)]}'
{"specs/juno/lbaas_metering.rst":[{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":18,"context_line":"Ceilometer currently has no support for metering Network Services. Cloud"},{"line_number":19,"context_line":"providers/Operators have the need to monitor and meter various aspects of"},{"line_number":20,"context_line":"the Network Services. This spec deals with metering Load Balancer as a"},{"line_number":21,"context_line":"service(LBaas)."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Proposed change"},{"line_number":24,"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":1,"id":"7adec928_ac77649e","line":21,"updated":"2014-05-22 18:14:52.000000000","message":"Nitpick (feel free to ignore):\n\ns/service(LBaas)/service (LBaaS)/","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":18,"context_line":"Ceilometer currently has no support for metering Network Services. Cloud"},{"line_number":19,"context_line":"providers/Operators have the need to monitor and meter various aspects of"},{"line_number":20,"context_line":"the Network Services. This spec deals with metering Load Balancer as a"},{"line_number":21,"context_line":"service(LBaas)."},{"line_number":22,"context_line":""},{"line_number":23,"context_line":"Proposed change"},{"line_number":24,"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":1,"id":"7adec928_e72345a5","line":21,"in_reply_to":"7adec928_ac77649e","updated":"2014-05-22 18:39:06.000000000","message":"sure, will fix","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":31,"context_line":""},{"line_number":32,"context_line":"    * Type of LB ( HAproxy, VPx etc..)"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":37,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_8c4d4875","line":34,"updated":"2014-05-22 18:14:52.000000000","message":"Where is the hourly aggregation actually done? Does this imply the ceilometer pollsters are on a one-hour cadence?","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":31,"context_line":""},{"line_number":32,"context_line":"    * Type of LB ( HAproxy, VPx etc..)"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":""},{"line_number":36,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":37,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_2c4fd4bf","line":34,"in_reply_to":"7adec928_8c4d4875","updated":"2014-05-22 18:39:06.000000000","message":"Was thinking based on the status of the pools(when ACTIVE), we could compute the usage time in some transformer.\n\nBut i’m not sure about the value of having this in ceilometer itself vs  let the operator gather sats from samples and compute themselves.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":49,"context_line":"    * Health monitors"},{"line_number":50,"context_line":"        * Status"},{"line_number":51,"context_line":"        * Response Time"},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"Alternatives"},{"line_number":54,"context_line":"------------"},{"line_number":55,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_cc0f50e0","line":52,"updated":"2014-05-22 18:14:52.000000000","message":"How are these health monitors to be represented in the data persisted to ceilometer?\n\nIs this per-endpoint status and response time? Or aggregated response times over an entire VIP? I\u0027m just thinking of how fine-grained the meters would end up being.\n\nSay we have a VIP in front of service scaled out to N instances - how many health monitor meters would that map onto?\n\nHow are the individual health monitors identified? buy UUID?","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":49,"context_line":"    * Health monitors"},{"line_number":50,"context_line":"        * Status"},{"line_number":51,"context_line":"        * Response Time"},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"Alternatives"},{"line_number":54,"context_line":"------------"},{"line_number":55,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_cc3bd0e9","line":52,"in_reply_to":"7adec928_cc0f50e0","updated":"2014-05-22 18:39:06.000000000","message":"when a health monitor is created, it has its own id and metadata info. Each probe could be tied to multiple pools. So if we have 10 probes tied to a vip we\u0027ll have 10 meters. If we have multiple pools using the same probe, we should have that in the pool metadata. So we should only have as many meters as probes. The status and response time are for the probes themselves. So we have a ping probe for example.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":74,"context_line":"---------------"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":77,"context_line":"number of connections and usage. Initially the plan is for the"},{"line_number":78,"context_line":"catch all to handle it. But once we have the new transformers in place,"},{"line_number":79,"context_line":"we\u0027ll need to include the network services metrics explicitly in pipeline.yaml."