Currently, we use a local node process to perform bulk generation of thumbnails, but this requires a lot of setup. Also, we do not have validation process to check if thumbnail exists before regeneration.
Based on code-review and dev of #169 card, we have come up with following action-items that need to be taken up for avni-media:
Capture errors in cloud-watch logs which are identifiable easily (Via tagging as a different topic or so)
Generate Alerts once a day on lambda failures and notifiy via Email
To Recover from failures by selectively reprocessing media
Mark media entities as failures in Lambda logs for quick retrieval of errored media content urls
Introduce APIs to trigger lambda functions for specified urls (Org/uuids) to regenerate thumbnails only if not present
To brute force recover for an entire org
Introduce API to invoke bulk regeneration of thumbnails for an entire Organisation only if not present
Currently, we use a local node process to perform bulk generation of thumbnails, but this requires a lot of setup. Also, we do not have validation process to check if thumbnail exists before regeneration.
Based on code-review and dev of #169 card, we have come up with following action-items that need to be taken up for avni-media: