Shard level indexing pressure improves the current Indexing Pressure framework which performs memory accounting at node level and rejects the requests. This takes a step further to have rejections based on the memory accounting at shard level along with other key performance factors like throughput and last successful requests.
**Key features**
- Granular tracking of indexing tasks performance, at every shard level, for each node role i.e. coordinator, primary and replica.
- Smarter rejections by discarding the requests intended only for problematic index or shard, while still allowing others to continue (fairness in rejection).
- Rejections thresholds governed by combination of configurable parameters (such as memory limits on node) and dynamic parameters (such as latency increase, throughput degradation).
- Node level and shard level indexing pressure statistics exposed through stats api.
- Integration of Indexing pressure stats with Plugins for for metric visibility and auto-tuning in future.
- Control knobs to tune to the key performance thresholds which control rejections, to address any specific requirement or issues.
- Control knobs to run the feature in shadow-mode or enforced-mode. In shadow-mode only internal rejection breakdown metrics will be published while no actual rejections will be performed.
The changes were divided into small manageable chunks as part of the following PRs against a feature branch.
- Add Shard Indexing Pressure Settings. #716
- Add Shard Indexing Pressure Tracker. #717
- Refactor IndexingPressure to allow extension. #718
- Add Shard Indexing Pressure Store #838
- Add Shard Indexing Pressure Memory Manager #945
- Add ShardIndexingPressure framework level construct and Stats #1015
- Add Indexing Pressure Service which acts as orchestrator for IP #1084
- Add plumbing logic for IndexingPressureService in Transport Actions. #1113
- Add shard indexing pressure metric/stats via rest end point. #1171
- Add shard indexing pressure integration tests. #1198
Signed-off-by: Saurabh Singh <sisurab@amazon.com>
Co-authored-by: Saurabh Singh <sisurab@amazon.com>
Co-authored-by: Rabi Panda <adnapibar@gmail.com>
* Support for Heap after GC stats
Signed-off-by: Andriy Redko <andriy.redko@aiven.io>
* Addressing code review comments
Signed-off-by: Andriy Redko <andriy.redko@aiven.io>
* Using the right version 2.0.0 (instead of 1.2.0) for the change
Signed-off-by: Andriy Redko <andriy.redko@aiven.io>
This commit adds test coverage for translog pruning setting to
MetadataCreateIndexServiceTests
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
* Rename translog pruing setting to CCR specific setting
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Rename to index.plugins.replication.translog.retention_lease.pruning.enabled as
index settings needs "index." as prefix
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Add deprecations to retention pruning controls
This commit adds deprecation flags to all added settings, variables, and methods
specific to ccr's retention lease pruning mechanism.
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
* Addressed CR comments
Signed-off-by: Sai Kumar <karanas@amazon.com>
* fix javadoc deprecation
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
* fix deprecation tag in TranslogDeletionPolicy
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
* Addressed test issue under translog tests
Signed-off-by: Sai Kumar <karanas@amazon.com>
Co-authored-by: Nicholas Walter Knize <nknize@apache.org>
The changes ensure that in the event of a partial zone failure, the surviving nodes in the minority zone don't get overloaded with shards, this is governed by a skewness limit.
Signed-off-by: Bukhtawar Khan <bukhtawa@amazon.com>
This will cause the leader stuck on IO during publication to step down and eventually trigger a leader election.
Issue Description
---
The publication of cluster state is time bound to 30s by a cluster.publish.timeout settings. If this time is reached before the new cluster state is committed, then the cluster state change is rejected and the leader considers itself to have failed. It stands down and starts trying to elect a new master.
There is a bug in leader that when it tries to publish the new cluster state it first tries acquire a lock to flush the new state under a mutex to disk. The same lock is used to cancel the publication on timeout. Below is the state of the timeout scheduler meant to cancel the publication. So essentially if the flushing of cluster state is stuck on IO, so will the cancellation of the publication since both of them share the same mutex. So leader will not step down and effectively block the cluster from making progress.
Signed-off-by: Bukhtawar Khan <bukhtawa@amazon.com>
* Drop mocksocket in favour of custom security manager checks (tests only)
Signed-off-by: Andriy Redko <andriy.redko@aiven.io>
* Slightly relaxed host checks to allow all local addresses
Signed-off-by: Andriy Redko <andriy.redko@aiven.io>
* Changes to support retrieval of operations from translog based on specified range
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Addressed CR comments
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Added testcases for internal engine
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Support for translog pruning based on retention leases
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Addressed CR Comments
Signed-off-by: Sai Kumar <karanas@amazon.com>
* Addressed test case issue
Signed-off-by: Sai Kumar <karanas@amazon.com>
The version framework only added support for OpenSearch 1.x bwc with legacy
clusters. This commit adds support for v2.0 which will be the last version with
bwc support for legacy clusters (v7.10)
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
This commit stages the branch to the next 1.0.1 patch release. BWC testing needs
this even if the next revision is never actually released.
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
PluginInfo should use .onOrAfter(Version.V_1_1_0) instead of
.after(Version.V_1_0_0) for the new custom folder name for plugin feature.
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
Lucene 9 removes support for SimpleFS File System format. This commit deprecates
the SimpleFS format in favor of NIOFS.
