)]}'
{"api-guide/source/server_concepts.rst":[{"author":{"_account_id":5754,"name":"Alex Xu","email":"hejie.xu@intel.com","username":"xuhj"},"change_message_id":"5eff4daaa60cbceb7e9c0a484fbe7a3b8fbf9e5f","unresolved":false,"context_lines":[{"line_number":394,"context_line":"~~~~~~~~~~~~~~"},{"line_number":395,"context_line":""},{"line_number":396,"context_line":"There are several actions that may result in a server moving from one"},{"line_number":397,"context_line":"compute host to another including shelve, resize and migrations. The"},{"line_number":398,"context_line":"following use cases demonstrate the intention of the actions and"},{"line_number":399,"context_line":"the consequence for operational procedures."},{"line_number":400,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_f809f084","line":397,"updated":"2015-11-19 09:02:59.000000000","message":"Should we count evacuate at here?","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"90f1d596a43af8e8a2aa8bf7e79da490f8069531","unresolved":false,"context_lines":[{"line_number":394,"context_line":"~~~~~~~~~~~~~~"},{"line_number":395,"context_line":""},{"line_number":396,"context_line":"There are several actions that may result in a server moving from one"},{"line_number":397,"context_line":"compute host to another including shelve, resize and migrations. The"},{"line_number":398,"context_line":"following use cases demonstrate the intention of the actions and"},{"line_number":399,"context_line":"the consequence for operational procedures."},{"line_number":400,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_ad4aa181","line":397,"in_reply_to":"ba8a016a_574aec56","updated":"2015-11-23 18:17:04.000000000","message":"Done","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"a927531855d455b97d50603a76f3562a902ed99c","unresolved":false,"context_lines":[{"line_number":394,"context_line":"~~~~~~~~~~~~~~"},{"line_number":395,"context_line":""},{"line_number":396,"context_line":"There are several actions that may result in a server moving from one"},{"line_number":397,"context_line":"compute host to another including shelve, resize and migrations. The"},{"line_number":398,"context_line":"following use cases demonstrate the intention of the actions and"},{"line_number":399,"context_line":"the consequence for operational procedures."},{"line_number":400,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_574aec56","line":397,"in_reply_to":"ba8a016a_f809f084","updated":"2015-11-20 16:57:46.000000000","message":"I did think about evacuate - I left it out because its not a deliberate move, but I can add it for completeness as a kind of move","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":10618,"name":"Pavel Kholkin","email":"p.v.holkin@mail.ru","username":"pkholkin"},"change_message_id":"61dc3036f9b2a9db98032f4574a6cf529c60ff9e","unresolved":false,"context_lines":[{"line_number":405,"context_line":"the cloud operator an opportunity to make better use of resources by"},{"line_number":406,"context_line":"freeing resources and rebalancing workloads across the infrastructure."},{"line_number":407,"context_line":""},{"line_number":408,"context_line":"When the user shelves a server the operator can chose to remove it"},{"line_number":409,"context_line":"from the compute hosts. When it is unshelved it is scheduled to a new"},{"line_number":410,"context_line":"host according to the operators policies for distributing work loads"},{"line_number":411,"context_line":"across the compute hosts, including taking disabled hosts into account."}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_52c7d795","line":408,"updated":"2015-11-18 17:45:27.000000000","message":"/s/chose/choose","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"90f1d596a43af8e8a2aa8bf7e79da490f8069531","unresolved":false,"context_lines":[{"line_number":405,"context_line":"the cloud operator an opportunity to make better use of resources by"},{"line_number":406,"context_line":"freeing resources and rebalancing workloads across the infrastructure."},{"line_number":407,"context_line":""},{"line_number":408,"context_line":"When the user shelves a server the operator can chose to remove it"},{"line_number":409,"context_line":"from the compute hosts. When it is unshelved it is scheduled to a new"},{"line_number":410,"context_line":"host according to the operators policies for distributing work loads"},{"line_number":411,"context_line":"across the compute hosts, including taking disabled hosts into account."}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_ed818990","line":408,"in_reply_to":"ba8a016a_52c7d795","updated":"2015-11-23 18:17:04.000000000","message":"Done","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":10618,"name":"Pavel Kholkin","email":"p.v.holkin@mail.ru","username":"pkholkin"},"change_message_id":"61dc3036f9b2a9db98032f4574a6cf529c60ff9e","unresolved":false,"context_lines":[{"line_number":420,"context_line":"**Resize**"},{"line_number":421,"context_line":""},{"line_number":422,"context_line":"Sometimes a user may want to change the flavor of a server, e.g. change"},{"line_number":423,"context_line":"the quantity of cpu, disk, memory or any other resource. This is done"},{"line_number":424,"context_line":"by rebuilding the server with a new flavor. As the server is being"},{"line_number":425,"context_line":"rebuilt it is normal to reschedule the server to another host"},{"line_number":426,"context_line":"(although resize to the same host is an option for the operator)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_0d1cd468","line":423,"updated":"2015-11-18 17:45:27.000000000","message":"should \u0027cpu\u0027 be \u0027cpus\u0027 or maybe \u0027CPUs\u0027?","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"90f1d596a43af8e8a2aa8bf7e79da490f8069531","unresolved":false,"context_lines":[{"line_number":420,"context_line":"**Resize**"},{"line_number":421,"context_line":""},{"line_number":422,"context_line":"Sometimes a user may want to change the flavor of a server, e.g. change"},{"line_number":423,"context_line":"the quantity of cpu, disk, memory or any other resource. This is done"},{"line_number":424,"context_line":"by rebuilding the server with a new flavor. As the server is being"},{"line_number":425,"context_line":"rebuilt it is normal to reschedule the server to another host"},{"line_number":426,"context_line":"(although resize to the same host is an option for the operator)."}],"source_content_type":"text/x-rst","patch_set":1,"id":"ba8a016a_5e695924","line":423,"in_reply_to":"ba8a016a_0d1cd468","updated":"2015-11-23 18:17:04.000000000","message":"Done","commit_id":"ef2df1b1d2df356edf37b5bf06c9d54d639a5e37"},{"author":{"_account_id":6167,"name":"Ken\u0027ichi Ohmichi","email":"ken1ohmichi@gmail.com","username":"oomichi"},"change_message_id":"b8874f2df65b8505d0d9a490c899368426e8caed","unresolved":false,"context_lines":[{"line_number":401,"context_line":""},{"line_number":402,"context_line":"**Shelving**"},{"line_number":403,"context_line":""},{"line_number":404,"context_line":"Sometimes a user does not require a server to be active for some time,"},{"line_number":405,"context_line":"perhaps over a weekend or at certain times of day. This gives"},{"line_number":406,"context_line":"the cloud operator an opportunity to make better use of resources by"},{"line_number":407,"context_line":"freeing resources and rebalancing workloads across the infrastructure."}],"source_content_type":"text/x-rst","patch_set":2,"id":"ba8a016a_d1dd2571","line":404,"updated":"2015-11-24 05:55:52.000000000","message":"\"for some time\" seems duplicated to the first \"Sometimes\" for me, is that different meaning?","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"bb35dd9e2fd4f7158823e4f80a6f7fc08aa2081a","unresolved":false,"context_lines":[{"line_number":401,"context_line":""},{"line_number":402,"context_line":"**Shelving**"},{"line_number":403,"context_line":""},{"line_number":404,"context_line":"Sometimes a user does not require a server to be active for some time,"},{"line_number":405,"context_line":"perhaps over a weekend or at certain times of day. This gives"},{"line_number":406,"context_line":"the cloud operator an opportunity to make better use of resources by"},{"line_number":407,"context_line":"freeing resources and rebalancing workloads across the infrastructure."}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a8ffd7b_1876e93e","line":404,"in_reply_to":"ba8a016a_cc54ae55","updated":"2015-11-24 11:57:20.000000000","message":"Done","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"f6afcec72d9245b27b54e8184b08d393993ba0c2","unresolved":false,"context_lines":[{"line_number":401,"context_line":""},{"line_number":402,"context_line":"**Shelving**"},{"line_number":403,"context_line":""},{"line_number":404,"context_line":"Sometimes a user does not require a server to be active for some time,"},{"line_number":405,"context_line":"perhaps over a weekend or at certain times of day. This gives"},{"line_number":406,"context_line":"the cloud operator an opportunity to make better use of resources by"},{"line_number":407,"context_line":"freeing resources and rebalancing workloads across the infrastructure."}],"source_content_type":"text/x-rst","patch_set":2,"id":"ba8a016a_cc54ae55","line":404,"in_reply_to":"ba8a016a_d1dd2571","updated":"2015-11-24 06:34:37.000000000","message":"It is different meaning: sometimes \u003d \"Occasionally it is the case that...