* Avoid race condition in ILMHistorySotre (#53039)
* Avoid race condition in ILMHistorySotre
This change modifies ILMHistoryStore to always apply correct settings and mappings,
even if template is deleted and not yet recreated. This ensures that ILM history index
is correctly managed by ILM and also fixes flaky history tests that were prone to
triggenring this race.
This commit also refactors and simplifies ILM history tests.
Closes#50353 and #52853
* Review comment
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* fixed tests
* backport #53306
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* Smarter copying of the rest specs and tests (#52114)
This PR addresses the unnecessary copying of the rest specs and allows
for better semantics for which specs and tests are copied. By default
the rest specs will get copied if the project applies
`elasticsearch.standalone-rest-test` or `esplugin` and the project
has rest tests or you configure the custom extension `restResources`.
This PR also removes the need for dozens of places where the x-pack
specs were copied by supporting copying of the x-pack rest specs too.
The plugin/task introduced here can also copy the rest tests to the
local project through a similar configuration.
The new plugin/task allows a user to minimize the surface area of
which rest specs are copied. Per project can be configured to include
only a subset of the specs (or tests). Configuring a project to only
copy the specs when actually needed should help with build cache hit
rates since we can better define what is actually in use.
However, project level optimizations for build cache hit rates are
not included with this PR.
Also, with this PR you can no longer use the includePackaged flag on
integTest task.
The following items are included in this PR:
* new plugin: `elasticsearch.rest-resources`
* new tasks: CopyRestApiTask and CopyRestTestsTask - performs the copy
* new extension 'restResources'
```
restResources {
restApi {
includeCore 'foo' , 'bar' //will include the core specs that start with foo and bar
includeXpack 'baz' //will include x-pack specs that start with baz
}
restTests {
includeCore 'foo', 'bar' //will include the core tests that start with foo and bar
includeXpack 'baz' //will include the x-pack tests that start with baz
}
}
```
* Make FreezeStep retryable
This change marks `FreezeStep` as retryable and adds test to make sure we can really run it again.
* refactor tests
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
* Make DeleteStep retryable
This change marks `DeleteStep` as retryable and adds test to make sure we really can invoke it again.
* Fix unused import
* revert unneeded changes
* test reworked
We marked the `init` ILM step as retryable but our test used `waitUntil`
without an assert so we didn’t catch the fact that we were not actually
able to retry this step as our ILM state didn’t contain any information
about the policy execution (as we were in the process of initialising
it).
This commit manually sets the current step to `init` when we’re moving
the ilm policy into the ERROR step (this enables us to successfully
move to the error step and later retry the step)
* ShrunkenIndexCheckStep: Use correct logger
(cherry picked from commit f78d4b3d91345a2a8fc0f48b90dd66c9959bd7ff)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
* Allow forcemerge in the hot phase for ILM policies
This commit changes the `forcemerge` action to also be allowed in the `hot` phase for policies. The
forcemerge will occur after a rollover, and allows users to take advantage of higher disk speeds for
performing the force merge (on a separate node type, for example).
On caveat with this is that a `forcemerge` in the `hot` phase *MUST* be accompanied by a `rollover`
action. ILM validates policies to ensure this is the case.
Resolves#43165
* Use anyMatch instead of findAny in validation
* Make randomTimeseriesLifecyclePolicy single-pass
* Adding best_compression (#49974)
This commit adds a `codec` parameter to the ILM `forcemerge` action. When setting the codec to `best_compression` ILM will close the index, then update the codec setting, re-open the index, and finally perform a force merge.
* Fix ForceMergeAction toSteps construction (#51825)
There was a duplicate force merge step and the test continued to fail. This commit clarifies the
`toStep` method and changes the `assertBestCompression` method for better readability.
Resolves#51822
* Update version constants
Co-authored-by: Sivagurunathan Velayutham <sivadeva.93@gmail.com>
We suspect the flakiness could’ve come from the fact that the rollover
step used to create the new index and roll the write alias to the new
index in separate cluster state updates. So the assertion that the
rolled index exists could’ve passed in the test but, before the
alias was rolled over to the new index, the subsequent write we execute
in the test (namely
`indexDocs("test_user", "x-pack-test-password", "foo_alias", 1)`)
would’ve sent the new document to the source index (ie. foo-logs-000001)
This would see the source index containing 3 documents and the rolled
index (foo-logs-000002) 0 documents.
However, we fixed this and the rollover step executes the “create index
and roll alias” in one single cluster update, so this situation should
not occur anymore.
(cherry picked from commit 834261c4fe7dd93f437eeec43c00d01ff2279f86)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
* Fix SnapshotLifecycleRestIT.testFullPolicySnapshot
This previously was missing some key information in the output of the failure. This captures that
information and adds logging at each step so we can determine the cause *if* it fails again.
Resolves#50358
Currently when an ILM policy finishes its execution, the index moves into the `TerminalPolicyStep`,
denoted by a completed/completed/completed phase/action/step lifecycle execution state.
This commit changes the behavior so that the index lifecycle execution state halts at the last
configured phase's `PhaseCompleteStep`, so for instance, if an index were configured with a policy
containing a `hot` and `cold` phase, the index would stop at the `cold/complete/complete`
`PhaseCompleteStep`. This allows an ILM user to update the policy to add any later phases and have
indices configured to use that policy pick up execution at the newly added "later" phase. For
example, if a `delete` phase were added to the policy specified about, the index would then move
from `cold/complete/complete` into the `delete` phase.
Relates to #48431
* Fix TimeSeriesLifecycleActionsIT.testShrinkAction
Shrinking a 6 shard index to 3 shards can be quite time consuming and
assertBusy probes the conditions at exponentially growing intervals.
This separates the one assertion that was used for all the conditions
into multiple assertBusy statements and increases the timeout for waiting
for the shrink to complete.
* Allow more time for shrink to complete
This commit allows more time for the shrink operation to complete in
testRetryFailedShrinkAction (separating the assertBusy calls too) and
testMoveToRolloverStep.
* Shrink to no more than 2 shards in tests
(cherry picked from commit 5fe780148fa3536915d61475b087896a5b9ace82)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
* Check all snapshots in SnapshotLifecycleRestIT.testFullPolicy
Rather than check the first returned snapshot for a snapshot starting with `snap-` in
SnapshotLifecycleRestIT.testFullPolicy, this commit changes the test to find any snapshots starting
with `snap-`.
In the event that there are no snapshots (the failure case), this also exposes the full results map
so we can diagnose why a failure occurred.
Relates to #50358
* Use a more imperative style for checking
* Separate aliases used for tests in TimeSeriesLifecycleActionsIT
This is related to #51375 and hopes to help illuminate why some of those tests are failing. This
commit switches the aliases used in the test to use a random alias name every time (since there were
some complaints in the tests about aliases having more than one write index). With this we hope to
determine the actual cause of the failure in the test.
This also adds additional information to the exception returned when calling move-to-step with the
incorrect current step.
* Fix rest test
After we rollover the index we wait for the configured number of shards for the
rolled index to become active (based on the index.write.wait_for_active_shards
setting which might be present in a template, or otherwise in the default case,
for the primaries to become active).
This wait might be long due to disk watermarks being tripped, replicas not
being able to spring to life due to cluster nodes reconfiguration and others
and, the RolloverStep might not complete successfully due to this inherent
transient situation, albeit the rolled index having been created.
(cherry picked from commit 457a92fb4c68c55976cc3c3e2f00a053dd2eac70)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
These policies store statistics, but since stats updating is asynchronous, it's
possible for the update from one test to bleed into a separate one. This change
switches the tests to use separate policy ids so that their stats are tracked
independently. It also relaxes the checking constraint in one of the tests.
Hopefully this:
Resolves#48531Resolves#48017
It's possible that the index could return no settings and thus throw a
`NullPointerException`.
I wasn't able to reproduce the original issue, but this should guard
against in the future.
Resolves#50646
Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com>
This test failed a couple of different ways, related to timing, as well
as concurrent snapshots, and also naming.
This commit splits the giant `assertBusy` into separate parts so that we don't
perform ~5 different requests and tests in the same loop. It also gives each
test a unique repository so that no other test can accidentally re-use
snapshots.
Resolves#50358 (hopefully!)
* ILM action to wait for SLM policy execution (#50454)
This change add new ILM action to wait for SLM policy execution to ensure that index has snapshot before deletion.
Closes#45067
* Fix flaky TimeSeriesLifecycleActionsIT#testWaitForSnapshot test
This change adds some randomness and cleanup step to TimeSeriesLifecycleActionsIT#testWaitForSnapshot and testWaitForSnapshotSlmExecutedBefore tests in attempt to make them stable.
Reletes to #50781
* Formatting changes
* Longer timeout
* Fix Map.of in Java8
* Unused import removed
* Refresh cached phase policy definition if possible on new policy
There are some cases when updating a policy does not change the
structure in a significant way. In these cases, we can reread the
policy definition for any indices using the updated policy.
This commit adds this refreshing to the `TransportPutLifecycleAction`
to allow this. It allows us to do things like change the configuration
values for a particular step, even when on that step (for example,
changing the rollover criteria while on the `check-rollover-ready` step).
There are more cases where the phase definition can be reread that just
the ones checked here (for example, removing an action that has already
been passed), and those will be added in subsequent work.
Relates to #48431
This commits makes the "init" ILM step retryable. It also adds a test
where an index is created with a non-parsable index name and then fails.
Related to #48183
This makes the "update-rollover-lifecycle-date" step, which is part of the
rollover action, retryable. It also adds an integration test to check the
step is retried and it eventually succeeds.
(cherry picked from commit 5bf068522deb2b6cd2563bcf80f34fdbf459c9f2)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
* Add aditional logging for ILM history store tests (#50624)
These tests use the same index name, making it hard to read logs when
diagnosing the failures. Additionally more information about the current
state of the index could be retrieved when failing.
This changes these two things in the hope of capturing more data about
why this fails on some CI nodes but not others.
Relates to #50353
This adds support for retrying AsyncActionSteps by triggering the async
step after ILM was moved back on the failed step (the async step we'll
be attempting to run after the cluster state reflects ILM being moved
back on the failed step).
This also marks the RolloverStep as retryable and adds an integration
test where the RolloverStep is failing to execute as the rolled over
index already exists to test that the async action RolloverStep is
retried until the rolled over index is deleted.
(cherry picked from commit 8bee5f4cb58a1242cc2ef4bc0317dae6c8be49d3)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
* Add ILM histore store index (#50287)
* Add ILM histore store index
This commit adds an ILM history store that tracks the lifecycle
execution state as an index progresses through its ILM policy. ILM
history documents store output similar to what the ILM explain API
returns.
An example document with ALL fields (not all documents will have all
fields) would look like:
```json
{
"@timestamp": 1203012389,
"policy": "my-ilm-policy",
"index": "index-2019.1.1-000023",
"index_age":123120,
"success": true,
"state": {
"phase": "warm",
"action": "allocate",
"step": "ERROR",
"failed_step": "update-settings",
"is_auto-retryable_error": true,
"creation_date": 12389012039,
"phase_time": 12908389120,
"action_time": 1283901209,
"step_time": 123904107140,
"phase_definition": "{\"policy\":\"ilm-history-ilm-policy\",\"phase_definition\":{\"min_age\":\"0ms\",\"actions\":{\"rollover\":{\"max_size\":\"50gb\",\"max_age\":\"30d\"}}},\"version\":1,\"modified_date_in_millis\":1576517253463}",
"step_info": "{... etc step info here as json ...}"
},
"error_details": "java.lang.RuntimeException: etc\n\tcaused by:etc etc etc full stacktrace"
}
```
These documents go into the `ilm-history-1-00000N` index to provide an
audit trail of the operations ILM has performed.
This history storage is enabled by default but can be disabled by setting
`index.lifecycle.history_index_enabled` to `false.`
Resolves#49180
* Make ILMHistoryStore.putAsync truly async (#50403)
This moves the `putAsync` method in `ILMHistoryStore` never to block.
Previously due to the way that the `BulkProcessor` works, it was possible
for `BulkProcessor#add` to block executing a bulk request. This was bad
as we may be adding things to the history store in cluster state update
threads.
This also moves the index creation to be done prior to the bulk request
execution, rather than being checked every time an operation was added
to the queue. This lessens the chance of the index being created, then
deleted (by some external force), and then recreated via a bulk indexing
request.
Resolves#50353
This commit refactors the `IndexLifecycleRunner` to split out and
consolidate the number of methods that change state from within ILM. It
adds a new class `IndexLifecycleTransition` that contains a number of
static methods used to modify ILM's state. These methods all return new
cluster states rather than making changes themselves (they can be
thought of as helpers for modifying ILM state).
Rather than having multiple ways to move an index to a particular step
(like `moveClusterStateToStep`, `moveClusterStateToNextStep`,
`moveClusterStateToPreviouslyFailedStep`, etc (there are others)) this
now consolidates those into three with (hopefully) useful names:
- `moveClusterStateToStep`
- `moveClusterStateToErrorStep`
- `moveClusterStateToPreviouslyFailedStep`
In the move, I was also able to consolidate duplicate or redundant
arguments to these functions. Prior to this commit there were many calls
that provided duplicate information (both `IndexMetaData` and
`LifecycleExecutionState` for example) where the duplicate argument
could be derived from a previous argument with no problems.