},{"line_number":80,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_0c13f85c","line":77,"updated":"2014-05-22 18:14:52.000000000","message":"I\u0027m not clear on what kind of transformations might be useful here. Can you elaborate?","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":74,"context_line":"---------------"},{"line_number":75,"context_line":""},{"line_number":76,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":77,"context_line":"number of connections and usage. Initially the plan is for the"},{"line_number":78,"context_line":"catch all to handle it. But once we have the new transformers in place,"},{"line_number":79,"context_line":"we\u0027ll need to include the network services metrics explicitly in pipeline.yaml."},{"line_number":80,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_cc12f061","line":77,"in_reply_to":"7adec928_0c13f85c","updated":"2014-05-22 18:39:06.000000000","message":"The transformations as i mentioned above would be for say usage. We pool the provider info periodically, based on the status of how long the pool is ACTIVE, we compute the usage.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":75,"context_line":""},{"line_number":76,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":77,"context_line":"number of connections and usage. Initially the plan is for the"},{"line_number":78,"context_line":"catch all to handle it. But once we have the new transformers in place,"},{"line_number":79,"context_line":"we\u0027ll need to include the network services metrics explicitly in pipeline.yaml."},{"line_number":80,"context_line":""},{"line_number":81,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_b11b3f7f","line":78,"updated":"2014-05-22 18:14:52.000000000","message":"Nitpick: might be clearer as:\n\n  s/catch all/catch-all source \u0026 sink/","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":75,"context_line":""},{"line_number":76,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":77,"context_line":"number of connections and usage. Initially the plan is for the"},{"line_number":78,"context_line":"catch all to handle it. But once we have the new transformers in place,"},{"line_number":79,"context_line":"we\u0027ll need to include the network services metrics explicitly in pipeline.yaml."},{"line_number":80,"context_line":""},{"line_number":81,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_67181576","line":78,"in_reply_to":"7adec928_b11b3f7f","updated":"2014-05-22 18:39:06.000000000","message":"will fix.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":125,"context_line":"Work Items"},{"line_number":126,"context_line":"----------"},{"line_number":127,"context_line":""},{"line_number":128,"context_line":"* Add neutron client APIs to query LBaas calls in neutron_client.py"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"* Add new Pollsters and notification handlers"},{"line_number":131,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_4c1a6018","line":128,"updated":"2014-05-22 18:14:52.000000000","message":"Will there be a separate neutron spec for this?","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":125,"context_line":"Work Items"},{"line_number":126,"context_line":"----------"},{"line_number":127,"context_line":""},{"line_number":128,"context_line":"* Add neutron client APIs to query LBaas calls in neutron_client.py"},{"line_number":129,"context_line":""},{"line_number":130,"context_line":"* Add new Pollsters and notification handlers"},{"line_number":131,"context_line":""}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_87ffe100","line":128,"updated":"2014-05-22 18:39:06.000000000","message":"most of these calls are already available on neutron side. The neutron_client here is on ceilometer side. We dont need a spec for this.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"3c7419d685b06d34e740d063c46912204675ad3b","unresolved":false,"context_lines":[{"line_number":152,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":153,"context_line":""},{"line_number":154,"context_line":"Unit Tests will be added to cover the necessary neutron_client calls, pollsters"},{"line_number":155,"context_line":"and notifications."},{"line_number":156,"context_line":""},{"line_number":157,"context_line":""},{"line_number":158,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_2ce27401","line":155,"updated":"2014-05-22 18:14:52.000000000","message":"What about integration testing against neutron?","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"a5c6704af5407d63b6b5fe94ece7ef02077047ab","unresolved":false,"context_lines":[{"line_number":152,"context_line":"\u003d\u003d\u003d\u003d\u003d\u003d\u003d"},{"line_number":153,"context_line":""},{"line_number":154,"context_line":"Unit Tests will be added to cover the necessary neutron_client calls, pollsters"},{"line_number":155,"context_line":"and notifications."},{"line_number":156,"context_line":""},{"line_number":157,"context_line":""},{"line_number":158,"context_line":"Documentation Impact"}],"source_content_type":"text/x-rst","patch_set":1,"id":"7adec928_471dd965","line":155,"in_reply_to":"7adec928_2ce27401","updated":"2014-05-22 18:39:06.000000000","message":"sure i can include them.","commit_id":"5f542aec5f94081b06b2cbb58c6f84f452b45a24"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"e757606f2c54f91bee3ec62059b55c9315efb3ea","unresolved":false,"context_lines":[{"line_number":32,"context_line":"    * Type of LB ( HAproxy, VPx etc..)"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"}],"source_content_type":"text/x-rst","patch_set":3,"id":"5adbc538_1ec18e80","line":35,"updated":"2014-05-26 15:13:58.000000000","message":"does this mean we need to store an hour worth of samples to correctly build usage meter?","commit_id":"2e9888ea60cc63e48a127cbc9c2cecc09475ea41"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"b2b26514aa3d6f4c587276a1fe6475e5aef9c720","unresolved":false,"context_lines":[{"line_number":32,"context_line":"    * Type of LB ( HAproxy, VPx etc..)"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"}],"source_content_type":"text/x-rst","patch_set":3,"id":"3ae8d1ca_57eea4ed","line":35,"in_reply_to":"5adbc538_1ec18e80","updated":"2014-05-27 13:54:41.000000000","message":"that or we poll every hour and compare the status.","commit_id":"2e9888ea60cc63e48a127cbc9c2cecc09475ea41"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"e757606f2c54f91bee3ec62059b55c9315efb3ea","unresolved":false,"context_lines":[{"line_number":134,"context_line":""},{"line_number":135,"context_line":"* Add neutron client APIs to query LBaas calls in neutron_client.py"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"* Add new Pollsters and notification handlers"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"* Add Unit/Integration test coverage"},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"5adbc538_a31de375","line":137,"updated":"2014-05-26 15:13:58.000000000","message":"could you add proposed metrics names, type, unit for each new metric.. either here or above.","commit_id":"2e9888ea60cc63e48a127cbc9c2cecc09475ea41"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"b2b26514aa3d6f4c587276a1fe6475e5aef9c720","unresolved":false,"context_lines":[{"line_number":134,"context_line":""},{"line_number":135,"context_line":"* Add neutron client APIs to query LBaas calls in neutron_client.py"},{"line_number":136,"context_line":""},{"line_number":137,"context_line":"* Add new Pollsters and notification handlers"},{"line_number":138,"context_line":""},{"line_number":139,"context_line":"* Add Unit/Integration test coverage"},{"line_number":140,"context_line":""}],"source_content_type":"text/x-rst","patch_set":3,"id":"3ae8d1ca_02acdcec","line":137,"in_reply_to":"5adbc538_a31de375","updated":"2014-05-27 13:54:41.000000000","message":"Done","commit_id":"2e9888ea60cc63e48a127cbc9c2cecc09475ea41"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"ba2574b10f4238d567a9a68eb152108f0c6f2744","unresolved":false,"context_lines":[{"line_number":32,"context_line":"    * Type of LB ( HAproxy, VPx etc..)"},{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_651651f3","line":35,"updated":"2014-05-28 20:21:47.000000000","message":"I\u0027d be a bit concerned here if the \"period of time\" was longer than a few minutes.\n\nRemember that the \"pooling\" would be done by the transformer in volatile memory.\n\nWhat happens to data that hasn\u0027t been pushed yet when the agent is restarted?","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"ba2574b10f4238d567a9a68eb152108f0c6f2744","unresolved":false,"context_lines":[{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_659bb172","line":36,"updated":"2014-05-28 20:21:47.000000000","message":"Would the transformer also be responsible for computing the \"usage time\" based on the status being active?\n\nHow would that work in practice WRT to the kind of polling interval needed to accurately track changes in connection state?