\"; for some time \u003d \"for a period of time.\"\n\nBut it does sound repetitious so I will change it.","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":6167,"name":"Ken\u0027ichi Ohmichi","email":"ken1ohmichi@gmail.com","username":"oomichi"},"change_message_id":"b8874f2df65b8505d0d9a490c899368426e8caed","unresolved":false,"context_lines":[{"line_number":447,"context_line":"server that is not active) and live migration (moving a server that"},{"line_number":448,"context_line":"is active)."},{"line_number":449,"context_line":""},{"line_number":450,"context_line":"Migration moves a server from one host to another by copying its"},{"line_number":451,"context_line":"state, local storage and network configuration to new resources"},{"line_number":452,"context_line":"allocated on a new host selected by scheduling policies or as"},{"line_number":453,"context_line":"an explicit decision. The operation is relatively quick as the"}],"source_content_type":"text/x-rst","patch_set":2,"id":"ba8a016a_71b9d9ca","line":450,"updated":"2015-11-24 05:55:52.000000000","message":"I was a little confused when seeing this \"Migration\" because I could not get it as cold migration.\nSo how about being \"Cold migration\" here?\n\nThe title (**Migration (including ...)) also is Migration, so nice to clarify here.","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"bb35dd9e2fd4f7158823e4f80a6f7fc08aa2081a","unresolved":false,"context_lines":[{"line_number":447,"context_line":"server that is not active) and live migration (moving a server that"},{"line_number":448,"context_line":"is active)."},{"line_number":449,"context_line":""},{"line_number":450,"context_line":"Migration moves a server from one host to another by copying its"},{"line_number":451,"context_line":"state, local storage and network configuration to new resources"},{"line_number":452,"context_line":"allocated on a new host selected by scheduling policies or as"},{"line_number":453,"context_line":"an explicit decision. The operation is relatively quick as the"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a8ffd7b_987bb97b","line":450,"in_reply_to":"ba8a016a_71b9d9ca","updated":"2015-11-24 11:57:20.000000000","message":"Done","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"f6afcec72d9245b27b54e8184b08d393993ba0c2","unresolved":false,"context_lines":[{"line_number":477,"context_line":""},{"line_number":478,"context_line":"-  **Rolling updates**"},{"line_number":479,"context_line":""},{"line_number":480,"context_line":"   Often it is necessary to perform an update on all compute hosts"},{"line_number":481,"context_line":"   that requires them to be rebooted. In this case it is not"},{"line_number":482,"context_line":"   strictly necessary to move inactive instances because they"},{"line_number":483,"context_line":"   will be available after the reboot. However, active instances would"}],"source_content_type":"text/x-rst","patch_set":2,"id":"ba8a016a_6cb902cb","line":480,"updated":"2015-11-24 06:34:37.000000000","message":"...necessary to perform an update that requires a reboot on all compute hosts.","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"bb35dd9e2fd4f7158823e4f80a6f7fc08aa2081a","unresolved":false,"context_lines":[{"line_number":477,"context_line":""},{"line_number":478,"context_line":"-  **Rolling updates**"},{"line_number":479,"context_line":""},{"line_number":480,"context_line":"   Often it is necessary to perform an update on all compute hosts"},{"line_number":481,"context_line":"   that requires them to be rebooted. In this case it is not"},{"line_number":482,"context_line":"   strictly necessary to move inactive instances because they"},{"line_number":483,"context_line":"   will be available after the reboot. However, active instances would"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a8ffd7b_38b64d3e","line":480,"in_reply_to":"ba8a016a_6cb902cb","updated":"2015-11-24 11:57:20.000000000","message":"Won\u0027t change - looks ok as it is.","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":6167,"name":"Ken\u0027ichi Ohmichi","email":"ken1ohmichi@gmail.com","username":"oomichi"},"change_message_id":"b8874f2df65b8505d0d9a490c899368426e8caed","unresolved":false,"context_lines":[{"line_number":491,"context_line":"   new host continues normal operation. The old host will be empty"},{"line_number":492,"context_line":"   and can be updated and rebooted. Then it becomes the new target for"},{"line_number":493,"context_line":"   another round of migrations."},{"line_number":494,"context_line":"   "},{"line_number":495,"context_line":"   This process can be repeated until the whole cloud has been updated,"},{"line_number":496,"context_line":"   usually using a pool of empty hosts instead of just one."},{"line_number":497,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"ba8a016a_71e0f9cc","line":494,"updated":"2015-11-24 05:55:52.000000000","message":"unnecessary spaces at line 494.","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":7461,"name":"Paul Murray","email":"pmurray@hpe.com","username":"ptm"},"change_message_id":"bb35dd9e2fd4f7158823e4f80a6f7fc08aa2081a","unresolved":false,"context_lines":[{"line_number":491,"context_line":"   new host continues normal operation. The old host will be empty"},{"line_number":492,"context_line":"   and can be updated and rebooted. Then it becomes the new target for"},{"line_number":493,"context_line":"   another round of migrations."},{"line_number":494,"context_line":"   "},{"line_number":495,"context_line":"   This process can be repeated until the whole cloud has been updated,"},{"line_number":496,"context_line":"   usually using a pool of empty hosts instead of just one."},{"line_number":497,"context_line":""}],"source_content_type":"text/x-rst","patch_set":2,"id":"9a8ffd7b_5841d138","line":494,"in_reply_to":"ba8a016a_71e0f9cc","updated":"2015-11-24 11:57:20.000000000","message":"Done","commit_id":"3aa1575245d97f22aa482597c2b390c8332659b1"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":131,"context_line":"   Evacuate is a work around for this that lets an administrator"},{"line_number":132,"context_line":"   forceably rebuild these zombie instances on another node. It makes"},{"line_number":133,"context_line":"   no guarantees that the instance was actually down, so fencing is"},{"line_number":134,"context_line":"   left as an exercise to the deployer."},{"line_number":135,"context_line":""},{"line_number":136,"context_line":"-  **Resize** (including **Confirm resize**, **Revert resize**)"},{"line_number":137,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_18a024f4","line":134,"updated":"2015-11-24 16:22:31.000000000","message":"Hmm, this seems to reference the details below?","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":250,"context_line":"   Set the root/admin password for the given instance, it wil uses an"},{"line_number":251,"context_line":"   optional installed agent to inject the admin password."},{"line_number":252,"context_line":""},{"line_number":253,"context_line":"-  **Migrate**, **Live migrate**"},{"line_number":254,"context_line":""},{"line_number":255,"context_line":"   Migrate is usually utilized by admin, it will move an instance to another"},{"line_number":256,"context_line":"   host; it utilize the \u0027resize\u0027 action but with same flavor, so during"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_d8a93ce1","line":253,"updated":"2015-11-24 16:22:31.000000000","message":"similar here? should we link?","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":398,"context_line":"compute host to another including shelve, resize, migrations and"},{"line_number":399,"context_line":"evacuate. The following use cases demonstrate the intention of the"},{"line_number":400,"context_line":"actions and the consequence for operational procedures."},{"line_number":401,"context_line":""},{"line_number":402,"context_line":"**Shelving**"},{"line_number":403,"context_line":""},{"line_number":404,"context_line":"Sometimes a user does not require a server to be active for a while,"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_6790a896","line":401,"updated":"2015-11-24 16:22:31.000000000","message":"This actually reads a little strangely, trying to think about how to describe this.\n\nMaybe we can say:\n* sometimes the user wants something, and to do that we need to move the instance, e.g. resize, shelve\n* sometimes the operator needs to do something that requires them to move the instance, e.g. live-migrate, migrate","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":399,"context_line":"evacuate. The following use cases demonstrate the intention of the"},{"line_number":400,"context_line":"actions and the consequence for operational procedures."},{"line_number":401,"context_line":""},{"line_number":402,"context_line":"**Shelving**"},{"line_number":403,"context_line":""},{"line_number":404,"context_line":"Sometimes a user does not require a server to be active for a while,"},{"line_number":405,"context_line":"perhaps over a weekend or at certain times of day. This gives"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_d8805c49","line":402,"updated":"2015-11-24 16:22:31.000000000","message":"Nit: Feels like these should be headings and not just bold","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"778e1a1fdcb8e9eba6f758645a61b41e03725d94","unresolved":false,"context_lines":[{"line_number":407,"context_line":"freeing resources and rebalancing workloads across the infrastructure."},{"line_number":408,"context_line":""},{"line_number":409,"context_line":"When the user shelves a server the operator can choose to remove it"},{"line_number":410,"context_line":"from the compute hosts. When it is unshelved it is scheduled to a new"},{"line_number":411,"context_line":"host according to the operators policies for distributing work loads"},{"line_number":412,"context_line":"across the compute hosts, including taking disabled hosts into account."},{"line_number":413,"context_line":"This will contribute to increased overall capacity, freeing hosts that"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_5d7456fc","line":410,"range":{"start_line":410,"start_character":23,"end_line":410,"end_character":44},"updated":"2015-11-24 15:16:22.000000000","message":"that sounds picky but only if it\u0027s shelved offloaded, right?","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"778e1a1fdcb8e9eba6f758645a61b41e03725d94","unresolved":false,"context_lines":[{"line_number":414,"context_line":"are ear-marked for maintenance and providing contiguous blocks"},{"line_number":415,"context_line":"of resources on single hosts due to moving out old servers."},{"line_number":416,"context_line":""},{"line_number":417,"context_line":"Shelving a server is not normally a choice that is available to"},{"line_number":418,"context_line":"the cloud operator because it affects the availability of the server"},{"line_number":419,"context_line":"being provided to the user."},{"line_number":420,"context_line":""},{"line_number":421,"context_line":"**Resize**"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_e11fa859","line":418,"range":{"start_line":417,"start_character":0,"end_line":418,"end_character":18},"updated":"2015-11-24 15:16:22.000000000","message":"Meh, the default policy accepts admin/user for a shelve but admin only for shelve-offload, so your comment is a bit confusing.\nhttps://github.com/openstack/nova/blob/e1f05750417eec26e973bc1af6f8e7dbacb07679/etc/nova/policy.json#L182-L183\n\nHere I guess you want to explain that reasonable operators should avoid VM owners to rebalance their VMs, right? If so, \"cloud operator\" doesn\u0027t sound the right name to use.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":414,"context_line":"are ear-marked for maintenance and providing contiguous blocks"},{"line_number":415,"context_line":"of resources on single hosts due to moving out old servers."},{"line_number":416,"context_line":""},{"line_number":417,"context_line":"Shelving a server is not normally a choice that is available to"},{"line_number":418,"context_line":"the cloud operator because it affects the availability of the server"},{"line_number":419,"context_line":"being provided to the user."},{"line_number":420,"context_line":""},{"line_number":421,"context_line":"**Resize**"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_e7c9f8ad","line":418,"range":{"start_line":417,"start_character":0,"end_line":418,"end_character":18},"in_reply_to":"9a8ffd7b_e11fa859","updated":"2015-11-24 16:22:31.000000000","message":"So I think this is OK, its just a bit irrelevant in some ways.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"778e1a1fdcb8e9eba6f758645a61b41e03725d94","unresolved":false,"context_lines":[{"line_number":431,"context_line":"to the operators scheduling policy, providing the same benefits as"},{"line_number":432,"context_line":"above."},{"line_number":433,"context_line":""},{"line_number":434,"context_line":"Resizing a server is not normally a choice that is available to"},{"line_number":435,"context_line":"the cloud operator because it changes the nature of the server"},{"line_number":436,"context_line":"being provided to the user."},{"line_number":437,"context_line":""},{"line_number":438,"context_line":"**Migration (including cold and live migration)**"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_c19384a2","line":435,"range":{"start_line":434,"start_character":0,"end_line":435,"end_character":18},"updated":"2015-11-24 15:16:22.000000000","message":"same comment here, by default we allow any VM owner to resize its instance. If you consider that a wrong RBAC policy, you should propose a change against that policy but you can\u0027t state the above because the default is not that.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":431,"context_line":"to the operators scheduling policy, providing the same benefits as"},{"line_number":432,"context_line":"above."