kibana/x-pack/plugins/task_manager/server/polling_lifecycle.test.ts
Mike Côté cb2e28d1e4
Fix task manager polling flow controls (#153491)
Fixes https://github.com/elastic/kibana/issues/151938

In this PR, I'm re-writing the Task Manager poller so it doesn't run
concurrently when timeouts occur while also fixing the issue where
polling requests would pile up when polling takes time. To support this,
I've also made the following changes:
- Removed the observable monitor and the
`xpack.task_manager.max_poll_inactivity_cycles` setting
- Make the task store `search` and `updateByQuery` functions have no
retries. This prevents the request from retrying 5x whenever a timeout
occurs, causing each call taking up to 2 1/2 minutes before Kibana sees
the error (now down to 30s each). We have polling to manage retries in
these situations.
- Switch the task poller tests to use `sinon` for faking timers
- Removing the `assertStillInSetup` checks on plugin setup. Felt like a
maintenance burden that wasn't necessary to fix with my code changes.

The main code changes are within these files (to review thoroughly so
the polling cycle doesn't suddenly stop):
- x-pack/plugins/task_manager/server/polling/task_poller.ts
- x-pack/plugins/task_manager/server/polling_lifecycle.ts (easier to
review if you disregard whitespace `?w=1`)

## To verify
1. Tasks run normally (create a rule or something that goes through task
manager regularly).
2. When the update by query takes a while, the request is cancelled
after 30s or the time manually configured.
4. When the search for claimed tasks query takes a while, the request is
cancelled after 30s or the time manually configured.

**Tips:**
<details><summary>how to slowdown search for claimed task
queries</summary>

```
diff --git a/x-pack/plugins/task_manager/server/queries/task_claiming.ts b/x-pack/plugins/task_manager/server/queries/task_claiming.ts
index 07042650a37..2caefd63672 100644
--- a/x-pack/plugins/task_manager/server/queries/task_claiming.ts
+++ b/x-pack/plugins/task_manager/server/queries/task_claiming.ts
@@ -247,7 +247,7 @@ export class TaskClaiming {
         taskTypes,
       });

-    const docs = tasksUpdated > 0 ? await this.sweepForClaimedTasks(taskTypes, size) : [];
+    const docs = await this.sweepForClaimedTasks(taskTypes, size);

     this.emitEvents(docs.map((doc) => asTaskClaimEvent(doc.id, asOk(doc))));

@@ -346,6 +346,13 @@ export class TaskClaiming {
       size,
       sort: SortByRunAtAndRetryAt,
       seq_no_primary_term: true,
+      aggs: {
+        delay: {
+          shard_delay: {
+            value: '40s',
+          },
+        },
+      },
     });

     return docs;
```
</details>

<details><summary>how to slow down update by queries</summary>
Not the cleanest way but you'll see occasional request timeouts from the
updateByQuery calls. I had more luck creating rules running every 1s.

```
diff --git a/x-pack/plugins/task_manager/server/task_store.ts b/x-pack/plugins/task_manager/server/task_store.ts
index a06ee7b918a..07aa81e5388 100644
--- a/x-pack/plugins/task_manager/server/task_store.ts
+++ b/x-pack/plugins/task_manager/server/task_store.ts
@@ -126,6 +126,7 @@ export class TaskStore {
       // Timeouts are retried and make requests timeout after (requestTimeout * (1 + maxRetries))
       // The poller doesn't need retry logic because it will try again at the next polling cycle
       maxRetries: 0,
+      requestTimeout: 900,
     });
   }

@@ -458,6 +459,7 @@ export class TaskStore {
           ignore_unavailable: true,
           refresh: true,
           conflicts: 'proceed',
+          requests_per_second: 1,
           body: {
             ...opts,
             max_docs,
```
</details>

---------

Co-authored-by: Kibana Machine <42973632+kibanamachine@users.noreply.github.com>
2023-05-03 09:33:10 -04:00