Signed-off-by: Nicholas Walter Knize <nknize@apache.org>
In some cases as one shared with issue #1099, the maxConcurrentSearchRequests was chosen as 0 which
will compute the final value during execution of the request based on processor counts. When this
computed value is less than number of search request in msearch request, it will execute all the
requests in multiple iterations causing the failure since test will only wait for one such
iteration. Hence setting the maxConcurrentSearchRequests explicitly to number of search requests
being added in the test to ensure correct behavior
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Part 1: Support for cancel_after_timeinterval parameter in search and msearch request
This commit introduces the new request level parameter to configure the timeout interval after which
a search request will be cancelled. For msearch request the parameter is supported both at parent
request and at sub child search requests. If it is provided at parent level and child search request
doesn't have it then the parent level value is set at such child request. The parent level msearch
is not used to cancel the parent request as it may be tricky to come up with correct value in cases
when child search request can have different runtimes
TEST: Added test for ser/de with new parameter
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Part 2: Support for cancel_after_timeinterval parameter in search and msearch request
This commit adds the handling of the new request level parameter and schedule cancellation task. It
also adds a cluster setting to set a global cancellation timeout for search request which will be
used in absence of request level timeout.
TEST: Added new tests in SearchCancellationIT
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Address Review feedback for Part 1
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Address review feedback for Part 2
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Update CancellableTask to remove the cancelOnTimeout boolean flag
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Replace search.cancellation.timeout cluster setting with search.enforce_server.timeout.cancellation to control if cluster level cancel_after_time_interval should take precedence over request level cancel_after_time_interval value
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
* Removing the search.enforce_server.timeout.cancellation cluster setting and just keeping search.cancel_after_time_interval setting with request level parameter taking the precedence.
Signed-off-by: Sorabh Hamirwasia <sohami.apache@gmail.com>
Co-authored-by: Sorabh Hamirwasia <hsorabh@amazon.com>
_cat/master is a fundamental API to know the master instance in the cluster. Given RestClusterState is exempted from tripping already, doesn't make sense for RestMasterAction to trip
Signed-off-by: Bukhtawar Khan bukhtawa@amazon.com
This change adds the initial version of a new CLI tool `opensearch-upgrade` as part of the OpenSearch distribution. This tool is meant for assisting during an upgrade from an existing Elasticsearch v7.10.2/v6.8.0 node to OpenSearch. It automates the process of importing existing configurations and installing of core plugins.
Signed-off-by: Rabi Panda <adnapibar@gmail.com>
* Lower build requirement from Java 14+ to Java 11+
Avoid use of -Werror -Xlint:all, which may change significantly across
java releases (new warnings could be added). Instead, just list the
warnings individually.
Workaround JDK 11 compiler bug (JDK-8209058) that only impacts test fixture
code in the build itself.
Signed-off-by: Robert Muir <rmuir@apache.org>
* Disable warning around -source 7 -release 7 for java version checker
The java version checker triggers some default warnings because it
targets java7:
```
> Task :distribution:tools:java-version-checker:compileJava FAILED
warning: [options] source value 7 is obsolete and will be removed in a future release
warning: [options] target value 7 is obsolete and will be removed in a future release
warning: [options] To suppress warnings about obsolete options, use -Xlint:-options.
error: warnings found and -Werror specified
```
Suppress this warning explicitly for this module.
Signed-off-by: Robert Muir <rmuir@apache.org>
* more java14 -> java11 cleanup
Signed-off-by: Robert Muir <rmuir@apache.org>
Co-authored-by: Robert Muir <rmuir@apache.org>
* Add "tagline" field back to "MainResponse" in sever side (not in rest-high-level-client side) that removed in PR #427 .
* Replace with a new tagline "The OpenSearch Project: https://opensearch.org/".
* Turn the tagline into a constant in server/src/main/java/org/opensearch/action/main/MainResponse.java.
This Change removes version.distribution when the version.number is
overridden with the cluster setting compatibility.override_main_response_version.
Signed-off-by: Marc Handalian <handalm@amazon.com>
This change adds a new cluster setting "compatibility.override_main_response_version"
that when enabled spoofs the version.number returned from MainResponse
for REST clients expecting legacy version 7.10.2.
Signed-off-by: Marc Handalian <handalm@amazon.com>
* Add Plugin name for verbose Plugin not found exception
* Make the plugin loading failure exception more verbose
* Throw Opensearch in place of RuntimeException for plugin load failure
* Nit fix, added ... to make logging standout
Signed-off-by: Jayesh Hathila <sharma.jayesh52@gmail.com>
* Address a kind of issue suggested by Amazon CodeGuru Reviewer:
* Add try-with-resources block to automatically close the resources after using to avoid resource leak, in `SymbolicLinkPreservingTarIT`, `LicenseAnalyzer`, `SymbolicLinkPreservingUntarTransform`, `ConcurrentSeqNoVersioningIT` in `VersionProperties`, `GeoFilterIT`, `XContentHelper`, `Json` and `IndexShard` class
* Add try-finally block to close the resources after using to avoid resource leak, in `ServerChannelContext` class.
* Add try-catch block to close the resources when exception occurs in `FsBlobContainer` class (when XContentFactory.xContentType throws an exception).
* Close resources when assertion error occurs, in `ServerChannelContext` class.
