)]}'
{"/COMMIT_MSG":[{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":8,"context_line":""},{"line_number":9,"context_line":"Currently, Cinder services have to be restarted when a change is made"},{"line_number":10,"context_line":"to configuration options in cinder.conf in order to pick up the"},{"line_number":11,"context_line":"modifications. There is no option nor capability of having this changes"},{"line_number":12,"context_line":"picked up automatically when the file is modified."},{"line_number":13,"context_line":""},{"line_number":14,"context_line":"Change-Id: I90b241859917165929189e8fdd6ea36487831c66"}],"source_content_type":"text/x-gerrit-commit-message","patch_set":1,"id":"7af24918_53258608","line":11,"range":{"start_line":11,"start_character":59,"end_line":11,"end_character":63},"updated":"2016-03-04 19:39:03.000000000","message":"these","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"}],"specs/newton/dynamic_reconfiguration.rst":[{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":18,"context_line":"be dealing with services shutting down and spinning back up when some aspect"},{"line_number":19,"context_line":"of configuration gets changed. In order to stay consistent with this, constantly"},{"line_number":20,"context_line":"running services that can be dynamically updated as changes are made provides a"},{"line_number":21,"context_line":"much more realistic view of how user interact with services."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Use Cases"},{"line_number":24,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7af24918_934d4ec7","line":21,"range":{"start_line":21,"start_character":32,"end_line":21,"end_character":36},"updated":"2016-03-04 19:39:03.000000000","message":"users","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":38,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":39,"context_line":"Code will be added to the cinder processes to watch for changes to the"},{"line_number":40,"context_line":"cinder.conf file. Once the processes see that changes have been made,"},{"line_number":41,"context_line":"the process will drain and take the necessary actions to reconfiure itself"},{"line_number":42,"context_line":"and then auto-restart. This capability could also be controlled by a"},{"line_number":43,"context_line":"configuration option so that if the user didn\u0027t want dynamic reconfiguration"},{"line_number":44,"context_line":"enabled, they could just disable it in the cinder.conf file."}],"source_content_type":"text/x-rst","patch_set":1,"id":"7af24918_13b5be7d","line":41,"range":{"start_line":41,"start_character":57,"end_line":41,"end_character":67},"updated":"2016-03-04 19:39:03.000000000","message":"reconfigure","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":57,"context_line":"missed or not flushed properly."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"*SIGHUP Approach: The administrator will send the SIGHUP signal to be caught by"},{"line_number":60,"context_line":"the Cinder processes.  THe processes would then be instrumented to catch the"},{"line_number":61,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":62,"context_line":"The downside to this approach is it still requires the administrator"},{"line_number":63,"context_line":"to take an additional manual step to send the SIGHUP after changing the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7af24918_9380ae8d","line":60,"range":{"start_line":60,"start_character":23,"end_line":60,"end_character":26},"updated":"2016-03-04 19:39:03.000000000","message":"The","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":78,"context_line":"---------------"},{"line_number":79,"context_line":""},{"line_number":80,"context_line":"In order to edit the cinder.conf file, the user must have root authority."},{"line_number":81,"context_line":"So, there is not additional security impact."},{"line_number":82,"context_line":""},{"line_number":83,"context_line":"Notifications impact"},{"line_number":84,"context_line":"--------------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7af24918_73540221","line":81,"range":{"start_line":81,"start_character":13,"end_line":81,"end_character":27},"updated":"2016-03-04 19:39:03.000000000","message":"not an additional","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"14416fbb7ceff95c994552f6440e2c7ef98fd0fb","unresolved":false,"context_lines":[{"line_number":83,"context_line":"Notifications impact"},{"line_number":84,"context_line":"--------------------"},{"line_number":85,"context_line":""},{"line_number":86,"context_line":"None"},{"line_number":87,"context_line":""},{"line_number":88,"context_line":"Other end user impact"},{"line_number":89,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7af24918_c5026909","line":86,"range":{"start_line":86,"start_character":0,"end_line":86,"end_character":4},"updated":"2016-03-04 19:39:03.000000000","message":"I am wondering if there is a notifications impact here.  If we need to monitor in process actions, drain them and then do the reload of configuration, seems like there might be an impact.  At least a notification of when the reload of options starts and ends?  Requires some more thought.","commit_id":"7c0e43bec67b1a61555f6e3f69be1faf476587c1"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"fb854e15d7882a1cf3e1a06d4c80e9c54075dcda","unresolved":false,"context_lines":[{"line_number":23,"context_line":"Use Cases"},{"line_number":24,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":25,"context_line":""},{"line_number":26,"context_line":"*Allows developers testing a new configuration option to change what the"},{"line_number":27,"context_line":"option is set to and test without having to restart all the Cinder services."},{"line_number":28,"context_line":""},{"line_number":29,"context_line":"*Allows developers to test new functionality implemented in a driver by"},{"line_number":30,"context_line":"enabling and disabling configuration options in cinder.conf without"},{"line_number":31,"context_line":"restarting Cinder services each time the option value is changed from"},{"line_number":32,"context_line":"\u0027disabled\u0027 to \u0027enabled\u0027."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"*Allows admins to manage ip addresses for various backends in cinder.conf and"},{"line_number":35,"context_line":"have the connections dynamically update."},{"line_number":36,"context_line":""},{"line_number":37,"context_line":"Proposed change"},{"line_number":38,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1af94dfe_2b6477e3","line":35,"range":{"start_line":26,"start_character":0,"end_line":35,"end_character":40},"updated":"2016-03-16 15:58:30.000000000","message":"You can make this a nice list when you follow RST rules [1].\n\n[1] http://docutils.sourceforge.net/docs/user/rst/quickref.html#bullet-lists","commit_id":"7f6d2911b05fca3dbcd98b2350976fbcfa34bafb"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"fb854e15d7882a1cf3e1a06d4c80e9c54075dcda","unresolved":false,"context_lines":[{"line_number":36,"context_line":""},{"line_number":37,"context_line":"Proposed change"},{"line_number":38,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":39,"context_line":"Code will be added to the cinder processes to watch for changes to the"},{"line_number":40,"context_line":"cinder.conf file. Once the processes see that changes have been made,"},{"line_number":41,"context_line":"the process will drain and take the necessary actions to reconfigure itself"},{"line_number":42,"context_line":"and then auto-restart. This capability could also be controlled by a"},{"line_number":43,"context_line":"configuration option so that if the user didn\u0027t want dynamic reconfiguration"},{"line_number":44,"context_line":"enabled, they could just disable it in the cinder.conf file."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1af94dfe_4bcbc353","line":44,"range":{"start_line":39,"start_character":0,"end_line":44,"end_character":60},"updated":"2016-03-16 15:58:30.000000000","message":"What if the value of config option is saved to a variable (or is part of a variable) and is kept in the memory through all the live of the service process? How do we update such variables? How do we even find places where we do such things?","commit_id":"7f6d2911b05fca3dbcd98b2350976fbcfa34bafb"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"fb854e15d7882a1cf3e1a06d4c80e9c54075dcda","unresolved":false,"context_lines":[{"line_number":56,"context_line":"possibilities for failure since each cache has to be flushed- things could get"},{"line_number":57,"context_line":"missed or not flushed properly."},{"line_number":58,"context_line":""},{"line_number":59,"context_line":"*SIGHUP Approach: The administrator will send the SIGHUP signal to be caught by"},{"line_number":60,"context_line":"the Cinder processes.  The processes would then be instrumented to catch the"},{"line_number":61,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":62,"context_line":"The downside to this approach is it still requires the administrator"},{"line_number":63,"context_line":"to take an additional manual step to send the SIGHUP after changing the"},{"line_number":64,"context_line":"configuration file.  Additionally, this approach requires deployers to"},{"line_number":65,"context_line":"update their sysinit scripts to support reloading the processes."},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Data model impact"},{"line_number":68,"context_line":"-----------------"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1af94dfe_8b4aebc3","line":65,"range":{"start_line":59,"start_character":0,"end_line":65,"end_character":64},"updated":"2016-03-16 15:58:30.000000000","message":"+100 - this will eradicate the issue I\u0027m mentioning earlier. On SIGHUP we\u0027re shutting everything down (cleaning up the memory) and start it again - causing conf options to be reread.","commit_id":"7f6d2911b05fca3dbcd98b2350976fbcfa34bafb"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"fb854e15d7882a1cf3e1a06d4c80e9c54075dcda","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"None"},{"line_number":102,"context_line":""},{"line_number":103,"context_line":"Developer impact"},{"line_number":104,"context_line":"----------------"},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"None"},{"line_number":107,"context_line":""},{"line_number":108,"context_line":""},{"line_number":109,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":2,"id":"1af94dfe_8bc94b7b","line":106,"range":{"start_line":103,"start_character":0,"end_line":106,"end_character":4},"updated":"2016-03-16 15:58:30.000000000","message":"There is a *very* significant impact for developers. With the approach you\u0027re proposing, in the code we can never save a config value to a variable or constant. Let\u0027s say that topic in [1] is normally derived from a config var. That would mean that if we\u0027re keeping the RPC client in the memory (like we do for example in backup.manager), then the config value of that will never be updated with config reload.