289 lines
9.9 KiB
TypeScript

/*
* Copyright Elasticsearch B.V. and/or licensed to Elasticsearch B.V. under one
* or more contributor license agreements. Licensed under the Elastic License
* 2.0; you may not use this file except in compliance with the Elastic License
* 2.0.
*/
import sinon from 'sinon';
import { Observable, of, Subject } from 'rxjs';
import { TaskPollingLifecycle, claimAvailableTasks } from './polling_lifecycle';
import { createInitialMiddleware } from './lib/middleware';
import { TaskTypeDictionary } from './task_type_dictionary';
import { taskStoreMock } from './task_store.mock';
import { mockLogger } from './test_utils';
import { taskClaimingMock } from './queries/task_claiming.mock';
import { TaskClaiming, ClaimOwnershipResult } from './queries/task_claiming';
import type { TaskClaiming as TaskClaimingClass } from './queries/task_claiming';
import { asOk, Err, isErr, isOk, Result } from './lib/result_type';
import { FillPoolResult } from './lib/fill_pool';
import { ElasticsearchResponseError } from './lib/identify_es_error';
import { executionContextServiceMock } from '@kbn/core/server/mocks';
const executionContext = executionContextServiceMock.createSetupContract();
let mockTaskClaiming = taskClaimingMock.create({});
jest.mock('./queries/task_claiming', () => {
return {
TaskClaiming: jest.fn().mockImplementation(() => {
return mockTaskClaiming;
}),
};
});
jest.mock('./constants', () => ({
CONCURRENCY_ALLOW_LIST_BY_TASK_TYPE: ['report', 'quickReport'],
}));
describe('TaskPollingLifecycle', () => {
let clock: sinon.SinonFakeTimers;
const taskManagerLogger = mockLogger();
const mockTaskStore = taskStoreMock.create({});
const taskManagerOpts = {
config: {
enabled: true,
max_workers: 10,
index: 'foo',
max_attempts: 9,
poll_interval: 6000000,
version_conflict_threshold: 80,
request_capacity: 1000,
monitored_aggregated_stats_refresh_rate: 5000,
monitored_stats_health_verbose_log: {
enabled: false,
level: 'debug' as const,
warn_delayed_task_start_in_seconds: 60,
},
monitored_stats_required_freshness: 5000,
monitored_stats_running_average_window: 50,
monitored_task_execution_thresholds: {
default: {
error_threshold: 90,
warn_threshold: 80,
},
custom: {},
},
ephemeral_tasks: {
enabled: true,
request_capacity: 10,
},
unsafe: {
exclude_task_types: [],
},
event_loop_delay: {
monitor: true,
warn_threshold: 5000,
},
worker_utilization_running_average_window: 5,
},
taskStore: mockTaskStore,
logger: taskManagerLogger,
unusedTypes: [],
definitions: new TaskTypeDictionary(taskManagerLogger),
middleware: createInitialMiddleware(),
maxWorkersConfiguration$: of(100),
pollIntervalConfiguration$: of(100),
executionContext,
};
beforeEach(() => {
mockTaskClaiming = taskClaimingMock.create({});
(TaskClaiming as jest.Mock<TaskClaimingClass>).mockClear();
clock = sinon.useFakeTimers();
});
afterEach(() => clock.restore());
describe('start', () => {
test('begins polling once the ES and SavedObjects services are available', () => {
const elasticsearchAndSOAvailability$ = new Subject<boolean>();
new TaskPollingLifecycle({
...taskManagerOpts,
elasticsearchAndSOAvailability$,
});
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).not.toHaveBeenCalled();
elasticsearchAndSOAvailability$.next(true);
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).toHaveBeenCalled();
});
test('provides TaskClaiming with the capacity available', () => {
const elasticsearchAndSOAvailability$ = new Subject<boolean>();
const maxWorkers$ = new Subject<number>();
taskManagerOpts.definitions.registerTaskDefinitions({
report: {
title: 'report',
maxConcurrency: 1,
createTaskRunner: jest.fn(),
},
quickReport: {
title: 'quickReport',
maxConcurrency: 5,
createTaskRunner: jest.fn(),
},
});
new TaskPollingLifecycle({
...taskManagerOpts,
elasticsearchAndSOAvailability$,
maxWorkersConfiguration$: maxWorkers$,
});
const taskClaimingGetCapacity = (TaskClaiming as jest.Mock<TaskClaimingClass>).mock
.calls[0][0].getCapacity;
maxWorkers$.next(20);
expect(taskClaimingGetCapacity()).toEqual(20);
expect(taskClaimingGetCapacity('report')).toEqual(1);
