)]}'
{"specs/kilo/elasticsearch-event-db.rst":[{"author":{"_account_id":11564,"name":"Chris Dent","email":"cdent@anticdent.org","username":"chdent"},"change_message_id":"fbaf0c6c9de7a1103408160073327e10f4912ba7","unresolved":false,"context_lines":[{"line_number":110,"context_line":"  chungg"},{"line_number":111,"context_line":""},{"line_number":112,"context_line":"Ongoing maintainer:"},{"line_number":113,"context_line":"  chungg "},{"line_number":114,"context_line":""},{"line_number":115,"context_line":"Work Items"},{"line_number":116,"context_line":"----------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"baa201ad_cceb47da","line":113,"updated":"2014-10-07 14:39:24.000000000","message":"omg whitespace","commit_id":"11a87766c241430423d88c3e041c6665f6a87af0"},{"author":{"_account_id":3153,"name":"Emilien Macchi","email":"emilien@redhat.com","username":"emilienm"},"change_message_id":"6f90ab7fcef80a2c18786aa656b4602d944f48b4","unresolved":false,"context_lines":[{"line_number":91,"context_line":"Performance/Scalability Impacts"},{"line_number":92,"context_line":"-------------------------------"},{"line_number":93,"context_line":""},{"line_number":94,"context_line":"None, as it\u0027s an optional backend. ElasticSearch has the ability to cluster to"},{"line_number":95,"context_line":"provide HA and it is built to scale horizontally[6]"},{"line_number":96,"context_line":""},{"line_number":97,"context_line":"Other deployer impact"}],"source_content_type":"text/x-rst","patch_set":2,"id":"9aa7fdbe_f9d97922","line":94,"updated":"2014-10-28 16:07:50.000000000","message":"I think you should mention the eventual consistency issue in ES.","commit_id":"63b42de8b1460bbb6d84550f53bcca293239ea20"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"45a484ad27d8023db1935856d72ef8147704f22b","unresolved":false,"context_lines":[{"line_number":18,"context_line":""},{"line_number":19,"context_line":"While SQL is able to store unstructured data, it\u0027s true use case is to"},{"line_number":20,"context_line":"effectively store data in a defined schema and it\u0027s relationships so that"},{"line_number":21,"context_line":"it\u0027s easily queryable."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Events in OpenStack are for the most part schema-free; each notification"},{"line_number":24,"context_line":"message may contain any combination of attributes based on when and where the"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_38ed69b8","line":21,"updated":"2014-12-17 13:46:12.000000000","message":"My understanding of the purpose of the event traits model was to effectively convert unstructured event data into a collection of strongly-typed (int, string, datetime etc.) traits that could be queried over in a more structured way.\n\ni.e. is it totally accurate to think about our current events storage as unstructured data?","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"70497eaf8bdab3c6be841d60270f771e3a7fe937","unresolved":false,"context_lines":[{"line_number":18,"context_line":""},{"line_number":19,"context_line":"While SQL is able to store unstructured data, it\u0027s true use case is to"},{"line_number":20,"context_line":"effectively store data in a defined schema and it\u0027s relationships so that"},{"line_number":21,"context_line":"it\u0027s easily queryable."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Events in OpenStack are for the most part schema-free; each notification"},{"line_number":24,"context_line":"message may contain any combination of attributes based on when and where the"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_1b7ae487","line":21,"in_reply_to":"3a961159_38ed69b8","updated":"2014-12-22 18:43:19.000000000","message":"correct, it\u0027s not really unstructured. the comment was more that there is no standard schema... just collection of k-v pairs.\n\nthat said, what i kind of envision the ES backend is for error messages (which we currently ignore)... i would imagine error notifications are more rich in text which is something elasticsearch is good for.","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":3012,"name":"Dina Belova","email":"dbelova@mirantis.com","username":"dbelova"},"change_message_id":"c94172c445fbfd23f4b7aaf915474225629accc8","unresolved":false,"context_lines":[{"line_number":20,"context_line":"effectively store data in a defined schema and it\u0027s relationships so that"},{"line_number":21,"context_line":"it\u0027s easily queryable."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Events in OpenStack are for the most part schema-free; each notification"},{"line_number":24,"context_line":"message may contain any combination of attributes based on when and where the"},{"line_number":25,"context_line":"message originated from."