With this split, `IndexLifecycleRunner` now contains the methods used to
actually run steps as well as the methods that kick off cluster state
updates for state transitions. `IndexLifecycleTransition` contains only
the helpers for constructing new states from given scenarios.
This also adds Javadocs to all methods in both `IndexLifecycleRunner`
and `IndexLifecycleTransition` (this accounts for almost all of the
increase in code lines for this commit). It also makes all methods be as
restrictive in visibility, to limit the scope of where they are used.
This refactoring is part of work towards capturing actions and
transitions that ILM makes, by consolidating and simplifying the places
we make state changes, it will make adding operation auditing easier.
The rollover action is now a retryable step (see #48256)
so ILM will keep retrying until it succeeds as opposed to stopping and
moving the execution in the ERROR step.
Fixes#49073
(cherry picked from commit 3ae90898121b43032ec8f3b50514d93a86e14d0f)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
# Conflicts:
# x-pack/plugin/ilm/qa/multi-node/src/test/java/org/elasticsearch/xpack/ilm/TimeSeriesLifecycleActionsIT.java
Backport of #48849. Update `.editorconfig` to make the Java settings the
default for all files, and then apply a 2-space indent to all `*.gradle`
files. Then reformat all the files.
When using the move-to-step API, we should reread the phase JSON from
the latest version of the ILM policy. This allows a user to move to the
same step while re-reading the policy's latest version. For example,
when changing rollover criteria.
While manually messing around with some other things I discovered that
we only reread the policy when using the retry API, not the move-to-step
API. This commit changes the move-to-step API to always read the latest
version of the policy.
Previously this step moved to the forcemerge step, however, if the
machine running the test was fast enough, it would execute the
forcemerge and move to the next step (`segment-count`) so the comparison
would fail. This commit changes the step to be a step that will never go
anywhere else, the terminal step.
Resolves#48761
* ILM Test asserts on the same ilm/_explain output
With the introduction of retryable steps subsequent ilm/_explain calls
can see the state of an ilm cycle move out of the error step. This test
made several assertions assuming that the cycle remains in the error
step so this commit changes the test to make one _explain call and have
all the asserts work on the same ilm state (so subsequent assumptions to
the cycle being in the error step are valid).
* Drop unused field in test.
(cherry picked from commit 44c74bb487151c886a08b27f32b13f7a72056997)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
This test used an index without an alias to simulate a failure in the
`check-rollover-ready` step. However, with #48256 that step
automatically retries, meaning that the index may not always be in
the ERROR step.
This commit changes the test to use a shrink action with an invalid
number of shards so that it stays in the ERROR step.
Resolves#48767
This adds the infrastructure to be able to retry the execution of retryable
steps and makes the `check-rollover-ready` retryable as an initial step to
make the rollover action more resilient to transient errors.
(cherry picked from commit 454020ac8acb147eae97acb4ccd6fb470d1e5f48)
Signed-off-by: Andrei Dan <andrei.dan@elastic.co>
The failures in these tests have been remarkably difficult to track
down, in part because they will not reproduce locally. This commit
unmutes the flaky tests and increases logging, as well as introducing
some additional logging, to attempt to pin down the failures.
The cron schedule "1 2 3 4 5 ?" will run every May 4 at 03:02:01, which
may result in unnecessary test failures once a year. This commit
switches out uses of that schedule in tests for one which will never
execute (because it specifies a day which doesn't exist, Feb. 31). Also
factors the schedule out to a constant to make the intent clearer.
When checking for the existence of a document in the ILM/CCR integration
tests, `assertDocumentExists` makes an HTTP request and checks the
response code. However, if the repsonse code is not successful, the call
will throw a `ResponseException`. `assertDocumentExists` is often called
inside an `assertBusy`, and wrapping the `ResponseException` in an
`AssertionError` will allow the `assertBusy` to retry.
In particular, this fixes an issue with `testCCRUnfollowDuringSnapshot`
where the index in question may still be closed when the document is
requested.
Ioannis Kakavas
David Kyle
Przemko Robakowski
Jake Landis
Andrei Dan
Lee Hinman
David Roberts
Nhat Nguyen
Tim Vernum
Christoph Büscher
Rory Hunter
Gordon Brown
Dimitrios Liappis