expect(taskClaimingGetCapacity('quickReport')).toEqual(5);
maxWorkers$.next(30);
expect(taskClaimingGetCapacity()).toEqual(30);
expect(taskClaimingGetCapacity('report')).toEqual(1);
expect(taskClaimingGetCapacity('quickReport')).toEqual(5);
maxWorkers$.next(2);
expect(taskClaimingGetCapacity()).toEqual(2);
expect(taskClaimingGetCapacity('report')).toEqual(1);
expect(taskClaimingGetCapacity('quickReport')).toEqual(2);
});
});
describe('stop', () => {
test('stops polling once the ES and SavedObjects services become unavailable', () => {
const elasticsearchAndSOAvailability$ = new Subject<boolean>();
new TaskPollingLifecycle({
elasticsearchAndSOAvailability$,
...taskManagerOpts,
});
elasticsearchAndSOAvailability$.next(true);
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).toHaveBeenCalled();
elasticsearchAndSOAvailability$.next(false);
mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable.mockClear();
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).not.toHaveBeenCalled();
});
test('restarts polling once the ES and SavedObjects services become available again', () => {
const elasticsearchAndSOAvailability$ = new Subject<boolean>();
new TaskPollingLifecycle({
elasticsearchAndSOAvailability$,
...taskManagerOpts,
});
elasticsearchAndSOAvailability$.next(true);
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).toHaveBeenCalled();
elasticsearchAndSOAvailability$.next(false);
mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable.mockClear();
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).not.toHaveBeenCalled();
elasticsearchAndSOAvailability$.next(true);
clock.tick(150);
expect(mockTaskClaiming.claimAvailableTasksIfCapacityIsAvailable).toHaveBeenCalled();
});
});
describe('claimAvailableTasks', () => {
test('should claim Available Tasks when there are available workers', async () => {
const logger = mockLogger();
const taskClaiming = taskClaimingMock.create({});
taskClaiming.claimAvailableTasksIfCapacityIsAvailable.mockImplementation(() =>
of(
asOk({
docs: [],
stats: { tasksUpdated: 0, tasksConflicted: 0, tasksClaimed: 0, tasksRejected: 0 },
})
)
);
expect(isOk(await getFirstAsPromise(claimAvailableTasks(taskClaiming, logger)))).toBeTruthy();
expect(taskClaiming.claimAvailableTasksIfCapacityIsAvailable).toHaveBeenCalledTimes(1);
});
/**
* This handles the case in which Elasticsearch has had inline script disabled.
* This is achieved by setting the `script.allowed_types` flag on Elasticsearch to `none`
*/
test('handles failure due to inline scripts being disabled', async () => {
const logger = mockLogger();
const taskClaiming = taskClaimingMock.create({});
taskClaiming.claimAvailableTasksIfCapacityIsAvailable.mockImplementation(
() =>
new Observable<Result<ClaimOwnershipResult, FillPoolResult>>((observer) => {
observer.error({
name: 'ResponseError',
meta: {
body: {
error: {
root_cause: [
{
type: 'illegal_argument_exception',
reason: 'cannot execute [inline] scripts',
},
],
type: 'search_phase_execution_exception',
reason: 'all shards failed',
phase: 'query',
grouped: true,
failed_shards: [
{
shard: 0,
index: '.kibana_task_manager_1',
node: '24A4QbjHSK6prvtopAKLKw',
reason: {
type: 'illegal_argument_exception',
reason: 'cannot execute [inline] scripts',
},
},
],
caused_by: {
type: 'illegal_argument_exception',
reason: 'cannot execute [inline] scripts',
caused_by: {
type: 'illegal_argument_exception',
reason: 'cannot execute [inline] scripts',
},
},
},
status: 400,
},
},
statusCode: 400,
} as ElasticsearchResponseError);
})
);
const err = await getFirstAsPromise(claimAvailableTasks(taskClaiming, logger));
expect(isErr(err)).toBeTruthy();
expect((err as Err<FillPoolResult>).error).toEqual(FillPoolResult.Failed);
expect(logger.warn).toHaveBeenCalledTimes(1);
expect(logger.warn).toHaveBeenCalledWith(
`Task Manager cannot operate when inline scripts are disabled in Elasticsearch`
);
});
});
});
function getFirstAsPromise<T>(obs$: Observable<T>): Promise<T> {
return new Promise((resolve, reject) => {
obs$.subscribe(resolve, reject);
});
}