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

[Response Ops][Alerting] Migrate installation of context-specific component templates, index templates and concrete write index to framework for alerts-as-data #151792

Merged
merged 15 commits into from
Mar 8, 2023

Conversation

ymao1
Copy link
Contributor

@ymao1 ymao1 commented Feb 21, 2023

Resolves #151697

Summary

In a previous PR we started installing a context-specific component templates, index templates and concrete write indices for framework alerts as data when the xpack.alerting.enableFrameworkAlerts config flag is set to true. In that PR we used a different naming pattern than what is used by the rule registry for those resources. In this PR, we are aligning the naming of these resources with the rule registry and installing these resources on alerting plugin setup when enableFrameworkAlerts: true. If the flag is set to false, the rule registry will continue to handle this resource installation.

In this PR we are doing the following:

  • Registering all rules currently registered with the rule registry with the alerting framework. This registration allows the alerting framework to build context specific component templates. Because this PR only addresses resource installation, rules will continue to be registered with the rule registry.
  • When enableFrameworkAlerts: true:
    • The framework installs the context specific component template with the following naming convention: .alerts-{context}.alerts-mappings. This matches what the rule registry currently installs so the transition should be seamless
    • The framework installs the context specific index template for the default space with the following name: .alerts-{context}.alerts-default-index-template. Space awareness will be addressed in a followup PR. This matches the current rule registry naming.This index template will reference
      (1) ECS component template (if useEcs: true),
      (2) context-specific component template,
      (3) legacy alert component template and
      (4) framework component template
      where the legacy alert component template + framework component template = technical component template (from the rule registry).
    • The framework creates or updates the concrete write index for the default space with the naming convention: .internal.alerts-{context}.alerts-default-000001. Space awareness will be addressed in a followup PR. This matches the current rule registry naming.
    • The installation of the index template & write index differs from the rule registry in that it occurs on alerting plugin start vs the first rule run.
  • We modified the rule registry resource installer to skip installation of these resources when enableFrameworkAlerts: true. In addition, it will wait for the alerting resource installation promise so if a rule runs before its resources are fully initialized, it will wait for initialization to complete before writing.

To Verify

The following rule registry contexts are affected:
observability.apm
observability.logs
observability.metrics
observability.slo
observability.uptime
security

For each context, we should verify the following:

Note that if your rule context references the ECS mappings, there may be differences in those mappings between main and this branch depending on whether you're running main with enableFrameworkAlerts true or false. These differences are explained in the summary of this prior PR: https://github.com/elastic/kibana/pull/150384 but essentially we're aligning with the latest ECS fields. In the instructions, I suggest running main with enableFrameworkAlerts: true to minimize the differences caused by ECS changes

While running main with enableFrameworkAlerts: true:

  1. Get the context specific component template GET _component_template/.alerts-{context}.alerts-mappings
  2. Create rule for this context that creates an alert and then
  3. Get the index template GET _index_template/.alerts-{context}.alerts-default-index-template
  4. Get the index mapping for the concrete index: GET .internal.alerts-{context}.alerts-default-000001/_mapping

While running this branch with xpack.alerting.enableFrameworkAlerts: true (with a fresh ES instance):
5. Get the context specific component template GET _component_template/.alerts-{context}.alerts-mappings
6. Get the index template GET _index_template/.alerts-{context}.alerts-default-index-template
7. Get the index mapping for the concrete index: GET .internal.alerts-{context}.alerts-default-000001/_mapping
Note that you should not have to create a rule that generates alerts before seeing these resources installed.

Compare the component templates
Compare 1 and 5. The difference should be:

  • component template from this branch should have _meta.managed: true. This is a flag indicating to the user that these templates are system managed and should not be manually modified.

Compare the index templates
Compare 3 and 6. The differences should be:

  • index template from this branch should have managed: true in the _meta fields
  • index template from this branch should not have a priority field. This will be addressed in a followup PR
  • index template from this branch should be composed of .alerts-legacy-alert-mappings and .alerts-framework-mappings instead of .alerts-technical-mappings but under the hood, these mappings are equivalent.

Compare the index mappings
Compare 4 and 7. The difference should be:

  • index mappings from this branch should have _meta.managed: true.

Verify that installed resources templates work as expected

  1. Run this branch on a fresh ES install with xpack.alerting.enableFrameworkAlerts: true.
  2. Create a rule in your context that generates alerts.
  3. Verify that there are no errors during rule execution.
  4. Verify that the alerts show up in your alerts table as expected.
  5. (For detection rules only): Run this branch with xpack.alerting.enableFrameworkAlerts: true and verify rules in a non-default space continue to create resources on first rule run and run as expected.
  6. (For detection rules only): Run this branch with xpack.alerting.enableFrameworkAlerts: true and verify rule preview continue to work as expected

Verify that installed resources templates work with existing rule registry resources.