\n\nWould that level of fine-grained accounting be better achieved on the LBaaS side before being emitted as a notification?","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c6a8261b536894ee1149da6191cc4a1337f3dc63","unresolved":false,"context_lines":[{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_9e49fcd2","line":36,"updated":"2014-05-28 13:41:33.000000000","message":"i\u0027d be interested to see this implementation... as long as we\u0027re not storing more than a handful of samples in memory i\u0027m ok with this implementation","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8669d4d57def7159dbb17da734b76a5672f375c4","unresolved":false,"context_lines":[{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_09ff67ab","line":36,"in_reply_to":"3ae8d1ca_659bb172","updated":"2014-05-28 21:10:58.000000000","message":"hmm perhaps, I don\u0027t have a strong preference either or. So you\u0027re suggesting we should drop this metric and let the user calculate it outside of ceilometer. But LBaas doesn\u0027t provide us this today, it .as to be done either via status or create/delete time/","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8ccbd337d707fbe4da730ac1b19c3031b14bab0a","unresolved":false,"context_lines":[{"line_number":33,"context_line":""},{"line_number":34,"context_line":"    * Usage in hourly increments"},{"line_number":35,"context_line":"      - Add a transformer to pool the samples over a period of time"},{"line_number":36,"context_line":"      - Compute usage time based on status being ACTIVE"},{"line_number":37,"context_line":""},{"line_number":38,"context_line":"    * Features (SSL, Non-SSL, stickiness etc..)"},{"line_number":39,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_3e945ee1","line":36,"in_reply_to":"3ae8d1ca_9e49fcd2","updated":"2014-05-28 18:02:53.000000000","message":"agreed","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"ba2574b10f4238d567a9a68eb152108f0c6f2744","unresolved":false,"context_lines":[{"line_number":53,"context_line":""},{"line_number":54,"context_line":"    * Health monitors"},{"line_number":55,"context_line":"        * Status of the health probe"},{"line_number":56,"context_line":"        * Response Time of the health probe"},{"line_number":57,"context_line":""},{"line_number":58,"context_line":"* Metric Definitions:"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_a0784790","line":56,"updated":"2014-05-28 20:21:47.000000000","message":"I\u0027m still a bit concerned about this notion of response time.\n\nReturning to your response to my comments on:\n\nhttps://review.openstack.org/#/c/94958/1/specs/juno/lbaas_metering.rst\n\n\"when a health monitor is created, it has its own id and metadata info. Each probe could be tied to multiple pools. So if we have 10 probes tied to a vip we\u0027ll have 10 meters. If we have multiple pools using the same probe, we should have that in the pool metadata. So we should only have as many meters as probes. The status and response time are for the probes themselves. So we have a ping probe for example.\"\n\nIsn\u0027t the ping probe simply an implementation detail within LBaaS?\n\nThe fact the there are 10 probes tied to a single VIP, as opposed to 11 or 9, doesn\u0027t feel like something that ceilometer should be interested in.\n\nWhat useful information would it yield to surface the fact the responses measured by one probe differ from another, within the scope of a single VIP?\n\nThis is something that feels like it should be aggregated per VIP *before* being reported up to ceilometer.\n\nUnless there is a meaningful user-visible distinction between the individual probes tied to a single VIP (?)","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8669d4d57def7159dbb17da734b76a5672f375c4","unresolved":false,"context_lines":[{"line_number":53,"context_line":""},{"line_number":54,"context_line":"    * Health monitors"},{"line_number":55,"context_line":"        * Status of the health probe"},{"line_number":56,"context_line":"        * Response Time of the health probe"},{"line_number":57,"context_line":""},{"line_number":58,"context_line":"* Metric Definitions:"},{"line_number":59,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_895c57ef","line":56,"in_reply_to":"3ae8d1ca_a0784790","updated":"2014-05-28 21:10:58.000000000","message":"You\u0027re right. I thought the probes also emitted some response time info but thats internal to LB and is not reported via api/notification. I just checked the lbaas code as well. i think we can drop this from health probe measure and just track the probe itself as a provider could bill based on what probes are being created and used.\n\nSo for response time, we could track the response time of the vip itself. But that has to happen from the ceilometer side by probing the vip. If you think that metric would be useful I can add that in and remove the health probe response time which doesn\u0027t add much value here.","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"c6a8261b536894ee1149da6191cc4a1337f3dc63","unresolved":false,"context_lines":[{"line_number":70,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":71,"context_line":"network.services.lb.bytes_out                d               B               p"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"* Status would be part of the metadata for each pool, vip, members and monitors. "},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"Alternatives"},{"line_number":76,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_bef9584f","line":73,"updated":"2014-05-28 13:41:33.000000000","message":"thinking about future, how does status get represented in v3 api (assuming we keep resource metadata static)","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6537,"name":"gordon chung","email":"gord@live.ca","username":"chungg"},"change_message_id":"5ece71453ed32a5a62444648802dfe779108ee97","unresolved":false,"context_lines":[{"line_number":70,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":71,"context_line":"network.services.lb.bytes_out                d               B               p"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"* Status would be part of the metadata for each pool, vip, members and monitors. "},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"Alternatives"},{"line_number":76,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_54d749ed","line":73,"in_reply_to":"3ae8d1ca_3efd3e3b","updated":"2014-05-28 18:52:42.000000000","message":"i feel like this falls into the same category as all the other \u0027Existence of ...\u0027 meters we have... not really metrics but more events\n\ni\u0027m ok with this falling into that group of \u0027meters\u0027.. we should probably have a bp too handle them before we implement v3 though.","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8ccbd337d707fbe4da730ac1b19c3031b14bab0a","unresolved":false,"context_lines":[{"line_number":70,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":71,"context_line":"network.services.lb.bytes_out                d               B               p"},{"line_number":72,"context_line":""},{"line_number":73,"context_line":"* Status would be part of the metadata for each pool, vip, members and monitors. "},{"line_number":74,"context_line":""},{"line_number":75,"context_line":"Alternatives"},{"line_number":76,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_3efd3e3b","line":73,"in_reply_to":"3ae8d1ca_bef9584f","updated":"2014-05-28 18:02:53.000000000","message":"Yea i\u0027m not sure how the info in metadata would be represented on v3 api. I\u0027m focusing the spec on what\u0027s currently doable. We can always amend it as v3 apis take shape.","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"ba2574b10f4238d567a9a68eb152108f0c6f2744","unresolved":false,"context_lines":[{"line_number":99,"context_line":"number of connections and usage. For example, we need to compute the usage"},{"line_number":100,"context_line":"based on the status of the pools being active. Initially the plan is for the"},{"line_number":101,"context_line":"catch-all source \u0026 sink to handle it. But once we have the new transformers "},{"line_number":102,"context_line":"in place, we will need to include the network services metrics explicitly in "},{"line_number":103,"context_line":"pipeline.yaml."},{"line_number":104,"context_line":""},{"line_number":105,"context_line":""}],"source_content_type":"text/x-rst","patch_set":5,"id":"3ae8d1ca_406303dc","line":102,"updated":"2014-05-28 20:21:47.000000000","message":"Nitpick (only fix if uploading another patchset anyway):\n\nTrailing whitespace throughout.","commit_id":"0004e590e210dfbbd79aafb7484323b677881419"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"4fb2b7634d5ab4b313f14d39dd506bb001839206","unresolved":false,"context_lines":[{"line_number":48,"context_line":"          - Use statistics API call from LBaaS side"},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"    * Health monitors"},{"line_number":51,"context_line":"        * Status of the health probe"},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"* Metric Definitions:"},{"line_number":54,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_3b6837f5","line":51,"updated":"2014-05-29 21:02:53.000000000","message":"Aren\u0027t there multiple probes per VIP?\n\nHow would these statuses be aggregated? Or reported per probe?","commit_id":"373c31d785b3c61b9cdf99d712bdd9e0dd5cdb90"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8434310f77d7e44dc48087bb38ab35e6bbc5af6a","unresolved":false,"context_lines":[{"line_number":48,"context_line":"          - Use statistics API call from LBaaS side"},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"    * Health monitors"},{"line_number":51,"context_line":"        * Status of the health probe"},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"* Metric Definitions:"},{"line_number":54,"context_line":""}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_de33c1f4","line":51,"in_reply_to":"1ae5cdf2_3b6837f5","updated":"2014-05-29 21:10:11.000000000","message":"well a health probe can exists on its own and has its own status.if the probe is tied to a vip, the vip metadata has the probe info. we want to track probes and vips and their conjunction in the metadata. Providers could bill based on # of probes being active as a