\n\n[1] https://github.com/openstack/cinder/blob/master/cinder/rpc.py#L181","commit_id":"7f6d2911b05fca3dbcd98b2350976fbcfa34bafb"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"746649ec676a9373a7c1179beea3cd6967adf6be","unresolved":false,"context_lines":[{"line_number":37,"context_line":"the Cinder processes.  The processes would then be instrumented to catch the"},{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file.  Additionally, this approach"},{"line_number":41,"context_line":"requires deployers to update their sysinit scripts to support reloading the"},{"line_number":42,"context_line":"processes. Overall this is a cleaner approach because everything is shut down,"},{"line_number":43,"context_line":"the memory is cleaned up, and then the sysinit scripts will take care of"},{"line_number":44,"context_line":"reloading the processes."},{"line_number":45,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"ba0121b8_b6a2a4ac","line":42,"range":{"start_line":40,"start_character":51,"end_line":42,"end_character":10},"updated":"2016-03-30 10:23:38.000000000","message":"Can you elaborate on that? Do you mean to have \n\n service cinder-scheduler restart\n\nissue SIGHUP to cinder-scheduler?","commit_id":"b7b9b6e336647684f0e941e6627c3b467da72d50"},{"author":{"_account_id":4523,"name":"Eric Harney","email":"eharney@redhat.com","username":"eharney"},"change_message_id":"51e1dae3d1dd401a1f217a928f94894d01535f02","unresolved":false,"context_lines":[{"line_number":37,"context_line":"the Cinder processes.  The processes would then be instrumented to catch the"},{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file.  Additionally, this approach"},{"line_number":41,"context_line":"requires deployers to update their sysinit scripts to support reloading the"},{"line_number":42,"context_line":"processes. Overall this is a cleaner approach because everything is shut down,"},{"line_number":43,"context_line":"the memory is cleaned up, and then the sysinit scripts will take care of"},{"line_number":44,"context_line":"reloading the processes."},{"line_number":45,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a061dce_ebdbe219","line":42,"range":{"start_line":40,"start_character":51,"end_line":42,"end_character":10},"in_reply_to":"ba0121b8_b6a2a4ac","updated":"2016-04-04 15:47:36.000000000","message":"The normal pattern is to implement this as \"service cinder-scheduler reload\".","commit_id":"b7b9b6e336647684f0e941e6627c3b467da72d50"},{"author":{"_account_id":4523,"name":"Eric Harney","email":"eharney@redhat.com","username":"eharney"},"change_message_id":"3319ee5909fa9748cbb25e317c12721c44740fef","unresolved":false,"context_lines":[{"line_number":61,"context_line":"configuration option so that if the user didn\u0027t want dynamic reconfiguration"},{"line_number":62,"context_line":"enabled, they could just disable it in the cinder.conf file. This approach is"},{"line_number":63,"context_line":"dangerous because it wouldn\u0027t account for configuration options that are saved"},{"line_number":64,"context_line":"into variables. To fix these cases, there would be a sizeable impact for"},{"line_number":65,"context_line":"developers finding and replacing all instances of configuration variables."},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"Data model impact"}],"source_content_type":"text/x-rst","patch_set":3,"id":"9a061dce_4b880ee8","line":64,"updated":"2016-04-04 15:50:26.000000000","message":"There are other major reasons this is dangerous, too.  (Like breaking assumptions of deployment and config tools, and packaging.)","commit_id":"b7b9b6e336647684f0e941e6627c3b467da72d50"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"22edcbb0469942d4f0de6d3ee9eff03113d0b7eb","unresolved":false,"context_lines":[{"line_number":95,"context_line":""},{"line_number":96,"context_line":"Other deployer impact"},{"line_number":97,"context_line":"---------------------"},{"line_number":98,"context_line":"None"},{"line_number":99,"context_line":""},{"line_number":100,"context_line":"Developer impact"},{"line_number":101,"context_line":"----------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a122d0e_e8defb3d","line":98,"range":{"start_line":98,"start_character":0,"end_line":98,"end_character":4},"updated":"2016-04-18 19:42:02.000000000","message":"I think we should explicitly list the changes that the deployer needs to make as you have documented