[Obs AI Assistant] Serverless API integration tests #192219

neptunian · 2024-09-05T20:05:22Z

Tests for serverless

copies over and modifies all tests from stateful to work in serverless. ~~deployment agnostic tests do not yet support enterprise license for stateful, so are tests don't yet qualify as being deployment agnostic~~. Given how difficult it is to see differences from the stateful tests, I've added PR comments where I've changed something that might be of interest.
changes to createObservabilityAIAssistantApiClient to use supertest without basic auth and accept headers for serverless and use roles
removes creating persisted users when tests start and use roles within tests. its not possible to create custom users with the serverless test framework at the moment. See [Obs AI Assistant] Serverless tests should use cookie based auth instead of api keys for internal routes #192711

Skipped tests

knowledge base tests [Obs AI Assistant] fix knowledge base tests failing in serverless #192886
any test suite that uses the LLM proxy has been skipped on MKI [Obs AI Assistant] Fix tests using Llm proxy in serverless #192751
all tests that depend on the config.modelId skipped in MKI [Obs AI Assistant] Update knowledge base setup route to accept a model id #192757

TODO:

move over remaining tests
test in MKI environment before merging
create issues for skipped tests
this will not run on MKI (after merging) unless we ping the appex-qa team to add it to the pipeline. this is due to creating a separate config. ask appex-qa team to add our config.

Followup / related issues to be tracked in a newly created issue:

[Obs AI Assistant] Update knowledge base setup route to accept a model id #192757
[Obs AI Assistant] fix knowledge base tests failing in serverless #192886
[Obs AI Assistant] Fix tests using Llm proxy in serverless #192751
[Obs AI Assistant] Fetch user instructions using id instead of username for knowledge base instructions #192701
[Obs AI Assistant] unskip summarize tests #192497
[Obs AI Assistant] Serverless tests should use cookie based auth instead of api keys for internal routes #192711
[Obs AI Assistant] Deployment agnostic API integration tests #192718
serverless functional tests
inquire with ml-ui-team to have the ability to delete system indices which we do after uninstalling tiny elser with .ml indices

obltmachine · 2024-09-05T20:05:35Z

🤖 GitHub comments

Expand to view the GitHub comments

Just comment with:

/oblt-deploy : Deploy a Kibana instance using the Observability test environments.
run docs-build : Re-trigger the docs validation. (use unformatted text in the comment!)

neptunian · 2024-09-06T12:11:07Z

/ci

neptunian · 2024-09-09T16:57:48Z

...ck/plugins/observability_solution/observability_ai_assistant/common/conversation_complete.ts

@@ -104,7 +104,8 @@ export type StreamingChatResponseEvent =
 | ConversationUpdateEvent
 | MessageAddEvent
 | ChatCompletionErrorEvent
- | TokenCountEvent;
+ | TokenCountEvent
+ | BufferFlushEvent;


BufferFlushEvent can be returned as a streaming chat response event in cloud

Yes but it's intentionally not here to not overcomplicate the StreamingChatResponseEvent type

I only added it due to causing a type error in complete.spec.ts due to filtering out BufferFlushEvents appearing in the data. Is it okay to leave it?

neptunian · 2024-09-09T17:03:02Z

...rless/api_integration/test_suites/observability/ai_assistant/tests/complete/complete.spec.ts

+ .filter(Boolean)
+ .map((line) => JSON.parse(line) as StreamingChatResponseEvent)
+ // Filter BufferFlush events that appear if isCloudEnabled is true which is the case in serverless tests
+ .filter((event) => event.type !== StreamingChatResponseEventType.BufferFlush)


added filtering out bufferFlush events so as not to break test below that checks for the 4th event which gets pushed down in serverless from the BufferFlush event.

neptunian · 2024-09-09T19:16:37Z

...ration/test_suites/observability/ai_assistant/tests/complete/functions/elasticsearch.spec.ts

+export default function ApiTest({ getService }: FtrProviderContext) {
+ const supertest = getService('supertest');
+ const log = getService('log');
+ const synthtrace = getService('synthtrace');


since this was added as a common service already, i've used it and not added a local one. we should extend it to uninstall the apm package similar to stateful, as a cleanup function or via the kibana client as we have in stateful.

