David Turner

Results 52 issues of David Turner

Today the unit tests on the desired balance allocator feature branch still default to using the `BalancedShardsAllocator`. With this commit we randomly choose between `BalancedShardsAllocator` and `DesiredBalanceShardsAllocator`.

>test
:Distributed/Allocation
Team:Distributed

A user [reported running out of disk space](https://discuss.elastic.co/t/unable-to-delete-snapshot-no-space-left-on-device-solved/261675) in their shared filesystem repository which left it completely stuck, unable to take any further actions since everything that might delete any...

>enhancement
:Distributed/Snapshot/Restore
Team:Distributed

Also seen in a local run of `./gradlew :server:test` but doesn't easily reproduce, even when running many iterations in a loop. **Build scan:** https://gradle-enterprise.elastic.co/s/kbbpmnbim32ck/tests/:server:test/org.elasticsearch.index.engine.InternalEngineTests/testMergeThreadLogging **Reproduction line:** `./gradlew ':server:test' --tests "org.elasticsearch.index.engine.InternalEngineTests.testMergeThreadLogging"...

>test-failure
:Distributed/Engine
Team:Distributed

Today the master's pending task queue is just the `PriorityBlockingQueue` belonging to the underlying `ThreadPoolExecutor`. The reasons for this date back a long way but it doesn't really reflect the...

WIP
>refactoring
:Distributed/Cluster Coordination
Team:Distributed
>tech debt
v8.5.0

Snapshots work by writing to a blobstore in which the same blob can be accessed at the same path across all nodes. By default we check that the blobstore is...

>bug
:Distributed/Snapshot/Restore
Team:Distributed
Supportability

When `IndicesClusterStateService` applies a new cluster state it does substantial amounts of work inline on the applier thread, including some IO (or blocking waits for IO on other threads) which...

>bug
:Distributed/Recovery
Team:Distributed
>tech debt

Today `transport.port` and `http.port` accept a range of ports. At startup, Elasticsearch binds to the first available port in the range. This is useful for testing because it lets you...

:Distributed/Network
Team:Distributed
>tech debt

Today a node with a registered `URLRepository` will not shut down cleanly because it never releases the last of the `activityRefs`. This commit fixes that.

>bug
:Distributed/Snapshot/Restore
Team:Distributed
auto-backport-and-merge
v8.13.0
v8.12.2
v8.14.0

The get-snapshots APIs accumulate the details of all matching snapshots in memory before responding. In a high-index-count cluster with a long retention period on snapshots this can add up to...

>bug
:Distributed/Snapshot/Restore
Team:Distributed