)]}'
{"doc/source/code-review.rst":[{"author":{"_account_id":1779,"name":"Daniel Berrange","email":"berrange@redhat.com","username":"berrange"},"change_message_id":"6274b1833993add8b925f20a8e34800fa3b72382","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_4e041449","line":122,"updated":"2015-12-02 18:16:56.000000000","message":"I think we need to clarify this more before we turn it into a formal rule.\n\nie, \"things that can impact XenAPI\" covers essentially all Nova code except for nova/virt/\u003cother virt drivers\u003e.  So if i change nova/compute/manager.py, this policy is saying we must wait for +1 votes from *all* virt driver CI systems.  Given the unreliability of many 3rd party CIs this is quite a big step forward. eg do we want to block compute manager changes because PPC64  3rd party CI for libvirt KVM has been broken for weeks and giving -1\u0027s to everything ? Or if we have a Docker 3rd party CI system up that gives a -1, should we honour that even though Docker is not an in-tree virt driver. Traditionally we reserve the right to break any out of tree virt driver even if it has 3rd party CI setup.","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":1779,"name":"Daniel Berrange","email":"berrange@redhat.com","username":"berrange"},"change_message_id":"08531d5c7ea74fa9a9177651dd6709c201b349b2","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_a90eda89","line":122,"in_reply_to":"9a8ffd7b_0974c633","updated":"2015-12-02 18:31:14.000000000","message":"Yeah, I think explicitly saying that changes to nova/virt/driver/\u003cNNNN\u003e need vote from the respective driver CI is fine.\n\nAt a granularity beyond virt drivers, with the proliferation of 3rd party CI code it is hard to know/remember which 3rd party CI we should look at. eg if we end up with one 3rd party CI for each VIF type or each cinder volume driver, good luck remember which we should wait for a vote from for each change...  So I\u0027d think we want some better Gerrit vote planning + reporting UI before we make things more complicated than just vitt driver directory changes.  ie for each change, gerrit should tell us which 3rd party CI we should be expecting to vote","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"dee4ded53d0f213c43a7c0755dfb6c0f36d0c17c","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_69123259","line":122,"in_reply_to":"9a8ffd7b_4e041449","updated":"2015-12-02 18:20:37.000000000","message":"First, this would only apply to 3rd party CI to in-tree code.  Maybe that should be explicit in here, but we don\u0027t have any cases of out of tree code 3rd party CI do we?\n\nSecond, I think this is tailored more specifically to the obvious things, like vmware driver changes should pass NSX CI, same for hyper-v, xen, Ironic (which isn\u0027t 3rd party), etc.\n\nWe know there are other more global changes that could break those drivers, but I think at a minimum we need to enforce virt driver 3rd party CI approving a change that is restricted to that virt driver.  That should probably be clarified in here.","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":5441,"name":"Andrew Laski","email":"andrew@lascii.com","username":"alaski"},"change_message_id":"819b1776d1858e9c3019b9a64bcb31fc9ce76d7a","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_0974c633","line":122,"in_reply_to":"9a8ffd7b_69123259","updated":"2015-12-02 18:28:06.000000000","message":"Clarification that code changes within a virt driver must have an appropriate CI passing before approval would be nice.  I think that\u0027s a good scope to start with and then expand this out.\n\nBut I think the current wording is fairly clear with \"When a change is only covered by a third party test...\"  which means a change in compute/manager.py would be covered by infra tests.  Unless it involved something specifically uncovered like PCI changes in which case third party CI is needed.","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"a6c6da27d21e876960d7c29d68c084d2893f5075","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_29674ab8","line":122,"in_reply_to":"9a8ffd7b_69123259","updated":"2015-12-02 18:26:26.000000000","message":"I am thinking baby steps here, I will try clarify that a bit.","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":4393,"name":"Dan Smith","email":"dms@danplanet.com","username":"danms"},"change_message_id":"98c17f9a71d9f9e738d634c33f8c6b2c6ab68cb0","unresolved":false,"context_lines":[{"line_number":119,"context_line":"vote from that third party test, before a +2 vote can be applied."},{"line_number":120,"context_line":"For example, if you change something in the XenAPI virt driver, you must wait"},{"line_number":121,"context_line":"for a +1 from the XenServer CI on the latest patchset, before you can give"},{"line_number":122,"context_line":"that patch set a +2 vote."