We have an enterprise customer using our extension on Citrix. 2 of 40 users are encountering an "Internal error opening backing store for indexedDB.open" error on all IndexedDB calls.
Previously when seeing flakiness for other users, e.g., "
Failed to read large IndexedDB value" errors (which can indicate a missing blob), we added code to automatically mitigate by deleting the DB so it could be recreated
Our current guess for these 2 users is that their Chrome profile is becoming corrupted. Having the Citrix admin delete the IndexedDB directory on the filesystem temporarily fixes the problem. But the problems have returned, e.g., when coming in the next day or after a lunch break
Any ideas of how to automatically mitigate in our extension code? We suspect it's a Citrix profile syncing/streaming issue, so we're following up with the Citrix admin on their Profile Management settings
Thanks,
Todd
Resources: