)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":13702,"name":"yushiro","username":"yushiro","inactive":true},"change_message_id":"2caafd3921b2d4a9eeeab633a0a8287166927d2c","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"This spec proposes Ironic Cinder integration. With this feature,"},{"line_number":10,"context_line":"user can boot bare metal node from Cinder volume."},{"line_number":11,"context_line":""},{"line_number":12,"context_line":"Change-Id: Ib0953852295e41eb9217ddead1a5845bf6efacee"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":2,"id":"7a016987_6be075cd","line":11,"updated":"2015-05-22 20:25:17.000000000","message":"Please specify the related blueprint like as follows:\n\n* blueprint xxx-yyy-zzz","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"}],"specs/kilo/cinder-integration.rst":[{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":13,"context_line":"This blueprint is a placeholder for integrating Ironic with Cinder. The idea"},{"line_number":14,"context_line":"is to get Bare Metal hosts booted from Cinder provisioned boot volumes."},{"line_number":15,"context_line":""},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"Problem description"},{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_9ff16803","line":16,"updated":"2015-01-27 21:36:48.000000000","message":"FYI- Water Boring submitted a similar blueprint a while back.  The wiki for his proposal is at https://wiki.openstack.org/wiki/Ironic/blueprints/cinder-integration","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":13,"context_line":"This blueprint is a placeholder for integrating Ironic with Cinder. The idea"},{"line_number":14,"context_line":"is to get Bare Metal hosts booted from Cinder provisioned boot volumes."},{"line_number":15,"context_line":""},{"line_number":16,"context_line":""},{"line_number":17,"context_line":"Problem description"},{"line_number":18,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":19,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_d52b6d2e","line":16,"in_reply_to":"1a930d6b_9ff16803","updated":"2015-01-28 08:02:09.000000000","message":"Yes, wiki which you mentioned is detailed description of the blueprint which I wrote above and bp has the link to the wiki. \n\nAlso I discussed this topic with Walter and we agreed that I take over Walter as he doesn\u0027t have enough time for this feature.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":5805,"name":"Chris Krelle","email":"nobodycam@gmail.com","username":"nobodycam"},"change_message_id":"d33bed14ebdd32b6e29d1b8a7d3515eb2c6114b3","unresolved":false,"context_lines":[{"line_number":30,"context_line":"  that of server"},{"line_number":31,"context_line":""},{"line_number":32,"context_line":"  e.g)Immediate recovery from failed bare metal server by just re-attaching"},{"line_number":33,"context_line":"  all volumes to alternative bare metal server"},{"line_number":34,"context_line":""},{"line_number":35,"context_line":"* **In particular, for OpenStack**, easier Windows bare metal instance"},{"line_number":36,"context_line":"  preparation by utilizing \"boot-from-volume\" functions"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_3560d000","line":33,"updated":"2015-01-23 22:50:10.000000000","message":"Note that this maybe more complicated.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":30,"context_line":"  that of server"},{"line_number":31,"context_line":""},{"line_number":32,"context_line":"  e.g)Immediate recovery from failed bare metal server by just re-attaching"},{"line_number":33,"context_line":"  all volumes to alternative bare metal server"},{"line_number":34,"context_line":""},{"line_number":35,"context_line":"* **In particular, for OpenStack**, easier Windows bare metal instance"},{"line_number":36,"context_line":"  preparation by utilizing \"boot-from-volume\" functions"}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_f5c9515b","line":33,"in_reply_to":"1a930d6b_3560d000","updated":"2015-01-28 08:02:09.000000000","message":"Right. I understand that it\u0027s not so simple but this is one of the reason that customer would like to use external block storage.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":5805,"name":"Chris Krelle","email":"nobodycam@gmail.com","username":"nobodycam"},"change_message_id":"d33bed14ebdd32b6e29d1b8a7d3515eb2c6114b3","unresolved":false,"context_lines":[{"line_number":38,"context_line":"Currently, OpenStack can provides bare metal server with Ironic but"},{"line_number":39,"context_line":"unfortunately, Ironic has no integration with OpenStack block storage project"},{"line_number":40,"context_line":"a.k.a Cinder. Ironic assumes that the bare metal node has it\u0027s own block"},{"line_number":41,"context_line":"storage device that Ironic deploys a bootable image onto. Cinder is the block"},{"line_number":42,"context_line":"storage project that knows how to provision volumes and create bootable"},{"line_number":43,"context_line":"volumes from images that live in Glance."},{"line_number":44,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_5012ca57","line":41,"updated":"2015-01-23 22:50:10.000000000","message":"depending on the driver used the image may or may not be bootable. There is effort underway to enable all drivers to support both whole-disk(bootable) and partition only images.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":38,"context_line":"Currently, OpenStack can provides bare metal server with Ironic but"},{"line_number":39,"context_line":"unfortunately, Ironic has no integration with OpenStack block storage project"},{"line_number":40,"context_line":"a.k.a Cinder. Ironic assumes that the bare metal node has it\u0027s own block"},{"line_number":41,"context_line":"storage device that Ironic deploys a bootable image onto. Cinder is the block"},{"line_number":42,"context_line":"storage project that knows how to provision volumes and create bootable"},{"line_number":43,"context_line":"volumes from images that live in Glance."},{"line_number":44,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_d5ec4dca","line":41,"in_reply_to":"1a930d6b_5012ca57","updated":"2015-01-28 08:02:09.000000000","message":"Thanks. I\u0027ll update the description","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10910,"name":"Stig Telfer","email":"stig.openstack@telfer.org","username":"stigtelfer"},"change_message_id":"fb1c6c786a60fec0ca5c2a5f181c59758a2f1db1","unresolved":false,"context_lines":[{"line_number":43,"context_line":"volumes from images that live in Glance."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"What we would like is to have Ironic use external block storage managed by"},{"line_number":46,"context_line":"Cinder as data and/or bootable volumes."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"To achive the goal above, Ironic needs to have following functions:"},{"line_number":49,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_afe3b431","line":46,"updated":"2015-02-03 15:05:36.000000000","message":"This raises an interesting point.  What are the differences between \"boot from volume\" and \"boot with data volume attached\"?  As you say, we would want to support both.  How does this affect what Ironic does with the volume parameters?","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"080dd54b060fd38834dd228919fbc0049bd94575","unresolved":false,"context_lines":[{"line_number":43,"context_line":"volumes from images that live in Glance."},{"line_number":44,"context_line":""},{"line_number":45,"context_line":"What we would like is to have Ironic use external block storage managed by"},{"line_number":46,"context_line":"Cinder as data and/or bootable volumes."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"To achive the goal above, Ironic needs to have following functions:"},{"line_number":49,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_da4e2d96","line":46,"in_reply_to":"da86d52c_afe3b431","updated":"2015-02-09 08:35:56.000000000","message":"Attaching volume itself is same, I think. But when we boot from volume, we need to setup BIOS to boot from external volume.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":76,"context_line":""},{"line_number":77,"context_line":"We have following three basic assumptions in this spec:"},{"line_number":78,"context_line":""},{"line_number":79,"context_line":"* The bootable block storage volue should have a bootable OS image on it"},{"line_number":80,"context_line":""},{"line_number":81,"context_line":"  * Cinder provides volume creation feature and so it is out of the scope"},{"line_number":82,"context_line":"    in this spec"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_0f5e6576","line":79,"updated":"2015-01-27 21:36:48.000000000","message":"nit. s/volue/volume","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":76,"context_line":""},{"line_number":77,"context_line":"We have following three basic assumptions in this spec:"},{"line_number":78,"context_line":""},{"line_number":79,"context_line":"* The bootable block storage volue should have a bootable OS image on it"},{"line_number":80,"context_line":""},{"line_number":81,"context_line":"  * Cinder provides volume creation feature and so it is out of the scope"},{"line_number":82,"context_line":"    in this spec"}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_7504817e","line":79,"in_reply_to":"1a930d6b_0f5e6576","updated":"2015-01-28 08:02:09.000000000","message":"Thanks, I\u0027ll fix it.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":82,"context_line":"    in this spec"},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_6a12d8b5","line":85,"updated":"2015-02-02 02:17:16.000000000","message":"@Chris, I\u0027m not sure that\u0027s relevant to cinder, and would like to pose the question to Satoru: in order to \"clean\" a machine between tenants, there is work being done to allow for secure-erase of local HDD. I would think this does not apply to cinder volumes because they can be deleted and created through an API. There is, AIUI, never a case where a cinder volume is created by one tenant, then erased and given to another tenant.\n\nThat said, for a machine with local HDD -- even if it is booting from Cinder volume -- we must still support cleaning the local HDD between tenants (once that work is implemented).","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":5805,"name":"Chris Krelle","email":"nobodycam@gmail.com","username":"nobodycam"},"change_message_id":"d33bed14ebdd32b6e29d1b8a7d3515eb2c6114b3","unresolved":false,"context_lines":[{"line_number":82,"context_line":"    in this spec"},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_90b43236","line":85,"updated":"2015-01-23 22:50:10.000000000","message":"the New State machine is introducing a cleaning state, I would like to see Cinder volumes support that.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"080dd54b060fd38834dd228919fbc0049bd94575","unresolved":false,"context_lines":[{"line_number":82,"context_line":"    in this spec"},{"line_number":83,"context_line":""},{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_da1c6d7a","line":85,"in_reply_to":"da86d52c_6a12d8b5","updated":"2015-02-09 08:35:56.000000000","message":"OK, I\u0027ll add support of cleaning the local HDD feature.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_8a170ca3","line":87,"updated":"2015-02-02 02:17:16.000000000","message":"I am eager to see your proposed implementation for supporting boot-from-volume with only standard IPMI","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":5805,"name":"Chris Krelle","email":"nobodycam@gmail.com","username":"nobodycam"},"change_message_id":"d33bed14ebdd32b6e29d1b8a7d3515eb2c6114b3","unresolved":false,"context_lines":[{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_10f762f6","line":87,"updated":"2015-01-23 22:50:10.000000000","message":"Many of the Ironic drivers are adding this support this cycle.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_24c22d34","line":87,"in_reply_to":"1a930d6b_10f762f6","updated":"2015-01-27 21:36:48.000000000","message":"yes.  iLO driver is working on bios/uefi settings.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_588df47a","line":87,"in_reply_to":"1a930d6b_24c22d34","updated":"2015-01-28 08:02:09.000000000","message":"Yes, vendor driver like iLO driver can interact with BIOS/UEFI. But I\u0027d like this feature to work with the driver which uses standard technology such as ipmitool. What I\u0027d like to explain here is this feature should not depend on Ironic BIOS/UEFI support and so put the assumption above.\n\nAnyway, I\u0027ll update this description.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"080dd54b060fd38834dd228919fbc0049bd94575","unresolved":false,"context_lines":[{"line_number":84,"context_line":"* Don\u0027t support Multi-path, Multi-volume, software initiator environment local"},{"line_number":85,"context_line":"  HDD clean up"},{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_ae085a92","line":87,"in_reply_to":"da86d52c_8a170ca3","updated":"2015-02-09 08:35:56.000000000","message":"Thanks. I\u0027ll implement it to work with standard IPMI.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":86,"context_line":""},{"line_number":87,"context_line":"* Ironic can\u0027t interact with the BIOS/UEFI"},{"line_number":88,"context_line":""},{"line_number":89,"context_line":"  * Currently, we have **NO** standard mechanism defined by some authority"},{"line_number":90,"context_line":"    or organization to change BIOS configurations remotely"},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"  * Although some vendor specific Ironic drivers can interact with their BIOS,"}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_aa0cd095","line":89,"updated":"2015-02-02 02:17:16.000000000","message":"FYI, there is a proposed standard of sorts, called REDFISH - https://www.redfishspecification.org/","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":91,"context_line":""},{"line_number":92,"context_line":"  * Although some vendor specific Ironic drivers can interact with their BIOS,"},{"line_number":93,"context_line":"    we should not rely on those vendor specific functions for generic"},{"line_number":94,"context_line":"    functionality such as Ironic-Cinder integration"},{"line_number":95,"context_line":""},{"line_number":96,"context_line":"Last assumption introduce two more subsequent assumptions:"},{"line_number":97,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_8f8ed5ce","line":94,"updated":"2015-01-27 21:36:48.000000000","message":"The  spec and interface should allow vendor drivers to implement this feature using their bios/uefi config capabilities.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":91,"context_line":""},{"line_number":92,"context_line":"  * Although some vendor specific Ironic drivers can interact with their BIOS,"},{"line_number":93,"context_line":"    we should not rely on those vendor specific functions for generic"},{"line_number":94,"context_line":"    functionality such as Ironic-Cinder integration"},{"line_number":95,"context_line":""},{"line_number":96,"context_line":"Last assumption introduce two more subsequent assumptions:"},{"line_number":97,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_95cc9537","line":94,"in_reply_to":"1a930d6b_8f8ed5ce","updated":"2015-01-28 08:02:09.000000000","message":"Ah, yes. It should be. I\u0027ll update.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":97,"context_line":""},{"line_number":98,"context_line":"* iSCSI/FC initiator information is provided during node registration process"},{"line_number":99,"context_line":""},{"line_number":100,"context_line":"  * We need to access the BIOS to get/set iSCSI/FC initiator information."},{"line_number":101,"context_line":"    On the otherhand, standard management tools can\u0027t access the BIOS."},{"line_number":102,"context_line":"    Therefore user should register those information manually during node"},{"line_number":103,"context_line":"    registration process."}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_ca1104ae","line":100,"updated":"2015-02-02 02:17:16.000000000","message":"This will be challenging, but not impossible. We will need a capability to verify the provided information prior to attempting provisioning.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"* To avoid BIOS configuration in Ironic, we need to set up \"Boot Order\" in"},{"line_number":106,"context_line":"  BIOS as follows prior to node registration."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"  #. Boot from volume attached to certain iSCSI/FC devices"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_ea25e80a","line":106,"updated":"2015-02-02 02:17:16.000000000","message":"Chris is correct -- we already use ipmitool to set the next boot device, and this has been part of the IPMI spec for several years (ie, it was added before 2.0, so is fairly widely supported already)","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":5805,"name":"Chris Krelle","email":"nobodycam@gmail.com","username":"nobodycam"},"change_message_id":"d33bed14ebdd32b6e29d1b8a7d3515eb2c6114b3","unresolved":false,"context_lines":[{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"* To avoid BIOS configuration in Ironic, we need to set up \"Boot Order\" in"},{"line_number":106,"context_line":"  BIOS as follows prior to node registration."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"  #. Boot from volume attached to certain iSCSI/FC devices"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_f0dbb64c","line":106,"updated":"2015-01-23 22:50:10.000000000","message":"Do we have to access the BIOS for this? Can this be done via ipmi bootdev command? Ironic currently uses this to set the boot device of a node between local hdd and network (pxe).","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"* To avoid BIOS configuration in Ironic, we need to set up \"Boot Order\" in"},{"line_number":106,"context_line":"  BIOS as follows prior to node registration."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"  #. Boot from volume attached to certain iSCSI/FC devices"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_d887240e","line":106,"in_reply_to":"1a930d6b_f0dbb64c","updated":"2015-01-28 08:02:09.000000000","message":"Thanks, I\u0027ll double-check on our server.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10910,"name":"Stig Telfer","email":"stig.openstack@telfer.org","username":"stigtelfer"},"change_message_id":"fb1c6c786a60fec0ca5c2a5f181c59758a2f1db1","unresolved":false,"context_lines":[{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"* To avoid BIOS configuration in Ironic, we need to set up \"Boot Order\" in"},{"line_number":106,"context_line":"  BIOS as follows prior to node registration."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"  #. Boot from volume attached to certain iSCSI/FC devices"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_4f55d84e","line":106,"in_reply_to":"da86d52c_ea25e80a","updated":"2015-02-03 15:05:36.000000000","message":"The use case I would like to see would be to boot via PXE, transferring kernel and ramdisk as usual, and then configure the kernel boot parameters to identify the Cinder volume.  Dracut ramdisks (for example) already support mounting root devices on iSCSI and (IIRC) FCoE.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":110,"context_line":"  #. Boot from PXE devices"},{"line_number":111,"context_line":""},{"line_number":112,"context_line":"  #. Boot from local HDD devices"},{"line_number":113,"context_line":""},{"line_number":114,"context_line":"Based on the assumptions above, work flow for boot from cinder volume in this"},{"line_number":115,"context_line":"spec is following:"},{"line_number":116,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_5c7f0915","line":113,"updated":"2015-01-27 21:36:48.000000000","message":"Is falling back to PXE desirable? If a deployer wants to boot from a cinder volume, shouldn\u0027t he/she just set the boot device to iSCSI/FC devices?\r\n\r\nAlso, any assumption where iscsi capabilities is supported, e.g., uefi network stack, nic firmware, ipxe/bios?  IMO, we should allow all.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":110,"context_line":"  #. Boot from PXE devices"},{"line_number":111,"context_line":""},{"line_number":112,"context_line":"  #. Boot from local HDD devices"},{"line_number":113,"context_line":""},{"line_number":114,"context_line":"Based on the assumptions above, work flow for boot from cinder volume in this"},{"line_number":115,"context_line":"spec is following:"},{"line_number":116,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_185b7c55","line":113,"in_reply_to":"1a930d6b_5c7f0915","updated":"2015-01-28 08:02:09.000000000","message":"Hmm, I think it depends on customer but yes your comment is reasonable and it\u0027s simpler than mine. So I\u0027ll remove fall back feature from this spec.\n\nAlso I\u0027ll add other assumptions.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"#. User registers bare metal node with initiator information (IQN/WWN)"},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"#. User asks nova to boot a node from a cinder volume"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":122,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_0a231c06","line":119,"updated":"2015-02-02 02:17:16.000000000","message":"One step missing here -- User must create a cinder volume from a bootable image","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"080dd54b060fd38834dd228919fbc0049bd94575","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"#. User registers bare metal node with initiator information (IQN/WWN)"},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"#. User asks nova to boot a node from a cinder volume"},{"line_number":120,"context_line":""},{"line_number":121,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":122,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_ce7bce79","line":119,"in_reply_to":"da86d52c_0a231c06","updated":"2015-02-09 08:35:56.000000000","message":"Thanks, I\u0027ll add it.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":2889,"name":"Aeva Black","email":"aeva.online@gmail.com","username":"tenbrae"},"change_message_id":"d84f5a54d69823ddf5e73d928a932cf53eeebf9a","unresolved":false,"context_lines":[{"line_number":121,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"   * Implement get_volume_connector method in Ironic driver. It collect"},{"line_number":124,"context_line":"     iSCSI/FC initiator information and pass it Nova."},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"   * Ironic get that information from deploy driver"},{"line_number":127,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_2a20e0f6","line":124,"updated":"2015-02-02 02:17:16.000000000","message":"@Wan-yen, I don\u0027t think we need any change to Nova API for this.\n\nSatoru\u0027s proposal is to add a new method to nova.virt.driver API which asks the driver for connector information, much the same way that we added a method there to ask the driver for node\u0027s MAC address(es) two years ago, which Nova then passes to Neutron. I think this approach is quite reasonable.