- Issue created by @trackleft2
- Merge request !14Close #3474301 Add better error handling for ConfigSnapshotServiceProvider. β (Merged) created by trackleft2
- Issue was unassigned.
- Status changed to Needs review
7 months ago 12:00am 14 September 2024 - Status changed to RTBC
7 months ago 11:29pm 17 September 2024 Thanks for working on this @trackleft2 !
Tested out a few aspects of this:Verified that services are still being registered:
$ drush devel:services ... - config_snapshot.config_sync.module.pathauto - config_snapshot.config_sync.module.redirect - config_snapshot.config_sync.module.search - config_snapshot.config_sync.module.shortcut ...
Verified that errors are logged if unhandled exceptions occur elsewhere in provisioning the storage:
Error initializing config storage: Failed to retrieve config storage. (And - separate error forced): Error initializing config storage: An unrelated error somewhere else occurred.
Verified that errors are logged in the case of any of
snapshotSet
,extensionType
, andextensionName
being absent from malformed snapshots (in separate cases):Error processing config ID config_snapshot.snapshot..module.search: Invalid snapshot data for config ID config_snapshot.snapshot..module.search Error processing config ID config_snapshot.snapshot.config_sync..search: Invalid snapshot data for config ID config_snapshot.snapshot.config_sync..search Error processing config ID config_snapshot.snapshot.config_sync.module.: Invalid snapshot data for config ID config_snapshot.snapshot.config_sync.module.
The only thing I noted was that the PHP error_log is used, rather than the Drupal log. But given that this is the service discovery phase, I'm not sure if using the Drupal log channel would even make sense. Perhaps that's why you selected the PHP error log to begin with?
-
joegraduate β
committed 847676be on 2.0.x authored by
trackleft2 β
Issue #3474301 by trackleft2, tadean: Add better error handling for...
-
joegraduate β
committed 847676be on 2.0.x authored by
trackleft2 β
- Status changed to Fixed
7 months ago 12:10am 18 September 2024 Automatically closed - issue fixed for 2 weeks with no activity.