[ResponseOps] Cases analytics index synchronization#222820
[ResponseOps] Cases analytics index synchronization#222820adcoelho merged 7 commits intoelastic:cases-analytics-indexfrom
Conversation
|
Pinging @elastic/response-ops (Team:ResponseOps) |
|
Pinging @elastic/response-ops-cases (Feature:Cases) |
707b861 to
050f701
Compare
9946246 to
c393168
Compare
|
The failing functional tests with |
cnasikas
left a comment
There was a problem hiding this comment.
Great work!! LGTM! I left some comments. Most of them are nits.
| }); | ||
| }); | ||
|
|
||
| describe('Error handling', () => { |
There was a problem hiding this comment.
For tests with try/catch, we should put it at the beginning of the test expect.assertions(1), so the test will fail if the assertion in the catch is not called for some reason.
| expect(isRetryableError(e)).toBe(null); | ||
| } | ||
|
|
||
| expect(logger.error).toBeCalledWith( |
There was a problem hiding this comment.
Nit: I think expecting certain error messages can lead to tests that are not resilient to refactoring. For example, every time we make a small change in the message, we should also update all tests. What about just checking expect(logger.error).toHaveBeenCalled();
c393168 to
d97eeeb
Compare
7b1e564
into
elastic:cases-analytics-index
💔 Build Failed
Failed CI StepsTest Failures
Metrics [docs]
History
cc @adcoelho |
This PR is for a feature branch that is being merged into main. The relevant PRs are: - #219211 - #222820 - #223241 - #224388 - #224682 ## Summary This PR adds 4 new indexes with case analytics data, which are created when the cases plugin starts. - `.internal.cases` - `.internal.cases-comments` - `.internal.cases-attachments` - `.internal.cases-activity` After the indexes are created, a backfill task for each of them is scheduled to run 1 minute after creation. This task populates the indexes with relevant data from `.kibana_alerting_cases`. A second type of task is registered, the index synchronization task. Four of these tasks, one for each index, are scheduled to run every 5 minutes. The synchronization tasks populated the indexes with data from `.kibana_alerting_cases` that was created or updated in the last five minutes. ## How to test You might want to start Kibana with `--verbose` to see relevant index messages in the console. Alternatively(what I normally do), is go to `analytics_index.ts`, `backfill_task_runner.ts`, and `synchronization_task_runner.ts`, and change the `logDebug` function to call `this.logger.info` instead. This way, you will have less spam in the console. Every log message starts with the index name between square brackets, so you can look for `[.internal.cases-` and follow what is happening. 1. You should have some existing case data, so before anything else, please create some activity, attachments, etc. 2. Add `xpack.cases.analytics.index.enabled: true` to `kibana.dev.yml` and restart Kibana. 3. Check out [this branch](elastic/elasticsearch#129414) from the ES project. 4. Start Elastic Search with `yarn es source`. This will use the above version of Elasticsearch. 5. Wait a bit for the indexes to be created and populated(backfilled). 6. Using the dev tools: - Confirm the indexes exist. - Check the index mapping. Does it match the one in the code? Is the `_meta` field correct? - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/mappings.ts` - Check that the painless scripts match the code. - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/painless_scripts.ts` - Confirm your existing case data is in the indexes. (See **Queries** section below.) 7. Play around with cases. Some examples: - Create a case - Change status/severity - Attach alerts - Add files - Change category/tags - Add comments - etc 8. Go to the dev tools again and confirm all this shows up in the relevant indexes. (See **Queries** section below.) ## Queries In addition to the ones, below I have a few more. Things like reindexing with specific scripts or fetching relevant data from `.kibana_alerting_cases`. Ping me if you want those queries. ### Checking index content ``` GET /.internal.cases/_search GET /.internal.cases-comments/_search GET /.internal.cases-attachments/_search GET /.internal.cases-activity/_search ``` ### Checking index mappings ``` GET /.internal.cases GET /.internal.cases-comments GET /.internal.cases-attachments GET /.internal.cases-activity ``` ### Fetching the painless scripts ``` GET /_scripts/cai_cases_script_1 GET /_scripts/cai_attachments_script_1 GET /_scripts/cai_comments_script_1 GET /_scripts/cai_activity_script_1 ``` ### Emptying the indexes It is sometimes useful for testing. ``` POST /.internal.cases/_delete_by_query POST /.internal.cases-comments/_delete_by_query POST /.internal.cases-attachments/_delete_by_query POST /.internal.cases-activity/_delete_by_query ``` ### Deleting the indexes It is sometimes useful for testing. ``` DELETE /.internal.cases DELETE /.internal.cases-comments DELETE /.internal.cases-attachments DELETE /.internal.cases-activity ``` ## Release notes Four dedicated case analytics indexes were created, allowing users to build dashboards and metrics over case data. These indexes are created on Kibana startup and updated periodically with cases, comments, attachments, and activity data. --------- Co-authored-by: kibanamachine <42973632+kibanamachine@users.noreply.github.com> Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> Co-authored-by: Christos Nasikas <christos.nasikas@elastic.co>
This PR is for a feature branch that is being merged into main. The relevant PRs are: - elastic#219211 - elastic#222820 - elastic#223241 - elastic#224388 - elastic#224682 ## Summary This PR adds 4 new indexes with case analytics data, which are created when the cases plugin starts. - `.internal.cases` - `.internal.cases-comments` - `.internal.cases-attachments` - `.internal.cases-activity` After the indexes are created, a backfill task for each of them is scheduled to run 1 minute after creation. This task populates the indexes with relevant data from `.kibana_alerting_cases`. A second type of task is registered, the index synchronization task. Four of these tasks, one for each index, are scheduled to run every 5 minutes. The synchronization tasks populated the indexes with data from `.kibana_alerting_cases` that was created or updated in the last five minutes. ## How to test You might want to start Kibana with `--verbose` to see relevant index messages in the console. Alternatively(what I normally do), is go to `analytics_index.ts`, `backfill_task_runner.ts`, and `synchronization_task_runner.ts`, and change the `logDebug` function to call `this.logger.info` instead. This way, you will have less spam in the console. Every log message starts with the index name between square brackets, so you can look for `[.internal.cases-` and follow what is happening. 1. You should have some existing case data, so before anything else, please create some activity, attachments, etc. 2. Add `xpack.cases.analytics.index.enabled: true` to `kibana.dev.yml` and restart Kibana. 3. Check out [this branch](elastic/elasticsearch#129414) from the ES project. 4. Start Elastic Search with `yarn es source`. This will use the above version of Elasticsearch. 5. Wait a bit for the indexes to be created and populated(backfilled). 6. Using the dev tools: - Confirm the indexes exist. - Check the index mapping. Does it match the one in the code? Is the `_meta` field correct? - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/mappings.ts` - Check that the painless scripts match the code. - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/painless_scripts.ts` - Confirm your existing case data is in the indexes. (See **Queries** section below.) 7. Play around with cases. Some examples: - Create a case - Change status/severity - Attach alerts - Add files - Change category/tags - Add comments - etc 8. Go to the dev tools again and confirm all this shows up in the relevant indexes. (See **Queries** section below.) ## Queries In addition to the ones, below I have a few more. Things like reindexing with specific scripts or fetching relevant data from `.kibana_alerting_cases`. Ping me if you want those queries. ### Checking index content ``` GET /.internal.cases/_search GET /.internal.cases-comments/_search GET /.internal.cases-attachments/_search GET /.internal.cases-activity/_search ``` ### Checking index mappings ``` GET /.internal.cases GET /.internal.cases-comments GET /.internal.cases-attachments GET /.internal.cases-activity ``` ### Fetching the painless scripts ``` GET /_scripts/cai_cases_script_1 GET /_scripts/cai_attachments_script_1 GET /_scripts/cai_comments_script_1 GET /_scripts/cai_activity_script_1 ``` ### Emptying the indexes It is sometimes useful for testing. ``` POST /.internal.cases/_delete_by_query POST /.internal.cases-comments/_delete_by_query POST /.internal.cases-attachments/_delete_by_query POST /.internal.cases-activity/_delete_by_query ``` ### Deleting the indexes It is sometimes useful for testing. ``` DELETE /.internal.cases DELETE /.internal.cases-comments DELETE /.internal.cases-attachments DELETE /.internal.cases-activity ``` ## Release notes Four dedicated case analytics indexes were created, allowing users to build dashboards and metrics over case data. These indexes are created on Kibana startup and updated periodically with cases, comments, attachments, and activity data. --------- Co-authored-by: kibanamachine <42973632+kibanamachine@users.noreply.github.com> Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> Co-authored-by: Christos Nasikas <christos.nasikas@elastic.co> (cherry picked from commit e566fec) # Conflicts: # x-pack/platform/plugins/shared/task_manager/server/mocks.ts