above.","commit_id":"eaa9a8f90450142dff6b5f179b7479681831fd7b"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"7a890a173d4b7e8940483a81fa7348a990975ff0","unresolved":false,"context_lines":[{"line_number":115,"context_line":""},{"line_number":116,"context_line":"Work Items"},{"line_number":117,"context_line":"----------"},{"line_number":118,"context_line":"*Implement CLI command"},{"line_number":119,"context_line":"*Write Unit tests"},{"line_number":120,"context_line":"*Testing"},{"line_number":121,"context_line":""},{"line_number":122,"context_line":"Dependencies"},{"line_number":123,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a122d0e_4f5915fe","line":120,"range":{"start_line":118,"start_character":0,"end_line":120,"end_character":8},"updated":"2016-04-18 09:56:12.000000000","message":"Aren\u0027t you missing the actual implementation of handling SIGHUP here?","commit_id":"eaa9a8f90450142dff6b5f179b7479681831fd7b"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"22edcbb0469942d4f0de6d3ee9eff03113d0b7eb","unresolved":false,"context_lines":[{"line_number":127,"context_line":""},{"line_number":128,"context_line":"Documentation Impact"},{"line_number":129,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":130,"context_line":"Adding a configuration option will require a DocImpact. In addition, there"},{"line_number":131,"context_line":"will need to be documentation describing how to use this new capability."},{"line_number":132,"context_line":""},{"line_number":133,"context_line":"References"}],"source_content_type":"text/x-rst","patch_set":5,"id":"1a122d0e_e8b55bee","line":130,"range":{"start_line":130,"start_character":0,"end_line":130,"end_character":55},"updated":"2016-04-18 19:42:02.000000000","message":"The approach you are going with won\u0027t add a config option though.","commit_id":"eaa9a8f90450142dff6b5f179b7479681831fd7b"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"43422b9c37d9fe90a25f83e1390f9d0b4a54810d","unresolved":false,"context_lines":[{"line_number":95,"context_line":""},{"line_number":96,"context_line":"Other deployer impact"},{"line_number":97,"context_line":"---------------------"},{"line_number":98,"context_line":"Update sysinit script to support the new \u0027service cinder-scheduler reload\u0027"},{"line_number":99,"context_line":"command."},{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1a122d0e_cef7c7f3","line":98,"range":{"start_line":98,"start_character":57,"end_line":98,"end_character":66},"updated":"2016-04-18 20:32:01.000000000","message":"I think this should say \u0027\u003cservice\u003e\u0027 as the scripts for each of the different scripts will need to be updated.","commit_id":"53f86c456ab9bd9496b5797a6b46e11b38ab7cb5"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"43422b9c37d9fe90a25f83e1390f9d0b4a54810d","unresolved":false,"context_lines":[{"line_number":95,"context_line":""},{"line_number":96,"context_line":"Other deployer impact"},{"line_number":97,"context_line":"---------------------"},{"line_number":98,"context_line":"Update sysinit script to support the new \u0027service cinder-scheduler reload\u0027"},{"line_number":99,"context_line":"command."},{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1a122d0e_2edf5b6a","line":98,"range":{"start_line":98,"start_character":15,"end_line":98,"end_character":21},"updated":"2016-04-18 20:32:01.000000000","message":"scripts","commit_id":"53f86c456ab9bd9496b5797a6b46e11b38ab7cb5"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"43422b9c37d9fe90a25f83e1390f9d0b4a54810d","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement CLI command that sends SIGHUP signal"},{"line_number":120,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":121,"context_line":"*Write Unit tests"},{"line_number":122,"context_line":"*Testing"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1a122d0e_8ed4af47","line":119,"range":{"start_line":119,"start_character":0,"end_line":119,"end_character":47},"updated":"2016-04-18 20:32:01.000000000","message":"There shouldn\u0027t be a CLI impact since it will be up to the deployers to set up there sysinit scripts to send a sighup.","commit_id":"53f86c456ab9bd9496b5797a6b46e11b38ab7cb5"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"cd8fbe3cd89e891493ee1aa9c9790d93526da7d1","unresolved":false,"context_lines":[{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file. A command would be"},{"line_number":41,"context_line":"implemented in the CLI, \u0027cinder-\u003cservice\u003e reload\u0027, that would issue the SIGHUP"},{"line_number":42,"context_line":"signal to cinder-scheduler that changes have been made and a reload is"},{"line_number":43,"context_line":"necessary. Additionally, this approach requires deployers to update their"},{"line_number":44,"context_line":"sysinit scripts to support reloading the processes.  Overall, this is a"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_83e62e13","line":41,"range":{"start_line":41,"start_character":25,"end_line":41,"end_character":48},"updated":"2016-04-19 11:15:21.000000000","message":"Oh please, no. This will get our CLI for services to be significantly different than other OpenStack projects. And how do you plan to detect PID of cinder-\u003cservice\u003e to target with the signal? cinder-\u003cservice\u003e doesn\u0027t have a pidfile available.