Use circuit breaker in InternalHistogram when adding empty buckets

[bowenlan-amzn]

Description

Utilize circuit breaker in InternalHistogram when adding empty buckets

Here's a example response when circuit breaker trips. Previously such request will bring the whole cluster down because of OutOfMemory error.

{
    "error": {
        "root_cause": [],
        "type": "search_phase_execution_exception",
        "reason": "",
        "phase": "fetch",
        "grouped": true,
        "failed_shards": [],
        "caused_by": {
            "type": "circuit_breaking_exception",
            "reason": "[parent] Data too large, data for [allocated_buckets] would be [512967080/489.2mb], which is larger than the limit of [510027366/486.3mb], real usage: [512967080/489.2mb], new bytes reserved: [0/0b], usages [request=153/153b, fielddata=0/0b, in_flight_requests=342/342b]",
            "bytes_wanted": 512967080,
            "bytes_limit": 510027366,
            "durability": "TRANSIENT"
        }
    },
    "status": 429
}

Related Issues

Resolves #14558

Check List

• Functionality includes testing.
• API changes companion pull request created, if applicable.
• Public documentation issue/PR created, if applicable.

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.

[bowenlan-amzn]

Details about the code flow around the reduceContext.consumeBucketsAndMaybeBreak(0) added in this PR

every 1024 call, will use circuit breaker to do a check addEstimateBytesAndMaybeBreak

OpenSearch/server/src/main/java/org/opensearch/search/aggregations/MultiBucketConsumerService.java

Lines 195 to 202 in 618782d

	if ((sum >= 1024) && (sum & 0x3FF) <= availProcessors) {
	try {
	breaker.addEstimateBytesAndMaybeBreak(0, "allocated_buckets");
	} catch (CircuitBreakingException e) {
	circuitBreakerTripped = true;
	throw e;
	}
	}

this circuit breaker check will check the real heap memory usage of the node and throw exception if it trips the parent circuit breaker indices.breaker.total.limit which defaults to 95%

OpenSearch/server/src/main/java/org/opensearch/common/breaker/ChildMemoryCircuitBreaker.java

Line 137 in c0852f8

parent.checkParentLimit((long) (bytes * overheadConstant), label);

OpenSearch/server/src/main/java/org/opensearch/indices/breaker/HierarchyCircuitBreakerService.java

Line 413 in aca2e9d

final MemoryUsage memoryUsed = memoryUsed(newBytesReserved);

OpenSearch/server/src/main/java/org/opensearch/indices/breaker/HierarchyCircuitBreakerService.java

Line 379 in aca2e9d

final long current = currentMemoryUsage();

OpenSearch/server/src/main/java/org/opensearch/indices/breaker/HierarchyCircuitBreakerService.java

Lines 388 to 394 in aca2e9d

	long currentMemoryUsage() {
	return realMemoryUsage();
	}
	static long realMemoryUsage() {
	try {
	return MEMORY_MX_BEAN.getHeapMemoryUsage().getUsed();

[jainankitk]

LGTM!

[github-actions]

❕ Gradle check result for 06d52c4: UNSTABLE

Please review all flaky tests that succeeded after retry and create an issue if one does not already exist to track the flaky failure.

[opensearch-trigger-bot]

The backport to 2.x 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/OpenSearch/backport-2.x 2.x
# Navigate to the new working tree
pushd ../.worktrees/OpenSearch/backport-2.x
# Create a new branch
git switch --create backport/backport-14754-to-2.x
# Cherry-pick the merged commit of this pull request and resolve the conflicts
git cherry-pick -x --mainline 1 b980b12e7fa14620aab38363f316ce72af70df1e
# Push it to GitHub
git push --set-upstream origin backport/backport-14754-to-2.x
# Go back to the original working tree
popd
# Delete the working tree
git worktree remove ../.worktrees/OpenSearch/backport-2.x

Then, create a pull request where the base branch is 2.x and the compare/head branch is backport/backport-14754-to-2.x.