  1. Run main or a previous version and create a rule in your context that generates alerts.
  2. Using the same ES data, switch to this branch with xpack.alerting.enableFrameworkAlerts: false and verify Kibana starts with no rule registry errors and the rule continues to run as expected.
  3. Using the same ES data, switch to this branch with xpack.alerting.enableFrameworkAlerts: true and verify Kibana starts with no alerting or rule registry errors and the rule continues to run as expected.
  4. Verify the alerts show up on the alerts table as expected.
  5. (For detection rules only): Run this branch with xpack.alerting.enableFrameworkAlerts: true and verify rules in a non-default space continue to create resources on first rule run and run as expected.
  6. (For detection rules only): Run this branch with xpack.alerting.enableFrameworkAlerts: true and verify rule preview continue to work as expected

@ymao1 ymao1 force-pushed the alerting/faad-resources-component branch 2 times, most recently from c24a423 to 9ea206c Compare February 24, 2023 20:03
@@ -0,0 +1,20 @@
/*
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Moved from x-pack/plugins/rule_registry/common/assets/field_maps/experimental_rule_field_map.ts

@ymao1 ymao1 force-pushed the alerting/faad-resources-component branch from 22a7324 to cc24f4d Compare February 27, 2023 19:51
@ymao1 ymao1 changed the title Alerting/faad resources component [Response Ops][Alerting] Installing context-specific component templates, index templates and concrete write index for framework alerts-as-data Feb 27, 2023
@ymao1 ymao1 changed the title [Response Ops][Alerting] Installing context-specific component templates, index templates and concrete write index for framework alerts-as-data [Response Ops][Alerting] Migrate installation of context-specific component templates, index templates and concrete write index to framework for alerts-as-data Feb 27, 2023
if (!isEqual(fieldMap, registeredFieldMap)) {
throw new Error(`${context} has already been registered with a different mapping`);
const registeredOptions = this.registeredContexts.get(context);
if (!isEqual(opts, registeredOptions)) {
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Testing equality of all options to catch the possibility of rule types registering with the same context but different useEcs or useLegacyAlerts flags.

@@ -326,6 +361,14 @@ export class AlertsService implements IAlertsService {
) {
this.options.logger.info(`Installing index template ${indexPatterns.template}`);

const indexMetadata: Metadata = {
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Added this to align with metadata that rule registry is outputting

/**
* Optional secondary alias to use. This alias should not include the namespace.
*/
secondaryAlias?: string;
Copy link
Contributor Author

Choose a reason for hiding this comment

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

Added to allow detection alerts to specify siem.signals alias

@ymao1 ymao1 self-assigned this Feb 28, 2023
@ymao1 ymao1 added Feature:Alerting Team:ResponseOps Label for the ResponseOps team (formerly the Cases and Alerting teams) Feature:Alerting/Alerts-as-Data Issues related to Alerts-as-data and RuleRegistry v8.8.0 release_note:skip Skip the PR/issue when compiling release notes labels Feb 28, 2023
@ymao1 ymao1 marked this pull request as ready for review February 28, 2023 21:01
@ymao1 ymao1 requested review from a team as code owners February 28, 2023 21:01
Copy link
Contributor

@mikecote mikecote left a comment

Choose a reason for hiding this comment

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

Great work!! I finished reviewing the code and have the comments below from our discussion. I will move onto testing the PR 👍 🎉

@@ -16,6 +16,7 @@ import {
DEFAULT_ALERTS_ILM_POLICY,
Copy link
Contributor

Choose a reason for hiding this comment

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

I'm not sure if this is a concern given we can do a diff down the line but:

question: do we have something in place to prevent registrationContext from differentiating to what is provided to the rule type via alerts object? (thinking in case a new rule type is in the works somewhere or soemthing.. until we remove the need for registrationContext)

Copy link
Contributor Author

Choose a reason for hiding this comment

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

If a new rule type is added to the rule registry without also being registered with the framework and the framework handles the installation the context specific resources, any run of the rule type that generates an alert will throw an error on execution. I believe this should be caught during development but if we want an additional check, I think we could add a check to the initializeIndex function in the rule registry RuleDataService that verifies any context registered with the rule data service is also registered in the framework.

Copy link
Contributor

Choose a reason for hiding this comment

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

I think adding the function within the rule registry's initializeIndex or doing a manual check when switching the framework alerts feature flag will be ok. If it's trivial to do the manual check, maybe we just do that when validating the feature flag PR?

x-pack/plugins/alerting/server/plugin.ts Outdated Show resolved Hide resolved
Copy link
Contributor

@mikecote mikecote left a comment

Choose a reason for hiding this comment

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

Completed my testing this morning, everything looks great! I used observability.metrics for my testing and all the differences were as outlined in the PR. Great work!

