- 9月 09, 2020
- 9月 08, 2020
-
-
由 Kowshik Prakasam 创作于
-
由 Kowshik Prakasam 创作于
-
- 9月 07, 2020
-
-
由 Kowshik Prakasam 创作于
- 9月 05, 2020
-
-
由 Vikas Singh 创作于
* CNKAF-1087: Fix rack aware topic placement for cloud use case In cloud deployements the partition assigment is done by TenantPartitionAssignor class. The class relies on cluster metadata from the callback that is called by 'UpdateMetadataRequest' api. In this callback we have bug where rack information isn't propagated, hence we end up doing rack unaware assignment. This change fixes that by passign rack information in the callback. The fix is just one line in MetadataCache. Most of the change is adding integration tests to make sure TenantPartitionAssignor works as expected for cloud use case, for both rack aware and unaware case. The newly written rack aware test case fails w/o this change and passes with this change.
- 9月 04, 2020
-
-
由 Rajini Sivaram 创作于
Reviewers: Brian Byrne
-
由 Kowshik Prakasam 创作于
-
由 Kowshik Prakasam 创作于
-
Users of older brokers have reported that partition reassignment didn't work when performed against a topic partition which had the replica placement constraint set to the empty string. This has been fixed since KCORE-175.
- 9月 03, 2020
-
-
由 Kowshik Prakasam 创作于
-
由 Harini Rajendran 创作于
* Adding licenses,notices and licenses.html files generated by license finder under ce-metrics doc. Co-authored-by: Xiaodong Du <xdu@confluent.io>
-
由 Dhruvil Shah 创作于
With this change, we ensure that: - all tier fetch logged along with relevant details about the fetch - request is completed exceptionally and the exception is propagated back to the API layer to handle. Before this change, the purgatory would wait until a request timeout and would expire the fetch as the tier fetcher thread would have died silently. - the fetch exception metric is bumped as we'd expect
-
由 Steve Rodrigues 创作于
This property is set on Operator-managed clusters. Done as a Kafka property to avoid warnings about unknown properties and allow reporting by AdminClient.describeConfigs. * Add a test case in KafkaConfigTest for the confluent.operator.managed property.
-
由 Rohit Shekhar 创作于
* Durability Audit Manager, Materialization, Reporter and Event Generation. Co-authored-by: Soumyarka Mondal <smondal@confluent.io>
- 9月 02, 2020
-
-
由 aloknnikhil 创作于
* CNKAF-697 - High bandwidth throttles during cluster roll - The quotas are configured with a minimum and a maximum per producer/consumer per tenant per broker. - The maximum is universal and is applicable to both followers and leaders. However, only the leaders ever reach the maximum quota. - The minimum is exclusively followed by the followers for cases when the cluster needs to roll (i.e. follower becomes a leader). - During this transition, it's necessary to maintain the throughput before the maximum quota is applied. - Per-broker tenant quota is always greater than zero to avoid excessive throttling of requests received before cluster metadata or quota configs are refreshed. These expose the configuration allowing this to be tuned at broker startup. Parameters added: ```java confluent.quota.tenant.follower.broker.min.producer.rate confluent.quota.tenant.follower.broker.min.consumer.rate ``` Also, refactored quota configuration parameters into ConfluentConfigs and marked as internal. * Fix checkstyle issue * Addressing comments * CNKAF-902: Add metrics for SSL handshake failures + time - Max + avg (ns) time for a connection handshake - Cumulative count + rate of handshake failures * Add unit-tests for SSL handshake metrics * Timeout if handshakes are not drained
-
由 An Qi Wu 创作于
- 9月 01, 2020
-
-
由 Aishwarya Gune 创作于
* modify the error message in catch block for Exception
-
由 Justine Olshan 创作于
* I added a metric for controller loading time. * Changed positioning of timer and streamlined the test. * Fix typo in formatting
-
由 Alok Thatikunta 创作于
* Print tiertopic partition and timestamp * Update integration tests
-
由 Soumyarka Mondal 创作于
* Increased the free disk threshold to remove flakiness * Refactored DiskThrottleListenerManager into a trait to avoid singleton during tests