Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

add distributed locking to jobs in alerting #1403

Merged
merged 1 commit into from
Mar 9, 2024

Conversation

sbcd90
Copy link
Collaborator

@sbcd90 sbcd90 commented Feb 3, 2024

Issue #, if available:

Description of changes:
Alerting uses scheduled jobs to trigger monitor runs. In case a monitor run is a long running job, the job may not get completed before the next job is scheduled to run. This introduces a scenario where multiple jobs run in parallel resulting a race condition to acquire resources.
This pr introduces distributed locking to scheduled jobs in alerting. It ensures only 1 active job can run at an instant of time.

CheckList:

  • Commits are signed per the DCO using --signoff

By submitting this pull request, I confirm that my contribution is made under the terms of the Apache 2.0 license.
For more information on following Developer Certificate of Origin and signing off your commits, please check here.

@sbcd90 sbcd90 marked this pull request as draft February 3, 2024 04:04
@sbcd90 sbcd90 force-pushed the dist_lock branch 4 times, most recently from 47a4888 to e2a52df Compare February 6, 2024 08:52
@sbcd90 sbcd90 force-pushed the dist_lock branch 2 times, most recently from aa8eef4 to 7c90068 Compare March 8, 2024 20:52
@sbcd90 sbcd90 marked this pull request as ready for review March 8, 2024 20:53
} ?: return@launch
logger.debug("lock ${lock!!.lockId} acquired")
logger.debug(
"PERF_DEBUG: executing workflow ${job.id} on node " +
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Minor: I think we should get rid of the PERF_DEBUG prefixes in these log statements. I'm fine if we fix it later though

Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

+1

monitorCtx.clusterService!!.state().nodes().localNode.id
)
runJob(job, periodStart, periodEnd, false)
var lock: LockModel? = null
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Minor: this code looks identical to 303-317. We should create a private method to avoid the duplication

override fun onResponse(existingLock: LockModel?) {
if (existingLock != null) {
if (isLockReleased(existingLock)) {
log.debug("lock is released or expired: {}", existingLock)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Non-blocking - this line is 10x nested. We should avoid deeply nesting code as it hurts readability. We can break much of this into private methods to avoid the nesting

try {
lock = monitorCtx.client!!.suspendUntil<Client, LockModel?> {
monitorCtx.lockService!!.acquireLock(job, it)
} ?: return@launch
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

We could consider adding adding some delay between retries on acquiring the lock

private val log = LogManager.getLogger(LockService::class.java)

class LockService(private val client: Client, private val clusterService: ClusterService) {
private var testInstant: Instant? = null
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

why is this called testInstant??

}

override fun onFailure(ex: Exception) {
log.error("Failed to update config index schema", ex)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

log error only in else condition

listener.onResponse(true)
} else {
val indexRequest = CreateIndexRequest(LOCK_INDEX_NAME).mapping(lockMapping())
.settings(Settings.builder().put("index.hidden", true).build())
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

plz add primary and replica settings based on our general practices for system index.

@@ -148,6 +154,10 @@ object DeleteMonitorService :
}
}

private suspend fun deleteLock(monitor: Monitor) {
client.suspendUntil<Client, Boolean> { lockService.deleteLock(LockModel.generateLockId(monitor.id), it) }
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

try catch?

)

override fun toXContent(builder: XContentBuilder, params: ToXContent.Params): XContentBuilder {
builder.startObject()
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lockid missing?

}

companion object {
const val SCHEDULED_JOB_ID = "scheduled_job_id"
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lockid missing?

) {
val lockId = LockModel.generateLockId(scheduledJobId)
createLockIndex(
object : ActionListener<Boolean> {
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

plz use actionlistener.wrap()

listener.onResponse(null)
}
} else {
listener.onResponse(null)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

what does null mean? and what does failure mean?

listener.onResponse(LockModel.parse(parser, response.seqNo, response.primaryTerm))
} catch (e: IOException) {
log.error("IOException occurred finding lock", e)
listener.onResponse(null)
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

onFailure?

@eirsep eirsep merged commit bcaa4e3 into opensearch-project:main Mar 9, 2024
7 of 18 checks passed
opensearch-trigger-bot bot pushed a commit that referenced this pull request Mar 9, 2024
Signed-off-by: Subhobrata Dey <[email protected]>
(cherry picked from commit bcaa4e3)
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
sbcd90 pushed a commit that referenced this pull request Mar 9, 2024
sbcd90 pushed a commit to sbcd90/alerting that referenced this pull request Mar 10, 2024
…ject#1307)

* Added 2.11.1 release notes.

* Added 2.11.1 release notes.

---------

(cherry picked from commit 06c1b8a)

Signed-off-by: AWSHurneyt <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

fix workflow security tests in alerting (opensearch-project#1310) (opensearch-project#1311)

Signed-off-by: Subhobrata Dey <[email protected]>

Increment version to 2.12.0-SNAPSHOT (opensearch-project#1239)

Signed-off-by: opensearch-ci-bot <[email protected]>
Co-authored-by: opensearch-ci-bot <[email protected]>

[Backport 2.x] Reference get monitor and search monitor action / request / responses from common-utils (opensearch-project#1315)

* Use get monitor action / req / resp from common-utils

Signed-off-by: Tyler Ohlsen <[email protected]>

* Dummy commit to retrigger

Signed-off-by: Tyler Ohlsen <[email protected]>

---------

Signed-off-by: Tyler Ohlsen <[email protected]>

optimize doc-level monitor execution workflow for datastreams (opensearch-project#1302) (opensearch-project#1322)

Signed-off-by: Subhobrata Dey <[email protected]>

Update to Gradle 8.5 (opensearch-project#1369) (opensearch-project#1371)

Signed-off-by: Andriy Redko <[email protected]>

[Backport 2.x] Inject namedWriteableRegistry during ser/deser of SearchMonitorAction (opensearch-project#1382) (opensearch-project#1384)

* Inject namedWriteableRegistry during ser/deser of SearchMonitorAction (opensearch-project#1382)

Signed-off-by: Tyler Ohlsen <[email protected]>

* remove bin files

Signed-off-by: Tyler Ohlsen <[email protected]>

* remove core bin

Signed-off-by: Tyler Ohlsen <[email protected]>

---------

Signed-off-by: Tyler Ohlsen <[email protected]>

Don't attempt to parse workflow if it doesn't exist (opensearch-project#1346) (opensearch-project#1359)

(cherry picked from commit 733fd4e)

Signed-off-by: Chase Engelbrecht <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

Set docData to empty string if actual is null (opensearch-project#1325) (opensearch-project#1334)

(cherry picked from commit 008e076)

Signed-off-by: Chase Engelbrecht <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

removed default admin credentials for alerting (opensearch-project#1399) (opensearch-project#1400)

(cherry picked from commit 3c50f7d)

Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Dennis Toepker <[email protected]>

ipaddress lib upgrade as part of cve fix (opensearch-project#1397) (opensearch-project#1407)

(cherry picked from commit 8d59060)

Signed-off-by: Riya Saxena <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

Bulk index findings and sequentially invoke auto-correlations (opensearch-project#1355) (opensearch-project#1410)

* Bulk index findings and sequentially invoke auto-correlations

* Bulk index findings in batches of 10000 and make it configurable

* Addressing review comments

* Add integ tests to test bulk index findings

* Fix ktlint formatting

---------

(cherry picked from commit b561965)

Signed-off-by: Megha Goyal <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

Add 2.12 release notes (opensearch-project#1408) (opensearch-project#1413)

* Add 2.12 release notes

* Fix release notes PR

* Add 2 more PRs

---------

(cherry picked from commit b10eaad)

Signed-off-by: Chase Engelbrecht <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

[Backport 2.x] Implemented cross-cluster monitor support (opensearch-project#1404) (opensearch-project#1412)

* Implemented cross-cluster monitor support (opensearch-project#1404)

* Updated alert mappings to accommodate cross-cluster cluster metrics monitors.

Signed-off-by: AWSHurneyt <[email protected]>

* Implemented support for cross-cluster cluster metrics monitors. Implemented GetRemoteIndexes API to populate the frontend UI with details regarding the remote clusters, and indexes.

Signed-off-by: AWSHurneyt <[email protected]>

* Fixed a writeable test after changing QueryLevelTriggerRunResult from a data class to an open class for inheritability.

Signed-off-by: AWSHurneyt <[email protected]>

* Fixed ktlint errors.

Signed-off-by: AWSHurneyt <[email protected]>

* Removed changes to IndexUtils as they're only needed by doc monitors.

Signed-off-by: AWSHurneyt <[email protected]>

---------

Signed-off-by: AWSHurneyt <[email protected]>

(cherry picked from commit ea36996)
Signed-off-by: AWSHurneyt <[email protected]>

* Fixed a test.

Signed-off-by: AWSHurneyt <[email protected]>

---------

Signed-off-by: AWSHurneyt <[email protected]>

Add publishToMavenLocal in build.sh (opensearch-project#1418) (opensearch-project#1419)

(cherry picked from commit 4cdc1d1)
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

fix for MapperException[the [enabled] parameter can't be updated for the object mapping [metadata.source_to_query_index_mapping] (opensearch-project#1432) (opensearch-project#1434)

Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

bacport PRs opensearch-project#1445, opensearch-project#1430, opensearch-project#1441, opensearch-project#1435 to 2.x (opensearch-project#1452)

* Add jvm aware setting and max num docs settings for batching docs for percolate queries (opensearch-project#1435)

* add jvm aware and max docs settings for batching docs for percolate queries

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix stats logging

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add queryfieldnames field in findings mapping

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

* optimize to fetch only fields relevant to doc level queries in doc level monitor instead of entire _source for each doc (opensearch-project#1441)

* optimize to fetch only fields relevant to doc level queries in doc level monitor

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix test for settings check

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix ktlint

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

* clean up doc level queries on dry run (opensearch-project#1430)

Signed-off-by: Joanne Wang <[email protected]>

* optimize sequence number calculation and reduce search requests in doc level monitor execution (opensearch-project#1445)

* optimize sequence number calculation and reduce search requests by n where n is number of shards being queried in the executino

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix tests

Signed-off-by: Surya Sashank Nistala <[email protected]>

* optimize check indices and execute to query only write index of aliases and datastreams during monitor creation

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix test

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add javadoc

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add tests to verify seq_no calculation

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>
Signed-off-by: Joanne Wang <[email protected]>
Co-authored-by: Joanne Wang <[email protected]>

[Backport 2.x] Add an _exists_ check to document level monitor queries (opensearch-project#1425) (opensearch-project#1456)

* Add an _exists_ check to document level monitor queries (opensearch-project#1425)

* clean up and add integ tests

Signed-off-by: Joanne Wang <[email protected]>

* refactored out common method and renamed test

Signed-off-by: Joanne Wang <[email protected]>

* remove _exists_ flag

Signed-off-by: Joanne Wang <[email protected]>

---------

Signed-off-by: Joanne Wang <[email protected]>

* fix integ test

Signed-off-by: Joanne Wang <[email protected]>

---------

Signed-off-by: Joanne Wang <[email protected]>

add distributed locking to jobs in alerting (opensearch-project#1403) (opensearch-project#1458)

Signed-off-by: Subhobrata Dey <[email protected]>
eirsep pushed a commit to eirsep/alerting that referenced this pull request Mar 13, 2024
eirsep pushed a commit to eirsep/alerting that referenced this pull request Mar 14, 2024
eirsep added a commit that referenced this pull request Mar 14, 2024
…, #1441 to 2.9 (#1469)

* optimize doc-level monitor execution workflow for datastreams (#1302)

