curator_cli forcemerge exception #1704
Assignees
Labels
Comments
untergeek
added a commit
to untergeek/curator
that referenced
this issue
Apr 1, 2025
Replace:
```
action.do_singleton_action(dry_run=ctx.obj['dry_run'])
```
With:
```
try:
action.do_singleton_action(dry_run=ctx.obj['dry_run'])
except NoIndices: # Speficically to address elastic#1704
action.logger.info('No indices in list after filtering. Skipping action.')
```
untergeek
added a commit
to untergeek/curator
that referenced
this issue
Apr 1, 2025
This here addresses the problem in elastic#1704 ```diff --- a/curator/cli_singletons/object_class.py +++ b/curator/cli_singletons/object_class.py @@ -262,7 +262,11 @@ class CLIAction: action_obj.do_dry_run() else: action_obj.do_action() - + except NoIndices: # Speficically to address elastic#1704 + if not self.ignore: + self.logger.critical('No indices in list after filtering. Exiting.') + sys.exit(1) + self.logger.info('No indices in list after filtering. Skipping action.') except Exception as exc: self.logger.critical( 'Failed to complete action: %s. Exception: %s', self.action, exc ```
untergeek
added a commit
to untergeek/curator
that referenced
this issue
Apr 1, 2025
This also demonstrates what happened with elastic#1704 and that this resolves it. Two extra tests. One to test with `--ignore_empty_list` and one without that flag.
untergeek
added a commit
to untergeek/curator
that referenced
this issue
Apr 1, 2025
8.0.21 (1 April 2025) --------------------- **Bugfix Release** As was reported in elastic#1704, the ``--ignore_empty_list`` option was not being respected. Code changes were made to each singleton that uses the ``--ignore_empty_list`` option to ensure that the option is respected. **Changes** * Fix ``--ignore_empty_list`` option to be respected in all singletons. * Add debug message to IndexList class when an empty list condition is encountered. * Dependency version bump: ``es_client`` to ``8.17.5``
|
I finally took the time to resolve this. A release should be out later today. |
untergeek
added a commit
that referenced
this issue
Apr 1, 2025
* Add debug log message on empty index list * Add NoIndices try/except catch block This here addresses the problem in #1704 ```diff --- a/curator/cli_singletons/object_class.py +++ b/curator/cli_singletons/object_class.py @@ -262,7 +262,11 @@ class CLIAction: action_obj.do_dry_run() else: action_obj.do_action() - + except NoIndices: # Speficically to address #1704 + if not self.ignore: + self.logger.critical('No indices in list after filtering. Exiting.') + sys.exit(1) + self.logger.info('No indices in list after filtering. Skipping action.') except Exception as exc: self.logger.critical( 'Failed to complete action: %s. Exception: %s', self.action, exc ``` * Add tests for forcemerge to test the fix This also demonstrates what happened with #1704 and that this resolves it. Two extra tests. One to test with `--ignore_empty_list` and one without that flag. * Version bump for es_client Now at 8.17.5 * Version bump to 8.0.21 * Update Changelog 8.0.21 (1 April 2025) --------------------- **Bugfix Release** As was reported in #1704, the ``--ignore_empty_list`` option was not being respected. Code changes were made to each singleton that uses the ``--ignore_empty_list`` option to ensure that the option is respected. **Changes** * Fix ``--ignore_empty_list`` option to be respected in all singletons. * Add debug message to IndexList class when an empty list condition is encountered. * Dependency version bump: ``es_client`` to ``8.17.5``
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
This issue has already been discussed on the Elastic forum: https://discuss.elastic.co/t/curator-forcemerge-exception/354245
To submit a bug or report an issue
I've discovered that the current version of curator_cli (version 8.0.10) will throw an Exception on forcemerge when the index in question is already at/below the requested number of segments per shard.
Expected Behavior
The curator_cli program should filter out indices as candidates to forcemerge when they are already at/below the requested number of segments per shard.
Actual Behavior
The curator_cli program throws an exception and exits with a nonzero exit code when the specified index already has the requested number of segments per shard.
Steps to Reproduce the Problem
Specifications
Context (Environment)
I have a script that runs nightly to forcemerge cold indices down to a single segment. It depends on script exit codes to check for errors/issues with Elastic index processed. Per the discussion in the Elastic forum (see the link at the beginning of this issue), I'm putting in a feature request to ask that curator_cli not error out on indices that are already at or below the requested number of segments per shard.
The text was updated successfully, but these errors were encountered: