Monitor lambda execution to notify failures

[mahalakshme]

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