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":121,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"   * Implement get_volume_connector method in Ironic driver. It collect"},{"line_number":124,"context_line":"     iSCSI/FC initiator information and pass it Nova."},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"   * Ironic get that information from deploy driver"},{"line_number":127,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_df14d045","line":124,"updated":"2015-01-27 21:36:48.000000000","message":"Also need a Nova api for ironic to return iscsi/fc initiator info to Nova.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10910,"name":"Stig Telfer","email":"stig.openstack@telfer.org","username":"stigtelfer"},"change_message_id":"fb1c6c786a60fec0ca5c2a5f181c59758a2f1db1","unresolved":false,"context_lines":[{"line_number":121,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":122,"context_line":""},{"line_number":123,"context_line":"   * Implement get_volume_connector method in Ironic driver. It collect"},{"line_number":124,"context_line":"     iSCSI/FC initiator information and pass it Nova."},{"line_number":125,"context_line":""},{"line_number":126,"context_line":"   * Ironic get that information from deploy driver"},{"line_number":127,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_6fac5c36","line":124,"in_reply_to":"da86d52c_2a20e0f6","updated":"2015-02-03 15:05:36.000000000","message":"@Deva - isn\u0027t the requirement met by the existing API method, get_volume_connector, which nova.virt.ironic could derive and implement?  If so, no API change for nova.virt.driver - even better.\n\nDoes any attention need paying to iSCSI vs FCoE vs other block protocols?  It would be nice not to restrict the physical protocols available.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10202,"name":"Wan-yen Hsu","email":"wanyenhsu@gmail.com"},"change_message_id":"ab1136a16413f7caf38a1141763224890f06f683","unresolved":false,"context_lines":[{"line_number":128,"context_line":"#. Nova calls Cinder to attach the volume to the host"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"#. Cinder calls the volume driver associated with the volume,"},{"line_number":131,"context_line":"   initialize_connection()"},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"#. Nova calls Ironic to boot the node"},{"line_number":134,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1a930d6b_190cb393","line":131,"updated":"2015-01-27 21:36:48.000000000","message":"Can os continues to use iscsi after booted?  If we want bare metal node OS to be able to use iscsi then I think Cinder needs to return traget connection_info to Nova and Nova needs to pass that info to Ironic.  Ironic then passes the info to bare-metal node so that bare metal node os can use iscsi.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"b665f2834cc4c4b5390919c7cde0ae50f416fbef","unresolved":false,"context_lines":[{"line_number":128,"context_line":"#. Nova calls Cinder to attach the volume to the host"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"#. Cinder calls the volume driver associated with the volume,"},{"line_number":131,"context_line":"   initialize_connection()"},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"#. Nova calls Ironic to boot the node"},{"line_number":134,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"fa81d914_73d01954","line":131,"in_reply_to":"1a930d6b_190cb393","updated":"2015-01-28 08:02:09.000000000","message":"In my understanding, initialize_connection (Cinder) connects the volume to the server and return target connection_info. After that, Ironic just turns on server and we can use it as long as boot order is configured appropriately.\n\nI\u0027ll double-check before posting v2.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":10910,"name":"Stig Telfer","email":"stig.openstack@telfer.org","username":"stigtelfer"},"change_message_id":"fb1c6c786a60fec0ca5c2a5f181c59758a2f1db1","unresolved":false,"context_lines":[{"line_number":134,"context_line":""},{"line_number":135,"context_line":"#. Ironic powers on the node"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"   * Implement deploy driver for boot from volume. It just turn on power."},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"#. (After using), user asks nova to terminate the node"},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_0a52ee12","line":137,"updated":"2015-02-03 15:05:36.000000000","message":"When booting from a Cinder volume, does deployment even need to power on the node?  