* optimize doc-level monitor execution for datastreams

Signed-off-by: Subhobrata Dey <[email protected]>

* add more tests to address comments

Signed-off-by: Subhobrata Dey <[email protected]>

* add integTest for multiple datastreams inside a single index pattern

* add integTest for multiple datastreams inside a single index pattern

Signed-off-by: Subhobrata Dey <[email protected]>

---------

Signed-off-by: Subhobrata Dey <[email protected]>

* Bulk index findings and sequentially invoke auto-correlations (#1355)

* Bulk index findings and sequentially invoke auto-correlations

Signed-off-by: Megha Goyal <[email protected]>

* Bulk index findings in batches of 10000 and make it configurable

Signed-off-by: Megha Goyal <[email protected]>

* Addressing review comments

Signed-off-by: Megha Goyal <[email protected]>

* Add integ tests to test bulk index findings

Signed-off-by: Megha Goyal <[email protected]>

* Fix ktlint formatting

Signed-off-by: Megha Goyal <[email protected]>

---------

Signed-off-by: Megha Goyal <[email protected]>

* fix for MapperException[the [enabled] parameter can't be updated for the object mapping [metadata.source_to_query_index_mapping] (#1432) (#1434)

Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>

* Add jvm aware setting and max num docs settings for batching docs for percolate queries (#1435)

* add jvm aware and max docs settings for batching docs for percolate queries

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix stats logging

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add queryfieldnames field in findings mapping

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

* clean up doc level queries on dry run (#1430)

Signed-off-by: Joanne Wang <[email protected]>

* optimize to fetch only fields relevant to doc level queries in doc level monitor instead of entire _source for each doc (#1441)

* optimize to fetch only fields relevant to doc level queries in doc level monitor

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix test for settings check

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix ktlint

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

* optimize sequence number calculation and reduce search requests in doc level monitor execution (#1445)

* optimize sequence number calculation and reduce search requests by n where n is number of shards being queried in the executino

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix tests

Signed-off-by: Surya Sashank Nistala <[email protected]>

* optimize check indices and execute to query only write index of aliases and datastreams during monitor creation

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix test

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add javadoc

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add tests to verify seq_no calculation

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Surya Sashank Nistala <[email protected]>

* add distributed locking to jobs in alerting (#1403)

Signed-off-by: Subhobrata Dey <[email protected]>

* 2.9.1 version bump

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix compilation issues

Signed-off-by: Surya Sashank Nistala <[email protected]>

* dummy commit

Signed-off-by: Surya Sashank Nistala <[email protected]>

* fix findings index schema version tsts

Signed-off-by: Surya Sashank Nistala <[email protected]>

---------

Signed-off-by: Subhobrata Dey <[email protected]>
Signed-off-by: Megha Goyal <[email protected]>
Signed-off-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Signed-off-by: Surya Sashank Nistala <[email protected]>
Signed-off-by: Joanne Wang <[email protected]>
Co-authored-by: Subhobrata Dey <[email protected]>
Co-authored-by: Megha Goyal <[email protected]>
Co-authored-by: opensearch-trigger-bot[bot] <98922864+opensearch-trigger-bot[bot]@users.noreply.github.com>
Co-authored-by: Joanne Wang <[email protected]>
@opensearch-trigger-bot
Copy link
Contributor

The backport to 2.11 failed:

The process '/usr/bin/git' failed with exit code 128

To backport manually, run these commands in your terminal:

# Navigate to the root of your repository
cd $(git rev-parse --show-toplevel)
# Fetch latest updates from GitHub
git fetch
# Create a new working tree
git worktree add ../.worktrees/alerting/backport-2.11 2.11
# Navigate to the new working tree
pushd ../.worktrees/alerting/backport-2.11
# Create a new branch
git switch --create backport-1403-to-2.11
# Cherry-pick the merged commit of this pull request and resolve the conflicts
git cherry-pick -x --mainline 1 bcaa4e36d86675850b537425b066f3c15cda15cc
# Push it to GitHub
git push --set-upstream origin backport-1403-to-2.11
# Go back to the original working tree
popd
# Delete the working tree
git worktree remove ../.worktrees/alerting/backport-2.11

Then, create a pull request where the base branch is 2.11 and the compare/head branch is backport-1403-to-2.11.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants