)]}'
{"specs/ocata/ceilometer-status.rst":[{"author":{"_account_id":18336,"name":"Sumant Murke","email":"sumant.murke@intel.com","username":"sumantmurke"},"change_message_id":"915c8a928a79ad45ecce53f1d258d597ed665095","unresolved":false,"context_lines":[{"line_number":32,"context_line":"Proposed change"},{"line_number":33,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":34,"context_line":""},{"line_number":35,"context_line":"Introduce a command-line tool similar to keyston-manage or nova-manage to be"},{"line_number":36,"context_line":"used by, or on behalf of, operators to gain insight into the operation of"},{"line_number":37,"context_line":"Ceilometer."},{"line_number":38,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_34832358","line":35,"range":{"start_line":35,"start_character":41,"end_line":35,"end_character":48},"updated":"2016-08-22 18:24:45.000000000","message":"Nit: keystone","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c1d75fc62ce4a37d785500ececd005133ee417c0","unresolved":false,"context_lines":[{"line_number":38,"context_line":""},{"line_number":39,"context_line":"This may include several different sets of reports for the command line tool:"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_7be0a7cd","line":41,"updated":"2016-08-23 15:01:42.000000000","message":"is this done at per agent level? i don\u0027t know how else you\u0027d achieve this metric?","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"d894012e3984f93ff48475e7a5b0ad9ef2707331","unresolved":false,"context_lines":[{"line_number":38,"context_line":""},{"line_number":39,"context_line":"This may include several different sets of reports for the command line tool:"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_4d98831c","line":41,"in_reply_to":"1ac06dbe_7be0a7cd","updated":"2016-08-23 19:40:01.000000000","message":"per agent, yes.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c1d75fc62ce4a37d785500ececd005133ee417c0","unresolved":false,"context_lines":[{"line_number":39,"context_line":"This may include several different sets of reports for the command line tool:"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_5b668b51","line":42,"updated":"2016-08-23 15:01:42.000000000","message":"i imagine this is again per notification level metric?","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"d894012e3984f93ff48475e7a5b0ad9ef2707331","unresolved":false,"context_lines":[{"line_number":39,"context_line":"This may include several different sets of reports for the command line tool:"},{"line_number":40,"context_line":""},{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_ad75b7cb","line":42,"in_reply_to":"1ac06dbe_5b668b51","updated":"2016-08-23 19:40:01.000000000","message":"per agent, yes.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c1d75fc62ce4a37d785500ececd005133ee417c0","unresolved":false,"context_lines":[{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"},{"line_number":46,"context_line":"  each part of expected infrastructure."},{"line_number":47,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_bbe7bfaa","line":44,"updated":"2016-08-23 15:01:42.000000000","message":"is this rabbitmq specific? i don\u0027t think there\u0027s a way to do this in oslo.messaging.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"d894012e3984f93ff48475e7a5b0ad9ef2707331","unresolved":false,"context_lines":[{"line_number":41,"context_line":"* A counter of resources and metrics tracked in a floating time window"},{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"},{"line_number":46,"context_line":"  each part of expected infrastructure."},{"line_number":47,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_3088fc70","line":44,"in_reply_to":"1ac06dbe_bbe7bfaa","updated":"2016-08-23 19:40:01.000000000","message":"good point, I am not seeing a way to do this with oslo.messaging. This is one of the primary diagnostics so I would like to have a way to support it but it may require a driver mechanism for each bus. Maybe this gets deferred based on time and complexity.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c1d75fc62ce4a37d785500ececd005133ee417c0","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"},{"line_number":46,"context_line":"  each part of expected infrastructure."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"The Ceilometer API would incorporate the Oslo healthcheck middleware which"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_fb7ad7bb","line":45,"updated":"2016-08-23 15:01:42.000000000","message":"this seems useful. is this just a notificaiton on start/stop? i don\u0027t know how you\u0027d poll to ensure pollsters were still active.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"d894012e3984f93ff48475e7a5b0ad9ef2707331","unresolved":false,"context_lines":[{"line_number":42,"context_line":"* A counter of error, message, event, and metrics process in a floating time"},{"line_number":43,"context_line":"  window."},{"line_number":44,"context_line":"* Message bus volumes waiting to be processed, by component."},{"line_number":45,"context_line":"* An overall \"Up/Down\" health status, by component, indicating connectivity to"},{"line_number":46,"context_line":"  each part of expected infrastructure."},{"line_number":47,"context_line":""},{"line_number":48,"context_line":"The Ceilometer API would incorporate the Oslo healthcheck middleware which"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_10e7c040","line":45,"in_reply_to":"1ac06dbe_fb7ad7bb","updated":"2016-08-23 19:40:01.000000000","message":"if there is a notification for the polling loop when it enters/exits waiting state in addition, that would probably be good enough to complete the picture for pollsters.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c1d75fc62ce4a37d785500ececd005133ee417c0","unresolved":false,"context_lines":[{"line_number":65,"context_line":"through Ceilometer to expose metrics through the existing storage mechanisms."},{"line_number":66,"context_line":"If there are issues in configuration or processing backlogs or excessive error"},{"line_number":67,"context_line":"rates for any reason than those metrics would either not be available or not be"},{"line_number":68,"context_line":"correct."},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"Ceilometer could generate several of the measures above and deliver them"},{"line_number":71,"context_line":"directly to Gnocchi through the StatsD protocol but that requires the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_7ba427b9","line":68,"updated":"2016-08-23 15:01:42.000000000","message":"i think i like this idea the best. i don\u0027t mind adding a v3 api which strictly enforces Ceilometer as a data processing/generating tool but i\u0027d very much prefer we avoid all storage/analytic concepts from Ceilometer. we\u0027ve had a lot of trouble weening people off Ceilometer storage already and adding more storage concepts will mislead users to believing we still support Ceilometer storage.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"70f5816da48437eabf964c0cd9b6408e4e032504","unresolved":false,"context_lines":[{"line_number":65,"context_line":"through Ceilometer to expose metrics through the existing storage mechanisms."},{"line_number":66,"context_line":"If there are issues in configuration or processing backlogs or excessive error"},{"line_number":67,"context_line":"rates for any reason than those metrics would either not be available or not be"},{"line_number":68,"context_line":"correct."},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"Ceilometer could generate several of the measures above and deliver them"},{"line_number":71,"context_line":"directly to Gnocchi through the StatsD protocol but that requires the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_77c0617d","line":68,"in_reply_to":"1ac06dbe_5411e35c","updated":"2016-08-24 18:18:23.000000000","message":"That was the initial idea I was working from.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":19594,"name":"Yarko Tymciurak","email":"yarko.tymciurak@intel.com","username":"yarkot"},"change_message_id":"3e7aafdbbabd61b87f9e93a52571a714b6af1991","unresolved":false,"context_lines":[{"line_number":65,"context_line":"through Ceilometer to expose metrics through the existing storage mechanisms."},{"line_number":66,"context_line":"If there are issues in configuration or processing backlogs or excessive error"},{"line_number":67,"context_line":"rates for any reason than those metrics would either not be available or not be"},{"line_number":68,"context_line":"correct."},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"Ceilometer could generate several of the measures above and deliver them"},{"line_number":71,"context_line":"directly to Gnocchi through the StatsD protocol but that requires the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_efa17195","line":68,"in_reply_to":"1ac06dbe_7ba427b9","updated":"2016-08-23 15:44:47.000000000","message":"I\u0027d go one further to gordon\u0027s comment(s):\n- to move ceilometer to a component based system, each component should provide some monitoring, e.g. \"heartbeat\" and configuration info;\n- that would be it\u0027s own subsystem (and should be pretty light-weight), along the lines of gnocchi, aodh, panko;\n- I would even go so far as to say this itself should be its own \"heartbeat\" bus and storage system (otherwise, how would you monitor issues with, say, gnocchi storage? or one of the message buses?);\n\nThat suggests that this should not be a middleware, and more a monitoring system, with some very minimal requirements placed on components (i.e. pollsters) to indicate their presence (and \"heartbeat\" might be just the right concept here).