Should it do anything other than update state in the Ironic node object?\n\nAfter deployment is (instantly) complete, the power driver would turn on the node.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"080dd54b060fd38834dd228919fbc0049bd94575","unresolved":false,"context_lines":[{"line_number":134,"context_line":""},{"line_number":135,"context_line":"#. Ironic powers on the node"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"   * Implement deploy driver for boot from volume. It just turn on power."},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"#. (After using), user asks nova to terminate the node"},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_4e751e31","line":137,"in_reply_to":"da86d52c_0a52ee12","updated":"2015-02-09 08:35:56.000000000","message":"It\u0027s my mistake.\n\nPower driver turns on power, not deploy driver.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":13762,"name":"jiangfei","email":"jiangfei0622@huawei.com","username":"jiangfei"},"change_message_id":"6a5bfb469e328fc11db7e7bb40c33079ced8184f","unresolved":false,"context_lines":[{"line_number":142,"context_line":""},{"line_number":143,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":144,"context_line":""},{"line_number":145,"context_line":"#. Nova calls Cinder to detach the volume to the host"}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_2da19aa9","line":145,"updated":"2015-02-10 07:14:59.000000000","message":"when node is deploy success?call cinder to attach the volume? do we miss this step?","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"a4cc679e892e85d004fd5e8542e8fc2d0d2ff4a0","unresolved":false,"context_lines":[{"line_number":142,"context_line":""},{"line_number":143,"context_line":"#. Nova calls Ironic to collect iSCSI/FC initiator information"},{"line_number":144,"context_line":""},{"line_number":145,"context_line":"#. Nova calls Cinder to detach the volume to the host"}],"source_content_type":"text/x-rst","patch_set":1,"id":"da86d52c_6d355245","line":145,"in_reply_to":"da86d52c_2da19aa9","updated":"2015-02-10 07:22:36.000000000","message":"I think \"deploy success\" is the time when Ironic succeeds in turning on the node after Cinder attaching the volume.","commit_id":"c1638b74f1f6ed2ce3de9e5d6e0a0db013e6a9a6"},{"author":{"_account_id":13997,"name":"SHIGEMATSU Mitsuhiro","email":"shigematsu.mitsuhiro@lab.ntt.co.jp","username":"pshige"},"change_message_id":"ffb3bd22a0bcdfad4610fd34eb5c2d0f71ecf32e","unresolved":false,"context_lines":[{"line_number":21,"context_line":"predictable performance. Also, in enterprise area, external block storage"},{"line_number":22,"context_line":"is commonly-used to get following benefits:"},{"line_number":23,"context_line":""},{"line_number":24,"context_line":"* Bettter data consistency while keeping good throughput/latency"},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"* Less overhead backup by utilizing volume snapshot and out-of-band volume"},{"line_number":27,"context_line":"  copy features"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a80dd14_c2123108","line":24,"updated":"2015-03-09 05:31:28.000000000","message":"nit. s/bettter/better","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":13997,"name":"SHIGEMATSU Mitsuhiro","email":"shigematsu.mitsuhiro@lab.ntt.co.jp","username":"pshige"},"change_message_id":"ffb3bd22a0bcdfad4610fd34eb5c2d0f71ecf32e","unresolved":false,"context_lines":[{"line_number":72,"context_line":""},{"line_number":73,"context_line":""},{"line_number":74,"context_line":"Proposed change"},{"line_number":75,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":76,"context_line":""},{"line_number":77,"context_line":"We have following three basic assumptions in this spec:"},{"line_number":78,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a80dd14_a2561dcc","line":75,"updated":"2015-03-09 05:31:28.000000000","message":"Please fix this error from jenkins, first. \n\nMissing sections: [u\u0027References\u0027, u\u0027Dependencies\u0027, u\u0027Documentation Impact\u0027, u\u0027Upgrades and Backwards Compatibility\u0027, u\u0027Testing\u0027, u\u0027Implementation\u0027]\n\nSection \u0027Proposed change\u0027 is missing subsections: [u\u0027Alternatives\u0027, u\u0027Data model impact\u0027, u\u0027REST API impact\u0027, u\u0027RPC API impact\u0027, u\u0027Driver API impact\u0027, u\u0027Nova driver impact\u0027, u\u0027Security impact\u0027, u\u0027Other end user impact\u0027, u\u0027Scalability impact\u0027, u\u0027Performance Impact\u0027, u\u0027Other deployer impact\u0027, u\u0027Developer