juju backup fails due to probable mongodb corruption

[Steven Parker]

Back up of juju DB fails

This juju controller is hosted on a bootstack cloud.

After upgrade from 2.5.x to 2.6.3 on an HA version of juju.

ERROR while creating backup archive: while dumping juju state database: error dumping databases: error executing "/usr/lib/juju/mongo3.2/bin/mongodump": 2019-06-25T19:32:15.460+0000 Failed: error counting admin.system.users: node is recovering;

This prevents juju from making a backup of the database and potentially means that our HA mongodb may be at risk.

Please let me know what other information you may require and I'll be happy to provide it.

[Tim Penhey (thumper)]

I'm pretty sure this is because the backup command went to a controller that was not the mongo master. This needs to be addressed at the API level for backup. Effectively Juju needs to tell the client to reconnect to a particular API server so it is talking to the mongo master.

Trying to do a mongo dump on the secondaries or recovering nodes is a problem.

[Canonical Juju QA Bot (juju-qa-bot)]

This bug has not been updated in 2 years, so we're marking it Low importance. If you believe this is incorrect, please update the importance.