\n\nWhy not simply let operator handle that through his scripting and sysinit?\n\nIf you really want to implement that helper it\u0027s better to place it in cinder-manage as it gathers similar tools. But then please include the way you want to handling finding a correct PID of the service, as I don\u0027t think it\u0027s trivial.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"df6e8c21752124002c1f293c7e5b5d4c8a8e7734","unresolved":false,"context_lines":[{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file. A command would be"},{"line_number":41,"context_line":"implemented in the CLI, \u0027cinder-\u003cservice\u003e reload\u0027, that would issue the SIGHUP"},{"line_number":42,"context_line":"signal to cinder-scheduler that changes have been made and a reload is"},{"line_number":43,"context_line":"necessary. Additionally, this approach requires deployers to update their"},{"line_number":44,"context_line":"sysinit scripts to support reloading the processes.  Overall, this is a"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_f72f3c86","line":41,"range":{"start_line":41,"start_character":25,"end_line":41,"end_character":48},"in_reply_to":"1a122d0e_0961d0a8","updated":"2016-05-02 16:03:11.000000000","message":"Michal and Patrick, I think there is some confusion here.  Oh, and I see why ... the CLI part should be removed.  What was intended would be that the operator would handle it through sysinit scripting.  For RHEL I think that is still as cinder-\u003cservice name\u003e reload or it could be through systemd for other distros. \n\nKendall,\n\nSo, I think we maybe want to reword this to make it less confusing.  Lets remove:  \"A command would be\nimplemented in the CLI, \u0027cinder-\u003cservice\u003e reload\u0027, that would issue the SIGHUP signal to cinder-scheduler that changes have been made and a reload is necessary.\"  Just say \"This approach requires deployers to update their sysinit scripts to support reloading the processes.\"","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":12924,"name":"Patrick East","email":"east.patrick@gmail.com","username":"patrick.east"},"change_message_id":"62e8e346188790a66ff5355102a9c3d8bfbebfa8","unresolved":false,"context_lines":[{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file. A command would be"},{"line_number":41,"context_line":"implemented in the CLI, \u0027cinder-\u003cservice\u003e reload\u0027, that would issue the SIGHUP"},{"line_number":42,"context_line":"signal to cinder-scheduler that changes have been made and a reload is"},{"line_number":43,"context_line":"necessary. Additionally, this approach requires deployers to update their"},{"line_number":44,"context_line":"sysinit scripts to support reloading the processes.  Overall, this is a"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_0961d0a8","line":41,"range":{"start_line":41,"start_character":25,"end_line":41,"end_character":48},"in_reply_to":"1a122d0e_83e62e13","updated":"2016-04-22 21:02:17.000000000","message":"I tend to agree with Michal on this one, we should try and keep our CLI from being too different than other services in OpenStack.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"d3fd061c90dcc6e18d8627d61da03ac6032b6831","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_047989f6","line":103,"updated":"2016-04-19 10:39:19.000000000","message":"I think there\u0027s significant developer impact here - code will need to be designed (And in some cases redesigned) to be reloadable - any caches affected by config options need to be cleared on SIGHUP, existing connections need to be cleanly dropped (work in progress finished), anything calculated from a config option needs to be recalculated... this is non-trivial and will take developer and reviewer effort to catch.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"cd8fbe3cd89e891493ee1aa9c9790d93526da7d1","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_c39ae66d","line":103,"in_reply_to":"1a122d0e_047989f6","updated":"2016-04-19 11:15:21.000000000","message":"It\u0027s not *that* bad. SIGHUP sent to oslo.services kills all the service processes besides main one, so you should be safe unless you\u0027re using global variables inside the main thread (is it even possible?).","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"df6e8c21752124002c1f293c7e5b5d4c8a8e7734","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_5799d0cf","line":103,"in_reply_to":"1a122d0e_292c8c49","updated":"2016-05-02 16:03:11.000000000","message":"Patrick, good idea to add this to the devref.  I think that we probably should add a note here that driver developers will want to test the SIGHUP/reload functionality with their drivers to ensure they behave as anticipated.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":12924,"name":"Patrick East","email":"east.patrick@gmail.com","username":"patrick.east"},"change_message_id":"62e8e346188790a66ff5355102a9c3d8bfbebfa8","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_292c8c49","line":103,"in_reply_to":"1a122d0e_4d679dea","updated":"2016-04-22 21:02:17.000000000","message":"We might want to have that added to the devref or something, at least for drivers, so that new folks will know what happens when the config is reloaded.