},{"line_number":26,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_3185d159","line":23,"range":{"start_line":23,"start_character":42,"end_line":23,"end_character":53},"updated":"2014-12-15 13:44:05.000000000","message":"I hope to see that fixed in OpenStack soon :)","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"45a484ad27d8023db1935856d72ef8147704f22b","unresolved":false,"context_lines":[{"line_number":20,"context_line":"effectively store data in a defined schema and it\u0027s relationships so that"},{"line_number":21,"context_line":"it\u0027s easily queryable."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Events in OpenStack are for the most part schema-free; each notification"},{"line_number":24,"context_line":"message may contain any combination of attributes based on when and where the"},{"line_number":25,"context_line":"message originated from."},{"line_number":26,"context_line":""}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_0ebf84fd","line":23,"range":{"start_line":23,"start_character":42,"end_line":23,"end_character":53},"in_reply_to":"3a961159_3185d159","updated":"2014-12-17 13:46:12.000000000","message":"Yep, if sandywalsh has his way, notifications will be schematized for kilo.","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":3012,"name":"Dina Belova","email":"dbelova@mirantis.com","username":"dbelova"},"change_message_id":"c94172c445fbfd23f4b7aaf915474225629accc8","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_11b8d50f","line":34,"updated":"2014-12-15 13:44:05.000000000","message":"How this will be coordinated with ghoccni\u0027s idea of how events will be stored? I mean  record_event() idea we\u0027ve briefly went through during the summit? https://etherpad.openstack.org/p/kilo-ceilometer-gnocchi-integration","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c7940533aee57180f231df2744ee59410c535738","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_54d96b55","line":34,"in_reply_to":"3a961159_11b8d50f","updated":"2014-12-15 14:13:38.000000000","message":"this won\u0027t. afaik, events isn\u0027t in scope for gnocchi... just metric data. i believe the point you referenced in the etherpad is that resource data in gnocchi will only be updated by notification based data.","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"45a484ad27d8023db1935856d72ef8147704f22b","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_cea81cde","line":34,"in_reply_to":"3a961159_348bff1f","updated":"2014-12-17 13:46:12.000000000","message":"@Dina: so I don\u0027t really see the issue here around events, my understanding was that the whole purpose of splitting out the events \u0026 alarms storage logic from meters was to facilitate a mixed approach for gnocchi (i.e. timeseries \u0026 resource data only in gnocchi, events \u0026 alarms elsewhere)","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":3012,"name":"Dina Belova","email":"dbelova@mirantis.com","username":"dbelova"},"change_message_id":"7bb540cdb0c0ee0d513a4ba8ae84dcfb63ba6828","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_348bff1f","line":34,"in_reply_to":"3a961159_54d96b55","updated":"2014-12-15 14:15:40.000000000","message":"a-ha, ok","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":3012,"name":"Dina Belova","email":"dbelova@mirantis.com","username":"dbelova"},"change_message_id":"2bd4aada6efd12ca95dddd5d0c5150c889a1fcda","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_c258d258","line":34,"in_reply_to":"3a961159_843a7396","updated":"2014-12-18 07:44:49.000000000","message":"Well, yeah, it looks like the solution that cannot be used in all the cases. Ok, I got it, thanks!","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"2c0f6500fbb45acce1eb4d111a2fa12636be0f28","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_843a7396","line":34,"in_reply_to":"3a961159_ce857c08","updated":"2014-12-17 14:42:57.000000000","message":"Dina: a-ha, I get it.\n\nYes, gnocchi resource creation could be a side-effect of the first \u003cresource\u003e.create.start style event, though I think that aspect of event processing is orthogonal to the persistence of the actual event payload for later queries over the event traits.\n\n(Also I\u0027m not 100% sure that there an resource-creation event for every resource-type that we\u0027ll need to represent in gnocchi)","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":3012,"name":"Dina Belova","email":"dbelova@mirantis.com","username":"dbelova"},"change_message_id":"d5419e839eae38ffb3eb55d7cc21c50fee71ce90","unresolved":false,"context_lines":[{"line_number":31,"context_line":"similar to the events in OpenStack and has worked with relative success in"},{"line_number":32,"context_line":"other projects[1]."