@ymao1
Copy link
Contributor Author

ymao1 commented Mar 6, 2023

@elasticmachine merge upstream

@dominiqueclarke dominiqueclarke self-requested a review March 6, 2023 20:45
Copy link
Contributor

@kdelemme kdelemme left a comment

Choose a reason for hiding this comment

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

LGTM! Tested locally with an existing SLO burn rate rule running.

Copy link
Contributor

@dominiqueclarke dominiqueclarke left a comment

Choose a reason for hiding this comment

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

Uptime changes LGTM

Copy link
Contributor

@marshallmain marshallmain left a comment

Choose a reason for hiding this comment

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

Detection alerts changes LGTM

@crespocarlos crespocarlos self-requested a review March 7, 2023 09:10
Copy link
Contributor

@crespocarlos crespocarlos left a comment

Choose a reason for hiding this comment

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

LGTM on behalf of Infra Observability UI

@ymao1
Copy link
Contributor Author

ymao1 commented Mar 8, 2023

@elasticmachine merge upstream

@kibana-ci
Copy link
Collaborator

💚 Build Succeeded

Metrics [docs]

Module Count

Fewer modules leads to a faster build time

id before after diff
apm 1410 1411 +1
observability 550 549 -1
securitySolution 3751 3750 -1
triggersActionsUi 547 548 +1
total -0

Public APIs missing comments

Total count of every public API that lacks a comment. Target amount is 0. Run node scripts/build_api_docs --plugin [yourplugin] --stats comments for more detailed information.

id before after diff
@kbn/alerts-as-data-utils 12 14 +2
alerting 486 487 +1
total +3

Async chunks

Total size of all lazy-loaded chunks that will be downloaded as the user navigates the app

id before after diff
apm 3.8MB 3.8MB +8.0B
observability 1.1MB 1.1MB +158.0B
triggersActionsUi 1.4MB 1.4MB +10.0B
total +176.0B

Page load bundle

Size of the bundles that are downloaded on every page load. Target size is below 100kb

id before after diff
alerting 41.8KB 41.8KB +66.0B
triggersActionsUi 80.6KB 80.8KB +130.0B
total +196.0B
Unknown metric groups

API count

id before after diff
@kbn/alerts-as-data-utils 12 14 +2
alerting 497 503 +6
total +8

ESLint disabled line counts

id before after diff
alerting 77 78 +1
securitySolution 431 433 +2
total +3

Total ESLint disabled count

id before after diff
alerting 78 79 +1
securitySolution 508 510 +2
total +3

History

To update your PR or re-run it, just comment with:
@elasticmachine merge upstream

cc @ymao1

@ymao1 ymao1 merged commit 64a7df7 into elastic:main Mar 8, 2023
@kibanamachine kibanamachine added the backport:skip This commit does not require backporting label Mar 8, 2023
@ymao1 ymao1 deleted the alerting/faad-resources-component branch March 8, 2023 14:51
bmorelli25 pushed a commit to bmorelli25/kibana that referenced this pull request Mar 10, 2023
…ponent templates, index templates and concrete write index to framework for alerts-as-data (elastic#151792)

Resolves elastic#151697

## Summary

In a previous [PR](elastic#145581) we
started installing a context-specific component templates, index
templates and concrete write indices for framework alerts as data when
the `xpack.alerting.enableFrameworkAlerts` config flag is set to true.
In that PR we used a different naming pattern than what is used by the
rule registry for those resources. In this PR, we are aligning the
naming of these resources with the rule registry and installing these
resources on alerting plugin setup when `enableFrameworkAlerts: true`.
If the flag is set to false, the rule registry will continue to handle
this resource installation.

In this PR we are doing the following:
* Registering all rules currently registered with the rule registry with
the alerting framework. This registration allows the alerting framework
to build context specific component templates. Because this PR only
addresses resource installation, rules will continue to be registered
with the rule registry.
* When `enableFrameworkAlerts: true`:
* The framework installs the context specific component template with
the following naming convention: `.alerts-{context}.alerts-mappings`.
This matches what the rule registry currently installs so the transition
should be seamless
* The framework installs the context specific index template for the
`default` space with the following name:
`.alerts-{context}.alerts-default-index-template`. Space awareness will
be addressed in a followup PR. This matches the current rule registry
naming.This index template will reference
    (1) ECS component template (if `useEcs: true`), 
    (2) context-specific component template, 
    (3) legacy alert component template and 
    (4) framework component template
where the legacy alert component template + framework component template
= technical component template (from the rule registry).
* The framework creates or updates the concrete write index for the
`default` space with the naming convention:
`.internal.alerts-{context}.alerts-default-000001`. Space awareness will
be addressed in a followup PR. This matches the current rule registry
naming.
* The installation of the index template & write index differs from the
rule registry in that it occurs on alerting plugin start vs the first
rule run.
* We modified the rule registry resource installer to skip installation
of these resources when `enableFrameworkAlerts: true`. In addition, it
will wait for the alerting resource installation promise so if a rule
runs before its resources are fully initialized, it will wait for
initialization to complete before writing.