resource would be consumed. So at the very least we need this. We can extend this to aggregate status of all probes per vip etc. But i dint get into that here.","commit_id":"373c31d785b3c61b9cdf99d712bdd9e0dd5cdb90"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"4fb2b7634d5ab4b313f14d39dd506bb001839206","unresolved":false,"context_lines":[{"line_number":89,"context_line":"Pipeline impact"},{"line_number":90,"context_line":"---------------"},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":93,"context_line":"number of connections and usage. For example, we need to compute the usage"},{"line_number":94,"context_line":"based on the status of the pools being active. Initially the plan is for the"},{"line_number":95,"context_line":"catch-all source \u0026 sink to handle it. But once we have the new transformers"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_1bc7fb13","line":92,"updated":"2014-05-29 21:02:53.000000000","message":"Do we still need this language about a custom transformer, now that the metrics requiring in-memory aggregation on the agent-side are all removed?","commit_id":"373c31d785b3c61b9cdf99d712bdd9e0dd5cdb90"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"8434310f77d7e44dc48087bb38ab35e6bbc5af6a","unresolved":false,"context_lines":[{"line_number":89,"context_line":"Pipeline impact"},{"line_number":90,"context_line":"---------------"},{"line_number":91,"context_line":""},{"line_number":92,"context_line":"We might have to deal with custom transformer to handle bandwidth,"},{"line_number":93,"context_line":"number of connections and usage. For example, we need to compute the usage"},{"line_number":94,"context_line":"based on the status of the pools being active. Initially the plan is for the"},{"line_number":95,"context_line":"catch-all source \u0026 sink to handle it. But once we have the new transformers"}],"source_content_type":"text/x-rst","patch_set":7,"id":"1ae5cdf2_1e74499b","line":92,"in_reply_to":"1ae5cdf2_1bc7fb13","updated":"2014-05-29 21:10:11.000000000","message":"agreed. Will drop this part as we dont need any aggregation now.","commit_id":"373c31d785b3c61b9cdf99d712bdd9e0dd5cdb90"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"06965fd96f09d71c7524d36a805e37554c4cfcee","unresolved":false,"context_lines":[{"line_number":48,"context_line":"          - Use statistics API call from LBaaS side"},{"line_number":49,"context_line":""},{"line_number":50,"context_line":"    * Health monitors"},{"line_number":51,"context_line":"        * Status of the health probe"},{"line_number":52,"context_line":""},{"line_number":53,"context_line":"* Metric Definitions:"},{"line_number":54,"context_line":""}],"source_content_type":"text/x-rst","patch_set":8,"id":"1ae5cdf2_8095ca88","line":51,"updated":"2014-06-03 22:04:54.000000000","message":"Status implies something binary or ternary like up/down, healthy/unhealthy/unknown.\n\nWhereas line 61 below suggests a count of the number of monitors.","commit_id":"249d79c459581da4de55a4e49dd3e7689da3646b"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"06965fd96f09d71c7524d36a805e37554c4cfcee","unresolved":false,"context_lines":[{"line_number":58,"context_line":"network.services.lb.pools                    g               pool            p"},{"line_number":59,"context_line":"network.services.lb.vips                     g               vip             p"},{"line_number":60,"context_line":"network.services.lb.members                  g               member          p"},{"line_number":61,"context_line":"network.services.lb.health_monitors          g               health          p"},{"line_number":62,"context_line":"network.services.lb.total_connections        d               connections     p"},{"line_number":63,"context_line":"network.services.lb.active_connections       d               connections     p"},{"line_number":64,"context_line":"network.services.lb.bytes_in                 d               B               p"}],"source_content_type":"text/x-rst","patch_set":8,"id":"1ae5cdf2_e0ebee10","line":61,"updated":"2014-06-03 22:04:54.000000000","message":"Is this a count of the number of monitors?\n\nThen the unit should probably be monitor, not health.","commit_id":"249d79c459581da4de55a4e49dd3e7689da3646b"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"06965fd96f09d71c7524d36a805e37554c4cfcee","unresolved":false,"context_lines":[{"line_number":62,"context_line":"network.services.lb.total_connections        d               connections     p"},{"line_number":63,"context_line":"network.services.lb.active_connections       d               connections     p"},{"line_number":64,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":65,"context_line":"network.services.lb.bytes_out                d               B               p"},{"line_number":66,"context_line":""},{"line_number":67,"context_line":"* Status would be part of the metadata for each pool, vip, members and monitors."