impact\u0027]","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":10263,"name":"Vilobh Meshram","email":"vilobhmm@yahoo-inc.com","username":"vilobhmm"},"change_message_id":"45a96423b42ac0da38f4ef41e93433f063757c47","unresolved":false,"context_lines":[{"line_number":76,"context_line":""},{"line_number":77,"context_line":"We have following three basic assumptions in this spec:"},{"line_number":78,"context_line":""},{"line_number":79,"context_line":"* The bootable block storage volue should have a bootable OS image on it"},{"line_number":80,"context_line":""},{"line_number":81,"context_line":"  * Cinder provides volume creation feature and so it is out of the scope"},{"line_number":82,"context_line":"    in this spec"}],"source_content_type":"text/x-rst","patch_set":2,"id":"7a016987_9b570302","line":79,"updated":"2015-05-28 00:27:20.000000000","message":"*volume*","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f620644161d8801038be57c2eff4c6d766ce0c46","unresolved":false,"context_lines":[{"line_number":98,"context_line":"* iSCSI/FC initiator information is provided during node registration process"},{"line_number":99,"context_line":""},{"line_number":100,"context_line":"  * We need to access the BIOS to get/set iSCSI/FC initiator information."},{"line_number":101,"context_line":"    On the otherhand, standard management tools can\u0027t access the BIOS."},{"line_number":102,"context_line":"    Therefore user should register those information manually during node"},{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"7a016987_b3ea58d9","line":101,"updated":"2015-05-24 10:14:46.000000000","message":"nit: s/otherhand/other hand/","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f620644161d8801038be57c2eff4c6d766ce0c46","unresolved":false,"context_lines":[{"line_number":103,"context_line":"    registration process."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":"* To avoid BIOS configuration in Ironic, we need to set up \"Boot Order\" in"},{"line_number":106,"context_line":"  BIOS as follows prior to node registration."},{"line_number":107,"context_line":""},{"line_number":108,"context_line":"  #. Boot from volume attached to certain iSCSI/FC devices"},{"line_number":109,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"7a016987_93e51c0a","line":106,"updated":"2015-05-24 10:14:46.000000000","message":"How will it work with set_boot_device call that we do during deploy? Won\u0027t it override these settings?\nWe probably should somehow extend set_boot_device, or make sure it\u0027s not called in this case","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":10239,"name":"Dmitry Tantsur","email":"dtantsur@protonmail.com","username":"dtantsur"},"change_message_id":"f620644161d8801038be57c2eff4c6d766ce0c46","unresolved":false,"context_lines":[{"line_number":114,"context_line":"Based on the assumptions above, work flow for boot from cinder volume in this"},{"line_number":115,"context_line":"spec is following:"},{"line_number":116,"context_line":""},{"line_number":117,"context_line":"#. User registers bare metal node with initiator information (IQN/WWN)"},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"#. User asks nova to boot a node from a cinder volume"},{"line_number":120,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"7a016987_73f060c9","line":117,"updated":"2015-05-24 10:14:46.000000000","message":"How exactly is it going to be stored?","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"},{"author":{"_account_id":9066,"name":"Satoru Moriya","email":"satoru.moriya.br@hitachi.com","username":"satoru-moriya-br"},"change_message_id":"ac0581ee6121d6880b3cb198d183a63bb8933a87","unresolved":false,"context_lines":[{"line_number":114,"context_line":"Based on the assumptions above, work flow for boot from cinder volume in this"},{"line_number":115,"context_line":"spec is following:"},{"line_number":116,"context_line":""},{"line_number":117,"context_line":"#. User registers bare metal node with initiator information (IQN/WWN)"},{"line_number":118,"context_line":""},{"line_number":119,"context_line":"#. User asks nova to boot a node from a cinder volume"},{"line_number":120,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a41bdd9_7d656e43","line":117,"in_reply_to":"7a016987_73f060c9","updated":"2015-07-13 04:23:30.000000000","message":"I\u0027m thinking about adding two new attributes \u0027volume_initiator_info\u0027 and \u0027volume_target_info\u0027 to node table. User need to register node connection info such as iqn, ip, wwnn, wwpn to \u0027volume_initiator_info\u0027 attribute.","commit_id":"15990a28ffaf5b1bef569788dca6779dea2d54ff"}]}