## To Verify

The following rule registry contexts are affected:
`observability.apm`
`observability.logs`
`observability.metrics`
`observability.slo`
`observability.uptime`
`security`

For each context, we should verify the following:

`Note that if your rule context references the ECS mappings, there may
be differences in those mappings between main and this branch depending
on whether you're running main with enableFrameworkAlerts true or false.
These differences are explained in the summary of this prior PR:
elastic#150384 but essentially we're
aligning with the latest ECS fields. In the instructions, I suggest
running main with enableFrameworkAlerts: true to minimize the
differences caused by ECS changes`

**While running `main` with `enableFrameworkAlerts: true`:**
1. Get the context specific component template `GET
_component_template/.alerts-{context}.alerts-mappings`
2. Create rule for this context that creates an alert and then
3. Get the index template `GET
_index_template/.alerts-{context}.alerts-default-index-template`
4. Get the index mapping for the concrete index: `GET
.internal.alerts-{context}.alerts-default-000001/_mapping`

**While running this branch with `xpack.alerting.enableFrameworkAlerts:
true` (with a fresh ES instance):**
5. Get the context specific component template `GET
_component_template/.alerts-{context}.alerts-mappings`
6. Get the index template `GET
_index_template/.alerts-{context}.alerts-default-index-template`
7. Get the index mapping for the concrete index: `GET
.internal.alerts-{context}.alerts-default-000001/_mapping`
Note that you should not have to create a rule that generates alerts
before seeing these resources installed.

**Compare the component templates**
Compare 1 and 5. The difference should be:
* component template from this branch should have `_meta.managed: true`.
This is a flag indicating to the user that these templates are system
managed and should not be manually modified.

**Compare the index templates**
Compare 3 and 6. The differences should be:
* index template from this branch should have `managed: true` in the
`_meta` fields
* index template from this branch should not have a `priority` field.
This will be addressed in a followup PR
* index template from this branch should be composed of
`.alerts-legacy-alert-mappings` and `.alerts-framework-mappings` instead
of `.alerts-technical-mappings` but under the hood, these mappings are
equivalent.

**Compare the index mappings**
Compare 4 and 7. The difference should be:
* index mappings from this branch should have `_meta.managed: true`.

### Verify that installed resources templates work as expected
1. Run this branch on a fresh ES install with
`xpack.alerting.enableFrameworkAlerts: true`.
2. Create a rule in your context that generates alerts.
3. Verify that there are no errors during rule execution.
4. Verify that the alerts show up in your alerts table as expected.
5. (For detection rules only): Run this branch with
`xpack.alerting.enableFrameworkAlerts: true` and verify rules in a
non-default space continue to create resources on first rule run and run
as expected.
6. (For detection rules only): Run this branch with
`xpack.alerting.enableFrameworkAlerts: true` and verify rule preview
continue to work as expected

### Verify that installed resources templates work with existing rule
registry resources.

1. Run `main` or a previous version and create a rule in your context
that generates alerts.
2. Using the same ES data, switch to this branch with
`xpack.alerting.enableFrameworkAlerts: false` and verify Kibana starts
with no rule registry errors and the rule continues to run as expected.
3. Using the same ES data, switch to this branch with
`xpack.alerting.enableFrameworkAlerts: true` and verify Kibana starts
with no alerting or rule registry errors and the rule continues to run
as expected.
4. Verify the alerts show up on the alerts table as expected.
5. (For detection rules only): Run this branch with
`xpack.alerting.enableFrameworkAlerts: true` and verify rules in a
non-default space continue to create resources on first rule run and run
as expected.
6. (For detection rules only): Run this branch with
`xpack.alerting.enableFrameworkAlerts: true` and verify rule preview
continue to work as expected

---------

Co-authored-by: Kibana Machine <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
backport:skip This commit does not require backporting Feature:Alerting/Alerts-as-Data Issues related to Alerts-as-data and RuleRegistry Feature:Alerting release_note:skip Skip the PR/issue when compiling release notes Team:APM All issues that need APM UI Team support Team:ResponseOps Label for the ResponseOps team (formerly the Cases and Alerting teams) Team:Uptime - DEPRECATED Synthetics & RUM sub-team of Application Observability v8.8.0
Projects
None yet