},{"line_number":123,"context_line":""},{"line_number":124,"context_line":"This is important to ensure:"},{"line_number":125,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"9a8ffd7b_64c3398a","line":122,"in_reply_to":"9a8ffd7b_a90eda89","updated":"2015-12-02 18:43:59.000000000","message":"More specific language potentially excludes more cases as well.\n\nI think that arguing that this is too general is kinda silly -- I think most people know a change that needs validation when they see it. Clearly anything that touches the driver itself would need it, but anything that changes the driver interface itself, or something from the API that may result in different values making it down to the virt driver.\n\nI think fewer words means it\u0027s more likely that people will actually read this part of the guideline.","commit_id":"467d31cb6a21290a28e1265d57ca88259d27d98a"},{"author":{"_account_id":5441,"name":"Andrew Laski","email":"andrew@lascii.com","username":"alaski"},"change_message_id":"5e5f52acafa0500d36a15f93b401e35a18cb2dd7","unresolved":false,"context_lines":[{"line_number":141,"context_line":""},{"line_number":142,"context_line":"Please note:"},{"line_number":143,"context_line":""},{"line_number":144,"context_line":"* Long term, we should ensure that any patch a third part CI is allowed to"},{"line_number":145,"context_line":"  vote on is binding. But we need a lot more work to make something like"},{"line_number":146,"context_line":"  that feasible, hence the proposed compromise."},{"line_number":147,"context_line":"* While its possible to break a virt driver CI system by changing code that is"}],"source_content_type":"text/x-rst","patch_set":4,"id":"7a740942_1176ce3e","line":144,"range":{"start_line":144,"start_character":53,"end_line":144,"end_character":57},"updated":"2015-12-03 15:16:17.000000000","message":"party","commit_id":"68e7be3ce77b07656acc1ab192cfd27b1751a9cf"},{"author":{"_account_id":5441,"name":"Andrew Laski","email":"andrew@lascii.com","username":"alaski"},"change_message_id":"5e5f52acafa0500d36a15f93b401e35a18cb2dd7","unresolved":false,"context_lines":[{"line_number":147,"context_line":"* While its possible to break a virt driver CI system by changing code that is"},{"line_number":148,"context_line":"  outside the virt drivers, this policy is not focusing on fixing that."},{"line_number":149,"context_line":"  A third party test failure should always be investigated, but the failure of"},{"line_number":150,"context_line":"  a third party test to report in a timely manor should not block others."},{"line_number":151,"context_line":"* We are only talking about the testing of in-tree code. Please note the only"},{"line_number":152,"context_line":"  public API is our REST API, see: doc:`policies`"},{"line_number":153,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"7a740942_b186da11","line":150,"range":{"start_line":150,"start_character":43,"end_line":150,"end_character":48},"updated":"2015-12-03 15:16:17.000000000","message":"manner","commit_id":"68e7be3ce77b07656acc1ab192cfd27b1751a9cf"},{"author":{"_account_id":6873,"name":"Matt Riedemann","email":"mriedem.os@gmail.com","username":"mriedem"},"change_message_id":"53ded710e59a98c7e196a8e628828344bc6e597c","unresolved":false,"context_lines":[{"line_number":149,"context_line":"  A third party test failure should always be investigated, but the failure of"},{"line_number":150,"context_line":"  a third party test to report in a timely manor should not block others."},{"line_number":151,"context_line":"* We are only talking about the testing of in-tree code. Please note the only"},{"line_number":152,"context_line":"  public API is our REST API, see: doc:`policies`"},{"line_number":153,"context_line":""},{"line_number":154,"context_line":"Release Notes"},{"line_number":155,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"7a740942_ae566c04","line":152,"updated":"2015-12-03 17:19:10.000000000","message":"The link to the policies page doesn\u0027t actually render properly:\n\nhttp://docs-draft.openstack.org/43/252543/4/check/gate-nova-docs/75b124a//doc/build/html/code-review.html#notes","commit_id":"68e7be3ce77b07656acc1ab192cfd27b1751a9cf"},{"author":{"_account_id":782,"name":"John Garbutt","email":"john@johngarbutt.com","username":"johngarbutt"},"change_message_id":"1a2951324489d34565a9500aa69a1c0318796ca4","unresolved":false,"context_lines":[{"line_number":149,"context_line":"  A third party test failure should always be investigated, but the failure of"},{"line_number":150,"context_line":"  a third party test to report in a timely manor should not block others."},{"line_number":151,"context_line":"* We are only talking about the testing of in-tree code. Please note the only"},{"line_number":152,"context_line":"  public API is our REST API, see: doc:`policies`"},{"line_number":153,"context_line":""},{"line_number":154,"context_line":"Release Notes"},{"line_number":155,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"7a740942_09ae7385","line":152,"in_reply_to":"7a740942_ae566c04","updated":"2015-12-04 11:02:24.000000000","message":"gdi","commit_id":"68e7be3ce77b07656acc1ab192cfd27b1751a9cf"}]}