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"4e72150006852ae31dbaa60b07cc98e72caef9f6","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_4d679dea","line":103,"in_reply_to":"1a122d0e_b6dc3974","updated":"2016-04-21 13:45:11.000000000","message":"Oh, sorry, I could have worded it better. Actually it stops RPC server and lets the threads finish in progress work.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"9a0374422276ffb8367b27c9b39add6bb26477a2","unresolved":false,"context_lines":[{"line_number":100,"context_line":""},{"line_number":101,"context_line":"Developer impact"},{"line_number":102,"context_line":"----------------"},{"line_number":103,"context_line":"None"},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""},{"line_number":106,"context_line":"Implementation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_b6dc3974","line":103,"in_reply_to":"1a122d0e_c39ae66d","updated":"2016-04-19 15:10:41.000000000","message":"If you blindly kill all your worker threads, then you loose all you rwork in progress, and the only cleanup we have for that currently is in early manager startup. We need to make this clean regarding work in progress, else it\u0027s no better than kill \u0026 restart.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"d3fd061c90dcc6e18d8627d61da03ac6032b6831","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":120,"context_line":"*Write Unit tests"},{"line_number":121,"context_line":"*Testing"},{"line_number":122,"context_line":"*Update Documentation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_845bd986","line":119,"range":{"start_line":119,"start_character":30,"end_line":119,"end_character":36},"updated":"2016-04-19 10:39:19.000000000","message":"Audit existing code to ensure that caches are cleared, dependant variables are recalculated, connections are cleanly dropped, etc","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"df6e8c21752124002c1f293c7e5b5d4c8a8e7734","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":120,"context_line":"*Write Unit tests"},{"line_number":121,"context_line":"*Testing"},{"line_number":122,"context_line":"*Update Documentation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_f7673cb5","line":119,"range":{"start_line":119,"start_character":30,"end_line":119,"end_character":36},"in_reply_to":"1a122d0e_6d8c1980","updated":"2016-05-02 16:03:11.000000000","message":"Duncan, can you help Kendall understand what kind of things we need to look for here?","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"4e72150006852ae31dbaa60b07cc98e72caef9f6","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":120,"context_line":"*Write Unit tests"},{"line_number":121,"context_line":"*Testing"},{"line_number":122,"context_line":"*Update Documentation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_6d8c1980","line":119,"range":{"start_line":119,"start_character":30,"end_line":119,"end_character":36},"in_reply_to":"1a122d0e_7612a10a","updated":"2016-04-21 13:45:11.000000000","message":"I think workers are actually new processes, not threads. Just adding a data point.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"cd8fbe3cd89e891493ee1aa9c9790d93526da7d1","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":120,"context_line":"*Write Unit tests"},{"line_number":121,"context_line":"*Testing"},{"line_number":122,"context_line":"*Update Documentation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_c3714646","line":119,"range":{"start_line":119,"start_character":30,"end_line":119,"end_character":36},"in_reply_to":"1a122d0e_845bd986","updated":"2016-04-19 11:15:21.000000000","message":"See my previous comment.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":1207,"name":"Duncan Thomas","email":"duncan.thomas@gmail.com","username":"duncan-thomas"},"change_message_id":"9a0374422276ffb8367b27c9b39add6bb26477a2","unresolved":false,"context_lines":[{"line_number":116,"context_line":""},{"line_number":117,"context_line":"Work Items"},{"line_number":118,"context_line":"----------"},{"line_number":119,"context_line":"*Implement handling of SIGHUP signal"},{"line_number":120,"context_line":"*Write Unit tests"},{"line_number":121,"context_line":"*Testing"},{"line_number":122,"context_line":"*Update Documentation"}],"source_content_type":"text/x-rst","patch_set":9,"id":"1a122d0e_7612a10a","line":119,"range":{"start_line":119,"start_character":30,"end_line":119,"end_character":36},"in_reply_to":"1a122d0e_c3714646","updated":"2016-04-19 15:10:41.000000000","message":"If the cache is done as a singleton, then it will be global across threads. Similar scenarios are possible, we need to look for them. I don\u0027t know for certain that they are there, but looking at the ones we know of (db connection pool, ssh connection pools in drivers, rabbit, ...) and confirming they look fine should be part of this work.","commit_id":"6a48dab79805c36c9a7ec1fb2e6d88a3fd682b65"},{"author":{"_account_id":9171,"name":"Vipin Balachandran","email":"vipin.bl@gmail.com","username":"vbala"},"change_message_id":"e3ae44aad089750348321ae0492099f2793ced18","unresolved":false,"context_lines":[{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":"SIGHUP Approach: The administrator will send the SIGHUP signal to be caught by"},{"line_number":37,"context_line":"the Cinder processes.  The processes would then be instrumented to catch the"},{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file. This approach requires"},{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_cd99a3d2","line":38,"range":{"start_line":38,"start_character":38,"end_line":38,"end_character":67},"updated":"2016-05-03 21:29:53.000000000","message":"I assume this will also take care of re-initializing the drivers.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"48101dd1b14bb18dd37da362e9d07d9ed05288a0","unresolved":false,"context_lines":[{"line_number":35,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":36,"context_line":"SIGHUP Approach: The administrator will send the SIGHUP signal to be caught by"},{"line_number":37,"context_line":"the Cinder processes.  The processes would then be instrumented to catch the"},{"line_number":38,"context_line":"signal, drain in progress actions and reload the configuration file."},{"line_number":39,"context_line":"It still requires the administrator to take an additional manual step to send"},{"line_number":40,"context_line":"the SIGHUP after changing the configuration file. This approach requires"},{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_f610ac02","line":38,"range":{"start_line":38,"start_character":38,"end_line":38,"end_character":67},"in_reply_to":"1a122d0e_cd99a3d2","updated":"2016-05-03 23:21:53.000000000","message":"Good point.  Maybe should be explicit about that.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":5997,"name":"Walt","display_name":"Hemna","email":"waboring@hemna.com","username":"walter-boring","status":"SAP"},"change_message_id":"253d7b37c0d989ffb7ec13d40b7f5f5e86a763d1","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_f059d4fe","line":44,"updated":"2016-05-03 22:05:16.000000000","message":"You should also mention this is a per host basis.  If you have a multi-node setup, each node will have to be signaled.  The same is the case with an HA setup with multiple cinder nodes spread across many more machines.\n\nI can see a problem where you have an Active/Active HA setup deployed talking to the same backend.  You modify a driver attribute on host 1, do a SIGHUP on host 1.  Host 2 is now out of date and provisioning requests to create a volume on Host1 will be potentially different than volumes created from Host2 until host2\u0027s cinder.conf is in sync with host 1.\n\nWhat are we going to do in this case?","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"ecbda2eadd43d5e193a858b8b570c5f04b93bd75","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"bab6814e_0be37234","line":44,"in_reply_to":"1a122d0e_11f9b74e","updated":"2016-05-25 12:58:48.000000000","message":"There\u0027s a simple and obvious solution. Just scale down your c-vol backend cluster to one node, reload its config and restart rest of the nodes (also with new config in place). That\u0027s how you would do it anyway without dynamic reloading, so it\u0027s worth documenting, but on the side of c-vol A/A efforts.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":5997,"name":"Walt","display_name":"Hemna","email":"waboring@hemna.com","username":"walter-boring","status":"SAP"},"change_message_id":"175b9440db78c658fbd8c17312a7129fffeb32a0","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_f88ce581","line":44,"in_reply_to":"1a122d0e_e947014a","updated":"2016-05-04 22:13:29.000000000","message":"All I\u0027m saying is that we need to document this as a potential for major problems.   Just think about replication being on in one node, and off on another.  bad stuffs.  \n\nAlso, the spec should mention that this use case isn\u0027t covered and needs to be called out.   Active/active brings out all the worts in things like this and we need to be very clear what we support and don\u0027t.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"48101dd1b14bb18dd37da362e9d07d9ed05288a0","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_f6deec3c","line":44,"in_reply_to":"1a122d0e_f059d4fe","updated":"2016-05-03 23:21:53.000000000","message":"Oy vey.  :-)  So, this is a case where I want to make sure we don\u0027t let striving for perfection get in the way of progress.  It would be nice if there were some way to handle signalling multiple hosts, but I am not sure what that would be right now.  I think we should start by documenting the limitation here that you need to do this for each node involved and across nodes of an HA cluster.  This limitation, however, is no worse than the current situation.  Currently the changes would need to be propagated and each service restarted.  Now they can be propagated and the service reloaded.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":12924,"name":"Patrick East","email":"east.patrick@gmail.com","username":"patrick.east"},"change_message_id":"76ea40e19ef0bc77e347df6b78bdfd28d67baf02","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_e947014a","line":44,"in_reply_to":"1a122d0e_f6deec3c","updated":"2016-05-04 19:51:53.000000000","message":"I agree with Jay on this... we have that problem today with or without the sighup suppport to reload without restarting. Its worth mentioning that this *doesn\u0027t* solve that, but not worth trying to solve it with these changes.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"abc4bc6884326f20608152afda93ea27ceb4b6d2","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_11f9b74e","line":44,"in_reply_to":"1a122d0e_f88ce581","updated":"2016-05-05 04:01:18.000000000","message":"Agreed that we need to document the limitations and possible areas where things may go wrong.  No argument there.  :-)","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":7198,"name":"Jay Bryant","email":"jungleboyj@electronicjungle.net","username":"jsbryant"},"change_message_id":"3838fe886f25e670b0515affb2a616683bcf6b5b","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"bab6814e_a53c06c0","line":44,"in_reply_to":"bab6814e_0be37234","updated":"2016-05-25 15:39:30.000000000","message":"So, Kendall and I are trying to figure out what is really being asked for here.  Is it sufficient to document that this dynamic reconfiguration functionality doesn\u0027t prevent the user from configuring one node of an A/A c-vol configuration with incompatible options and that it is recommended that the cluster be scaled down to a single node while options are changed and then scaled back up?  It also seems to me that that is really something that should be documented for the A/A c-vol configuration regardless of whether dynamic reconfiguration is possible.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"1b4dc16d6c7b253e28cca8a10a533f49cf95f669","unresolved":false,"context_lines":[{"line_number":41,"context_line":"deployers to update their sysinit scripts to support reloading the processes."},{"line_number":42,"context_line":"Overall, this is a cleaner approach because everything is shut down, the memory"},{"line_number":43,"context_line":"is cleaned up, and then the sysinit scripts will take care of reloading"},{"line_number":44,"context_line":"the processes."},{"line_number":45,"context_line":""},{"line_number":46,"context_line":"Alternatives"},{"line_number":47,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"bab6814e_fc84bdd3","line":44,"in_reply_to":"bab6814e_a53c06c0","updated":"2016-05-26 11:24:34.000000000","message":"I can agree with Jay. To me it\u0027s enough to say that we\u0027re not centralizing configs anyhow with this spec - configs are still local.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":9171,"name":"Vipin Balachandran","email":"vipin.bl@gmail.com","username":"vbala"},"change_message_id":"e3ae44aad089750348321ae0492099f2793ced18","unresolved":false,"context_lines":[{"line_number":98,"context_line":""},{"line_number":99,"context_line":"Developer impact"},{"line_number":100,"context_line":"----------------"},{"line_number":101,"context_line":"Update devref to describe SIGHUP/reload process, and test functionality with"},{"line_number":102,"context_line":"vendor\u0027s drivers to ensure correct behavior."},{"line_number":103,"context_line":""},{"line_number":104,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"1a122d0e_4ddc7374","line":101,"range":{"start_line":101,"start_character":0,"end_line":101,"end_character":47},"updated":"2016-05-03 21:29:53.000000000","message":"This sounds like a work item.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"ecbda2eadd43d5e193a858b8b570c5f04b93bd75","unresolved":false,"context_lines":[{"line_number":98,"context_line":""},{"line_number":99,"context_line":"Developer impact"},{"line_number":100,"context_line":"----------------"},{"line_number":101,"context_line":"Update devref to describe SIGHUP/reload process, and test functionality with"},{"line_number":102,"context_line":"vendor\u0027s drivers to ensure correct behavior."},{"line_number":103,"context_line":""},{"line_number":104,"context_line":""}],"source_content_type":"text/x-rst","patch_set":10,"id":"bab6814e_8bb762f4","line":101,"range":{"start_line":101,"start_character":0,"end_line":101,"end_character":47},"in_reply_to":"1a122d0e_4ddc7374","updated":"2016-05-25 12:58:48.000000000","message":"Yes, that\u0027s not really impact for developers.","commit_id":"817f881ddd3a206b6514e4a4c1a60a9fb9052a05"},{"author":{"_account_id":11600,"name":"Michał Dulko","email":"michal.dulko@gmail.com","username":"dulek"},"change_message_id":"d9b8fb61c08381c04e17e527a23bf4c78ac8c857","unresolved":false,"context_lines":[{"line_number":124,"context_line":"**Connections are cleanly dropped"},{"line_number":125,"context_line":"*Write Unit tests"},{"line_number":126,"context_line":"*Testing on a variety of drivers"},{"line_number":127,"context_line":"*Update devref to describe SIGHUP/reload process"},{"line_number":128,"context_line":"Dependencies"},{"line_number":129,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":130,"context_line":""}],"source_content_type":"text/x-rst","patch_set":11,"id":"bab6814e_fce95d80","line":127,"updated":"2016-05-26 11:27:02.000000000","message":"Missing newline causes py27 failure. I think. ;)","commit_id":"bf908cb84be04c4f18cff6e061deed80e5c9f575"}]}