},{"line_number":433,"context_line":""},{"line_number":434,"context_line":"Resizing a server is not normally a choice that is available to"},{"line_number":435,"context_line":"the cloud operator because it changes the nature of the server"},{"line_number":436,"context_line":"being provided to the user."},{"line_number":437,"context_line":""},{"line_number":438,"context_line":"**Migration (including cold and live migration)**"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_847eeaac","line":435,"range":{"start_line":434,"start_character":0,"end_line":435,"end_character":18},"in_reply_to":"9a8ffd7b_c19384a2","updated":"2015-11-24 16:22:31.000000000","message":"Not sure. I agree with Paul\u0027s comment here, although its a little odd in this context.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":435,"context_line":"the cloud operator because it changes the nature of the server"},{"line_number":436,"context_line":"being provided to the user."},{"line_number":437,"context_line":""},{"line_number":438,"context_line":"**Migration (including cold and live migration)**"},{"line_number":439,"context_line":""},{"line_number":440,"context_line":"Sometimes a cloud operator may need to redistribute work loads for"},{"line_number":441,"context_line":"operational purposes. For example, the operator may need to remove"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_27baa0e2","line":438,"updated":"2015-11-24 16:22:31.000000000","message":"Maybe make this section first?","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"adb4a4cb07bd0b002753e8044a0df33e5f8e0c01","unresolved":false,"context_lines":[{"line_number":460,"context_line":"a limited set of management actions are available to the user."},{"line_number":461,"context_line":""},{"line_number":462,"context_line":"The following are two common patterns for employing migrations in"},{"line_number":463,"context_line":"a cloud:"},{"line_number":464,"context_line":""},{"line_number":465,"context_line":"-  **Host maintenance**"},{"line_number":466,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_e7dc9848","line":463,"updated":"2015-11-24 16:22:31.000000000","message":"I think this should include resource balancing.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":9555,"name":"Matthew Booth","email":"mbooth@redhat.com","username":"MatthewBooth"},"change_message_id":"1f8fc67eb47cdaabfe79c6f89edf9efed2b73522","unresolved":false,"context_lines":[{"line_number":504,"context_line":"Sometimes a compute host may fail. This is a rare occurrence, but when"},{"line_number":505,"context_line":"it happens during normal operation the servers running on the host may"},{"line_number":506,"context_line":"be lost. In this case the operator may recreate the servers on the"},{"line_number":507,"context_line":"remaining compute hosts using the evacuate action."},{"line_number":508,"context_line":""},{"line_number":509,"context_line":"Failure detection can be proved to be impossible in compute systems"},{"line_number":510,"context_line":"with asynchronous communication, so true failure detection cannot be"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_c4add232","line":507,"updated":"2015-11-24 15:38:50.000000000","message":"I\u0027d highlight here that this involves a loss of data. I know you\u0027ve already said \u0027lost\u0027, but I\u0027d drill it home explicitly. \u0027Evacuate\u0027 has connotations of recovering data from a failed host which is frequently confusing to users.","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"},{"author":{"_account_id":7166,"name":"Sylvain Bauza","email":"sbauza@redhat.com","username":"sbauza"},"change_message_id":"778e1a1fdcb8e9eba6f758645a61b41e03725d94","unresolved":false,"context_lines":[{"line_number":511,"context_line":"achieved. Usually when a host is considered to have failed it should be"},{"line_number":512,"context_line":"excluded from the cloud and any virtual networking or storage associated"},{"line_number":513,"context_line":"with servers on the failed host should be isolated from it. These steps"},{"line_number":514,"context_line":"are called fencing the host. Initiating these action is outside the scope"},{"line_number":515,"context_line":"of Nova."},{"line_number":516,"context_line":""},{"line_number":517,"context_line":"Once the host has been fenced its servers can be recreated on other"},{"line_number":518,"context_line":"hosts without worry of the old incarnations reappearing and trying to"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a8ffd7b_01466c23","line":515,"range":{"start_line":514,"start_character":29,"end_line":515,"end_character":8},"updated":"2015-11-24 15:16:22.000000000","message":"but now you can ask Nova to mark the host down, so you should add that :-)","commit_id":"192e6b1d03ce0d21f79a5afba209aa7e6a301d02"}]}