\n\nThinking of the gnocchi support in Grafana, this may be a useful target for this health monitoring component also.","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"b38a6a45b40038cdac142a1341843553ae18783f","unresolved":false,"context_lines":[{"line_number":65,"context_line":"through Ceilometer to expose metrics through the existing storage mechanisms."},{"line_number":66,"context_line":"If there are issues in configuration or processing backlogs or excessive error"},{"line_number":67,"context_line":"rates for any reason than those metrics would either not be available or not be"},{"line_number":68,"context_line":"correct."},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"Ceilometer could generate several of the measures above and deliver them"},{"line_number":71,"context_line":"directly to Gnocchi through the StatsD protocol but that requires the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_5411e35c","line":68,"in_reply_to":"1ac06dbe_dcbedfac","updated":"2016-08-24 18:14:16.000000000","message":"hmm.. so say we had 2 notifications agents up, with 5 workers each. each agent will store metrics in cache/redis? and API queries redis for data?","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"d894012e3984f93ff48475e7a5b0ad9ef2707331","unresolved":false,"context_lines":[{"line_number":65,"context_line":"through Ceilometer to expose metrics through the existing storage mechanisms."},{"line_number":66,"context_line":"If there are issues in configuration or processing backlogs or excessive error"},{"line_number":67,"context_line":"rates for any reason than those metrics would either not be available or not be"},{"line_number":68,"context_line":"correct."},{"line_number":69,"context_line":""},{"line_number":70,"context_line":"Ceilometer could generate several of the measures above and deliver them"},{"line_number":71,"context_line":"directly to Gnocchi through the StatsD protocol but that requires the"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_dcbedfac","line":68,"in_reply_to":"1ac06dbe_efa17195","updated":"2016-08-23 19:40:01.000000000","message":"@gordon and @yarko  \n\nTo be clear about what was intended in this draft:\nThe data to be emitted from components is effectively cached, not stored. Previous data is overwritten. Maybe this is still too close to \"storage\" but I\u0027m afraid I don\u0027t know what can be done about that.\n\nIf the CLI is invoked the latest data emitted can be drawn from the collective cache and organized for display. There is no history or visualization implied so I don\u0027t see analytics implied. I wanted to suggest the CLI is only run when it is being looked at but may have confused this point in lines 48-52.\n\nA monitoring system would be an active running thing. As proposed this is not creating an active running thing, yet. A monitor suggests a greater impact on the running system and a different model (persistent storage, or memory? how do you get the data back out of the monitor? API?). \n\nI\u0027m not strongly opposed to this model if that is preferred but it is larger in scope carry additional complexity which give me pause. Similarly, I am not opposed to expanding the scope over time to use the same pattern to include other Telemetry project services, possibly from the same tool.\n\nThe point of the proposal is to provide tools to spot and diagnose issues in telemetry. I\u0027m skeptical of trying to use the Telemetry project\u0027s tooling to instrument the telemetry project\u0027s components. As I noted above operators will not be able to clearly identify issues in the telemetry system that way.\n\nDoes any of this clarification give us a way forward?","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"},{"author":{"_account_id":12807,"name":"Steve Lewis (stevelle)","email":"stevelle@gmail.com","username":"stevelle"},"change_message_id":"dec89fd50eee858a6c9acaebfadc4048b1e39f8a","unresolved":false,"context_lines":[{"line_number":126,"context_line":""},{"line_number":127,"context_line":"TBD but expected to be minimal increase in cpu and memory footprint of"},{"line_number":128,"context_line":"Ceilometer component processes and a minimal increase in coordination traffic"},{"line_number":129,"context_line":"as each component process reports in periodically."},{"line_number":130,"context_line":""},{"line_number":131,"context_line":"Other deployer impact"},{"line_number":132,"context_line":"---------------------"}],"source_content_type":"text/x-rst","patch_set":1,"id":"1ac06dbe_ad5f1371","line":129,"updated":"2016-08-20 05:05:46.000000000","message":"additional load on message bus(es) is implied by polling for rates and queue sizes","commit_id":"1d290a7fa62c52c553675e1dd25afe1e8ee6fa79"}]}