},{"line_number":68,"context_line":""}],"source_content_type":"text/x-rst","patch_set":8,"id":"1ae5cdf2_e0d20e5b","line":65,"updated":"2014-06-03 22:04:54.000000000","message":"Needs a key? e.g.\n\n  g \u003d gauge, d \u003d delta, p \u003d pollster","commit_id":"249d79c459581da4de55a4e49dd3e7689da3646b"},{"author":{"_account_id":6676,"name":"zqfan","email":"aji.zqfan@gmail.com","username":"aji-zqfan"},"change_message_id":"353a9ac5fc0a19fcee386aeb6d22d1acc4c34640","unresolved":false,"context_lines":[{"line_number":67,"context_line":"g \u003d gauge, d \u003d delta, p \u003d pollster"},{"line_number":68,"context_line":""},{"line_number":69,"context_line":"* Status is captured in an enum-style value in the sample volume, as opposed to the resource metadata,"},{"line_number":70,"context_line":"for each pool, vip, members and monitor."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"Alternatives"},{"line_number":73,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":10,"id":"1ae5cdf2_d56785b5","line":70,"updated":"2014-06-04 04:17:51.000000000","message":"Sphinx complains: Bullet list ends without a blank line; unexpected unindent.","commit_id":"509c052c4f612d4846c41fbeeaf7ab5ed8619209"},{"author":{"_account_id":2284,"name":"Eoghan Glynn","email":"eglynn@redhat.com","username":"eglynn"},"change_message_id":"c3476887c7b55180fefa910113533ee970587eab","unresolved":false,"context_lines":[{"line_number":67,"context_line":"g \u003d gauge, d \u003d delta, p \u003d pollster"},{"line_number":68,"context_line":""},{"line_number":69,"context_line":"* Status is captured in an enum-style value in the sample volume, as opposed to"},{"line_number":70,"context_line":"the resource metadata, for each pool, vip, members and monitor."},{"line_number":71,"context_line":""},{"line_number":72,"context_line":"Alternatives"},{"line_number":73,"context_line":"------------"}],"source_content_type":"text/x-rst","patch_set":11,"id":"1ae5cdf2_11281385","line":70,"updated":"2014-06-04 11:50:11.000000000","message":"^^^ \u003cindent\u003e to line up \"the resource ...\" underneath \"Status is ...\" \n\nThis is needed to clear the gate failure.","commit_id":"3f24728fccc54d08cf624717eef379e870062e98"},{"author":{"_account_id":9412,"name":"Andres Buraschi","email":"andres.buraschi@intel.com","username":"aburaschi"},"change_message_id":"7a6de43d073131936efd77fc2bca30e1ac56c603","unresolved":false,"context_lines":[{"line_number":59,"context_line":"network.services.lb.vips                     g               vip             p"},{"line_number":60,"context_line":"network.services.lb.members                  g               member          p"},{"line_number":61,"context_line":"network.services.lb.health_monitors          g               monitor         p"},{"line_number":62,"context_line":"network.services.lb.total_connections        d               connections     p"},{"line_number":63,"context_line":"network.services.lb.active_connections       d               connections     p"},{"line_number":64,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":65,"context_line":"network.services.lb.bytes_out                d               B               p"}],"source_content_type":"text/x-rst","patch_set":12,"id":"1ae5cdf2_b2bd6fda","line":62,"updated":"2014-06-05 15:55:50.000000000","message":"Maybe the unit should be \"connection\". (Same for active connections, below)","commit_id":"8ea7066640aaf46afa8d368a5776708d7615d3f8"},{"author":{"_account_id":6924,"name":"Pradeep Kilambi","email":"pkilambi@redhat.com","username":"pkilambi"},"change_message_id":"582d39caff39edeb99cf45e241b30b73c834a2ae","unresolved":false,"context_lines":[{"line_number":59,"context_line":"network.services.lb.vips                     g               vip             p"},{"line_number":60,"context_line":"network.services.lb.members                  g               member          p"},{"line_number":61,"context_line":"network.services.lb.health_monitors          g               monitor         p"},{"line_number":62,"context_line":"network.services.lb.total_connections        d               connections     p"},{"line_number":63,"context_line":"network.services.lb.active_connections       d               connections     p"},{"line_number":64,"context_line":"network.services.lb.bytes_in                 d               B               p"},{"line_number":65,"context_line":"network.services.lb.bytes_out                d               B               p"}],"source_content_type":"text/x-rst","patch_set":12,"id":"1ae5cdf2_b263af40","line":62,"in_reply_to":"1ae5cdf2_b2bd6fda","updated":"2014-06-05 16:01:04.000000000","message":"agreed. Would be consistent with rest. please review https://review.openstack.org/98165","commit_id":"8ea7066640aaf46afa8d368a5776708d7615d3f8"}]}