not a viable alternative, but listed here for good measure.","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":121,"context_line":"--------------------"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"When a token is validated, it will need to annotate the use of this"},{"line_number":124,"context_line":"flag, both presence and absence."},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"Other End User Impact"},{"line_number":127,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_8c1baac9","line":124,"updated":"2016-10-07 16:03:12.000000000","message":"Can this be accomplished with the pyCADF format? Or do we need to add more stuff to the payload in general? If pyCADF supports this some how, we should consolidate all notification usage to be pyCADF in keystone (which feels like a prerequisite spec).","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":5046,"name":"Lance Bragstad","email":"lbragstad@redhat.com","username":"ldbragst"},"change_message_id":"2baf2bef14cc84e3a7a1e03c89a0970355451cc6","unresolved":false,"context_lines":[{"line_number":165,"context_line":""},{"line_number":166,"context_line":"UUID tokens stored in the database will require holding on to expired"},{"line_number":167,"context_line":"tokens in order to honor this change, which means providing a larger"},{"line_number":168,"context_line":"window for token flush. PKI tokens that are not stored in the database"},{"line_number":169,"context_line":"(or have been flushed) will require the full token body to honor this"},{"line_number":170,"context_line":"change."},{"line_number":171,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_6c91fef0","line":168,"updated":"2016-10-07 16:03:12.000000000","message":"Since PKI and PKIZ are being removed - do we need to consider this case in the spec?","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"},{"author":{"_account_id":6482,"name":"Steve Martinelli","email":"s.martinelli@gmail.com","username":"stevemar"},"change_message_id":"665026848ae6376c421c1bf5f67eeec1fa1c8553","unresolved":false,"context_lines":[{"line_number":192,"context_line":""},{"line_number":193,"context_line":"* Add flag to Keystone server API"},{"line_number":194,"context_line":""},{"line_number":195,"context_line":"* Add flag to keystoneclient which performs the token validation step."},{"line_number":196,"context_line":""},{"line_number":197,"context_line":"* Add step to keystonemiddleware that uses the above flags when a service token"},{"line_number":198,"context_line":"  is present."},{"line_number":199,"context_line":""},{"line_number":200,"context_line":"* Modify service to service communication to start passing a service token."},{"line_number":201,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a95cdbc_264d6ec4","line":198,"range":{"start_line":195,"start_character":0,"end_line":198,"end_character":13},"updated":"2016-10-07 05:32:19.000000000","message":"we\u0027ll need to bump the minimum level of these libs in ocata to match whatever level has this logic -- this is more of a note for myself than anything else","commit_id":"481015f8a4e7c6d8f3ef00eac9ad38140ad53582"}]}