This PR is for a feature branch that is being merged into main. The relevant PRs are: - elastic#219211 - elastic#222820 - elastic#223241 - elastic#224388 - elastic#224682 ## Summary This PR adds 4 new indexes with case analytics data, which are created when the cases plugin starts. - `.internal.cases` - `.internal.cases-comments` - `.internal.cases-attachments` - `.internal.cases-activity` After the indexes are created, a backfill task for each of them is scheduled to run 1 minute after creation. This task populates the indexes with relevant data from `.kibana_alerting_cases`. A second type of task is registered, the index synchronization task. Four of these tasks, one for each index, are scheduled to run every 5 minutes. The synchronization tasks populated the indexes with data from `.kibana_alerting_cases` that was created or updated in the last five minutes. ## How to test You might want to start Kibana with `--verbose` to see relevant index messages in the console. Alternatively(what I normally do), is go to `analytics_index.ts`, `backfill_task_runner.ts`, and `synchronization_task_runner.ts`, and change the `logDebug` function to call `this.logger.info` instead. This way, you will have less spam in the console. Every log message starts with the index name between square brackets, so you can look for `[.internal.cases-` and follow what is happening. 1. You should have some existing case data, so before anything else, please create some activity, attachments, etc. 2. Add `xpack.cases.analytics.index.enabled: true` to `kibana.dev.yml` and restart Kibana. 3. Check out [this branch](elastic/elasticsearch#129414) from the ES project. 4. Start Elastic Search with `yarn es source`. This will use the above version of Elasticsearch. 5. Wait a bit for the indexes to be created and populated(backfilled). 6. Using the dev tools: - Confirm the indexes exist. - Check the index mapping. Does it match the one in the code? Is the `_meta` field correct? - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/mappings.ts` - Check that the painless scripts match the code. - `x-pack/platform/plugins/shared/cases/server/cases_analytics/******_index/painless_scripts.ts` - Confirm your existing case data is in the indexes. (See **Queries** section below.) 7. Play around with cases. Some examples: - Create a case - Change status/severity - Attach alerts - Add files - Change category/tags - Add comments - etc 8. Go to the dev tools again and confirm all this shows up in the relevant indexes. (See **Queries** section below.) ## Queries In addition to the ones, below I have a few more. Things like reindexing with specific scripts or fetching relevant data from `.kibana_alerting_cases`. Ping me if you want those queries. ### Checking index content ``` GET /.internal.cases/_search GET /.internal.cases-comments/_search GET /.internal.cases-attachments/_search GET /.internal.cases-activity/_search ``` ### Checking index mappings ``` GET /.internal.cases GET /.internal.cases-comments GET /.internal.cases-attachments GET /.internal.cases-activity ``` ### Fetching the painless scripts ``` GET /_scripts/cai_cases_script_1 GET /_scripts/cai_attachments_script_1 GET /_scripts/cai_comments_script_1 GET /_scripts/cai_activity_script_1 ``` ### Emptying the indexes It is sometimes useful for testing. ``` POST /.internal.cases/_delete_by_query POST /.internal.cases-comments/_delete_by_query POST /.internal.cases-attachments/_delete_by_query POST /.internal.cases-activity/_delete_by_query ``` ### Deleting the indexes It is sometimes useful for testing. ``` DELETE /.internal.cases DELETE /.internal.cases-comments DELETE /.internal.cases-attachments DELETE /.internal.cases-activity ``` ## Release notes Four dedicated case analytics indexes were created, allowing users to build dashboards and metrics over case data. These indexes are created on Kibana startup and updated periodically with cases, comments, attachments, and activity data. --------- Co-authored-by: kibanamachine <42973632+kibanamachine@users.noreply.github.com> Co-authored-by: Elastic Machine <elasticmachine@users.noreply.github.com> Co-authored-by: Christos Nasikas <christos.nasikas@elastic.co>
Closes #221232
Merging into a feature branch
Summary
This PR adds:
How to test
A few things are needed.
index_creator. The password should bechangeme.Start Kibana with this user as the Elasticsearch user -
yarn startwith the option--elasticsearch.username=index_creator.You can check the content of these indexes in the Dev Tools.