If solrcloud_manager is being run in a different timezone than the SolrCloud servers, solrcloud_manager may fail to determine that a backup is complete for up to 24 hours.
This is because it records a timestamp locally, just before the backup command is issued, and waits until a SolrCloud-reported timestamp of the last completed backup is greater than that.
Seems like this should be changed to get the last completed backup time before the backup command starts, and use that instead of a local timestamp. (If there isn't one, some timestamp in the 1970s should probably suffice.)
This requires an extra request, but would ensure the timestamps are all from the cluster's perspective.
If solrcloud_manager is being run in a different timezone than the SolrCloud servers, solrcloud_manager may fail to determine that a backup is complete for up to 24 hours.
This is because it records a timestamp locally, just before the backup command is issued, and waits until a SolrCloud-reported timestamp of the last completed backup is greater than that.
Seems like this should be changed to get the last completed backup time before the backup command starts, and use that instead of a local timestamp. (If there isn't one, some timestamp in the 1970s should probably suffice.)
This requires an extra request, but would ensure the timestamps are all from the cluster's perspective.