…into serverless-tests

kibanamachine · 2024-09-30T23:11:40Z

💔 All backports failed

Status	Branch	Result
❌	8.x	Backport failed because of merge conflicts You might need to backport the following PRs to 8.x: - Revert "[UII] Fix client-side validation for agent policy timeout fields (#191674)" (#194338) - [Threat Hunting Investigations] Migrate all timeline routes to OpenAPI types (#190238) - [Security Solution] Rule `type` field diff algorithm (#193369) - Hide progress bar on initial load when client prefers reduced motion (#194365) - [Security Solution][Detection Engine] fixes empty item_id in exception item update request schema (#194175)

Manual backport

To create the backport manually run:

node scripts/backport --pr 192219

Questions ?

Please refer to the Backport tool documentation

Tests for serverless - copies over and modifies all tests from stateful to work in serverless. ~~deployment agnostic tests do not yet support enterprise license for stateful, so are tests don't yet qualify as being deployment agnostic~~. Given how difficult it is to see differences from the stateful tests, I've added PR comments where I've changed something that might be of interest. - changes to `createObservabilityAIAssistantApiClient` to use supertest without basic auth and accept headers for serverless and use roles - removes creating persisted users when tests start and [use roles](https:/elastic/kibana/blob/main/x-pack/test_serverless/README.md#roles-based-testing) within tests. its not possible to create custom users with the serverless test framework at the moment. See elastic#192711 Skipped tests - knowledge base tests elastic#192886 - any test suite that uses the LLM proxy has been skipped on MKI elastic#192751 - all tests that depend on the config.modelId skipped in MKI elastic#192757 TODO: - [x] move over remaining tests - [x] test in MKI environment before merging - [x] create issues for skipped tests - [ ] this will not run on MKI (after merging) unless we ping the appex-qa team to add it to the pipeline. this is due to creating a separate config. ask appex-qa team to add our config. Followup / related issues to be tracked in a newly created issue: - [ ] elastic#192757 - [ ] elastic#192886 - [ ] elastic#192751 - [ ] elastic#192701 - [ ] elastic#192497 - [ ] elastic#192711 - [ ] elastic#192718 - [ ] serverless functional tests - [ ] inquire with ml-ui-team to have the ability to delete system indices which we do after uninstalling tiny elser with .ml indices

…194582) Tests for serverless - copies over and modifies all tests from stateful to work in serverless. ~~deployment agnostic tests do not yet support enterprise license for stateful, so are tests don't yet qualify as being deployment agnostic~~. Given how difficult it is to see differences from the stateful tests, I've added PR comments where I've changed something that might be of interest. - changes to `createObservabilityAIAssistantApiClient` to use supertest without basic auth and accept headers for serverless and use roles - removes creating persisted users when tests start and [use roles](https:/elastic/kibana/blob/main/x-pack/test_serverless/README.md#roles-based-testing) within tests. its not possible to create custom users with the serverless test framework at the moment. See #192711 Skipped tests - knowledge base tests #192886 - any test suite that uses the LLM proxy has been skipped on MKI #192751 - all tests that depend on the config.modelId skipped in MKI #192757 TODO: - [x] move over remaining tests - [x] test in MKI environment before merging - [x] create issues for skipped tests - [ ] this will not run on MKI (after merging) unless we ping the appex-qa team to add it to the pipeline. this is due to creating a separate config. ask appex-qa team to add our config. Followup / related issues to be tracked in a newly created issue: - [ ] #192757 - [ ] #192886 - [ ] #192751 - [ ] #192701 - [ ] #192497 - [ ] #192711 - [ ] #192718 - [ ] serverless functional tests - [ ] inquire with ml-ui-team to have the ability to delete system indices which we do after uninstalling tiny elser with .ml indices ## Summary Summarize your PR. If it involves visual changes include a screenshot or gif. ### Checklist Delete any items that are not applicable to this PR. - [ ] Any text added follows [EUI's writing guidelines](https://elastic.github.io/eui/#/guidelines/writing), uses sentence case text and includes [i18n support](https:/elastic/kibana/blob/main/packages/kbn-i18n/README.md) - [ ] [Documentation](https://www.elastic.co/guide/en/kibana/master/development-documentation.html) was added for features that require explanation or tutorials - [ ] [Unit or functional tests](https://www.elastic.co/guide/en/kibana/master/development-tests.html) were updated or added to match the most common scenarios - [ ] [Flaky Test Runner](https://ci-stats.kibana.dev/trigger_flaky_test_runner/1) was used on any tests changed - [ ] Any UI touched in this PR is usable by keyboard only (learn more about [keyboard accessibility](https://webaim.org/techniques/keyboard/)) - [ ] Any UI touched in this PR does not create any new axe failures (run axe in browser: [FF](https://addons.mozilla.org/en-US/firefox/addon/axe-devtools/), [Chrome](https://chrome.google.com/webstore/detail/axe-web-accessibility-tes/lhdoppojpmngadmnindnejefpokejbdd?hl=en-US)) - [ ] If a plugin configuration key changed, check if it needs to be allowlisted in the cloud and added to the [docker list](https:/elastic/kibana/blob/main/src/dev/build/tasks/os_packages/docker_generator/resources/base/bin/kibana-docker) - [ ] This renders correctly on smaller devices using a responsive layout. (You can test this [in your browser](https://www.browserstack.com/guide/responsive-testing-on-local-server)) - [ ] This was checked for [cross-browser compatibility](https://www.elastic.co/support/matrix#matrix_browsers) ### Risk Matrix Delete this section if it is not applicable to this PR. Before closing this PR, invite QA, stakeholders, and other developers to identify risks that should be tested prior to the change/feature release. When forming the risk matrix, consider some of the following examples and how they may potentially impact the change: | Risk | Probability | Severity | Mitigation/Notes | |---------------------------|-------------|----------|-------------------------| | Multiple Spaces—unexpected behavior in non-default Kibana Space. | Low | High | Integration tests will verify that all features are still supported in non-default Kibana Space and when user switches between spaces. | | Multiple nodes—Elasticsearch polling might have race conditions when multiple Kibana nodes are polling for the same tasks. | High | Low | Tasks are idempotent, so executing them multiple times will not result in logical error, but will degrade performance. To test for this case we add plenty of unit tests around this logic and document manual testing procedure. | | Code should gracefully handle cases when feature X or plugin Y are disabled. | Medium | High | Unit tests will verify that any feature flag or plugin combination still results in our service operational. | | [See more potential risk examples](https:/elastic/kibana/blob/main/RISK_MATRIX.mdx) | ### For maintainers - [ ] This was checked for breaking API changes and was [labeled appropriately](https://www.elastic.co/guide/en/kibana/master/contributing.html#kibana-release-notes-process)

kibanamachine · 2024-10-01T23:48:37Z

Friendly reminder: Looks like this PR hasn’t been backported yet.
To create automatically backports add the label auto-backport or prevent reminders by adding the backport:skip label.
You can also create backports manually by running node scripts/backport --pr 192219 locally

Tests for serverless - copies over and modifies all tests from stateful to work in serverless. ~~deployment agnostic tests do not yet support enterprise license for stateful, so are tests don't yet qualify as being deployment agnostic~~. Given how difficult it is to see differences from the stateful tests, I've added PR comments where I've changed something that might be of interest. - changes to `createObservabilityAIAssistantApiClient` to use supertest without basic auth and accept headers for serverless and use roles - removes creating persisted users when tests start and [use roles](https:/elastic/kibana/blob/main/x-pack/test_serverless/README.md#roles-based-testing) within tests. its not possible to create custom users with the serverless test framework at the moment. See elastic#192711 Skipped tests - knowledge base tests elastic#192886 - any test suite that uses the LLM proxy has been skipped on MKI elastic#192751 - all tests that depend on the config.modelId skipped in MKI elastic#192757 TODO: - [x] move over remaining tests - [x] test in MKI environment before merging - [x] create issues for skipped tests - [ ] this will not run on MKI (after merging) unless we ping the appex-qa team to add it to the pipeline. this is due to creating a separate config. ask appex-qa team to add our config. Followup / related issues to be tracked in a newly created issue: - [ ] elastic#192757 - [ ] elastic#192886 - [ ] elastic#192751 - [ ] elastic#192701 - [ ] elastic#192497 - [ ] elastic#192711 - [ ] elastic#192718 - [ ] serverless functional tests - [ ] inquire with ml-ui-team to have the ability to delete system indices which we do after uninstalling tiny elser with .ml indices

kibanamachine · 2024-10-02T23:49:09Z