},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"ElasticSearch will be added in as an alternative backend to the current"},{"line_number":35,"context_line":"offerings of HBase, MongoDB, and SQL. The current implementation of filtering"},{"line_number":36,"context_line":"attributes using a definition file will continue to be used. A Logstash"},{"line_number":37,"context_line":"implementation of capturing and processing events is not in the scope of this"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_ce857c08","line":34,"in_reply_to":"3a961159_cea81cde","updated":"2014-12-17 13:52:33.000000000","message":"Eoghan, a-ha, cool. So basically datastore separation between events and meters will allow us quickly go to the gnocchi without any changes to the events features. I\u0027m kind of thinking - can we create resources in gnocchi as a result of \u0027create-instance\u0027 events for gnocchi? That will make measures writing to the gnocchi faster in case if measures are coming for the resource, never met by gnocchi before. What do you think about this? It does not refer to this change, I\u0027m just thinking about it","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":7336,"name":"Phil Neal","email":"phil.neal@hp.com","username":"nealph"},"change_message_id":"a1b1e6084f3c4d24472da12099d7c65328c664f2","unresolved":false,"context_lines":[{"line_number":108,"context_line":"Developer impact"},{"line_number":109,"context_line":"----------------"},{"line_number":110,"context_line":""},{"line_number":111,"context_line":"Devs will need to use ElasticSearch client when modifying ElasticSearch driver"},{"line_number":112,"context_line":""},{"line_number":113,"context_line":"Implementation"},{"line_number":114,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_2e417539","line":111,"updated":"2014-12-18 15:03:59.000000000","message":"can you clarify what config change would be likely? I\u0027m mainly concerned that the out-of-box setup is sufficient for deployment...is this saying that client/driver modification is only a developer concern if adding capabilities?","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"70497eaf8bdab3c6be841d60270f771e3a7fe937","unresolved":false,"context_lines":[{"line_number":108,"context_line":"Developer impact"},{"line_number":109,"context_line":"----------------"},{"line_number":110,"context_line":""},{"line_number":111,"context_line":"Devs will need to use ElasticSearch client when modifying ElasticSearch driver"},{"line_number":112,"context_line":""},{"line_number":113,"context_line":"Implementation"},{"line_number":114,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_fb28b887","line":111,"in_reply_to":"3a961159_2e417539","updated":"2014-12-22 18:43:19.000000000","message":"this is more just a general statement. if you want to edit es backend functionality, we\u0027re using es client to connect and make requests.  nothing markedly different from the other backends.","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"45a484ad27d8023db1935856d72ef8147704f22b","unresolved":false,"context_lines":[{"line_number":143,"context_line":"Testing"},{"line_number":144,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":145,"context_line":""},{"line_number":146,"context_line":"testing will need to be done against a mocked db or a real ElasticSearch database"},{"line_number":147,"context_line":""},{"line_number":148,"context_line":"Documentation Impact"},{"line_number":149,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_eedcf876","line":146,"updated":"2014-12-17 13:46:12.000000000","message":"In the real database case, would you invisage ElasticSearch being installed directly on the CI nodes, or are you thinking more in terms of a remote hosted ElasticSearch-as-a-Service?","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"70497eaf8bdab3c6be841d60270f771e3a7fe937","unresolved":false,"context_lines":[{"line_number":143,"context_line":"Testing"},{"line_number":144,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":145,"context_line":""},{"line_number":146,"context_line":"testing will need to be done against a mocked db or a real ElasticSearch database"},{"line_number":147,"context_line":""},{"line_number":148,"context_line":"Documentation Impact"},{"line_number":149,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"}],"source_content_type":"text/x-rst","patch_set":4,"id":"3a961159_fb9d18bd","line":146,"in_reply_to":"3a961159_eedcf876","updated":"2014-12-22 18:43:19.000000000","message":"based on conversation with clarkb, the recommended solution is to have the job pull in elasticsearch and install it... not have it installed directly on CI node.","commit_id":"81913b4fcea079d3330554231423314d05ffa2e0"}]}