[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17401270#comment-17401270 ] Kai Huang commented on KAFKA-12713: --- [~ijuma] Thanks for the pointer. I've bumped up the [discussion thread|https://lists.apache.org/thread.html/r261915b64c819129bc6017adaa12e8f9a0feb74c24ba331f4a08f30c%40%3Cdev.kafka.apache.org%3E] again. > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Assignee: Kai Huang >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17400039#comment-17400039 ] Ismael Juma commented on KAFKA-12713: - I had missed that. Worth bumping that thread for feedback. > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Assignee: Kai Huang >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17400037#comment-17400037 ] Kai Huang commented on KAFKA-12713: --- [~ijuma] here is the [discussion thread|https://lists.apache.org/thread.html/r261915b64c819129bc6017adaa12e8f9a0feb74c24ba331f4a08f30c%40%3Cdev.kafka.apache.org%3E] that Ming created a while ago. > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Assignee: Kai Huang >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=1730#comment-1730 ] Ismael Juma commented on KAFKA-12713: - [~kaihuang] Thanks for the KIP. I suggest starting a discuss thread for KIP-736 and we can discuss it further over there. :) > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Assignee: Kai Huang >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17399940#comment-17399940 ] Kai Huang commented on KAFKA-12713: --- [~ijuma] I would like to follow up on this ticket, and continue the discussion. I replied to your [discussion|https://lists.apache.org/thread.html/r7f82dde9133bf9d3a8b688ca7ae02ad761c52e7b79212c9247b276c5%40%3Cdev.kafka.apache.org%3E] thread and illustrate how the fetch latency metric should work in [KIP-736|https://cwiki.apache.org/confluence/display/KAFKA/KIP-736%3A+Report+the+true+end+to+end+fetch+latency]. Could you please take a look and see if that clarifies your question? > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Assignee: Kai Huang >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17390068#comment-17390068 ] Ming Liu commented on KAFKA-12713: -- Certainly, [~kaihuang], please take the ownership of the KIP and the PR. > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17388324#comment-17388324 ] Kai Huang commented on KAFKA-12713: --- Twitter would be interested to pursue this work to debug latency issues. [~mingaliu], I wonder do you mind if I take over this JIRA ticket and [KIP-736|https://cwiki.apache.org/confluence/display/KAFKA/KIP-736%3A+Report+the+true+end+to+end+fetch+latency]? > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)
[jira] [Commented] (KAFKA-12713) Report "REAL" follower/consumer fetch latency
[ https://issues.apache.org/jira/browse/KAFKA-12713?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=17342778#comment-17342778 ] Ming Liu commented on KAFKA-12713: -- Here is PR [https://github.com/apache/kafka/pull/10674/files] for the KIP. Please comment. > Report "REAL" follower/consumer fetch latency > - > > Key: KAFKA-12713 > URL: https://issues.apache.org/jira/browse/KAFKA-12713 > Project: Kafka > Issue Type: Bug >Reporter: Ming Liu >Priority: Major > > The fetch latency is an important metrics to monitor for the cluster > performance. With ACK=ALL, the produce latency is affected primarily by > broker fetch latency. > However, currently the reported fetch latency didn't reflect the true fetch > latency because it sometimes need to stay in purgatory and wait for > replica.fetch.wait.max.ms when data is not available. This greatly affect the > real P50, P99 etc. > I like to propose a KIP to be able track the real fetch latency for both > broker follower and consumer. > -- This message was sent by Atlassian Jira (v8.3.4#803005)