dm log intermittently fails with an error message

[prisamuel]

This occurs especially on the backup job as it does a dm log on each branch on each dot in the cluster to back up it's contents.

priya@Priyas-MBP:~/dmesh/ops/backup$ dm log
Response '{"jsonrpc":"2.0","error":{"code":-32000,"message":"Snapshots of '3c8919fc-27fb-4ca7-5837-5e91ac43cac5' not currently known on server '2a1f51b88403779'","data":null},"id":4037200794235010051}
' yields error Snapshots of '3c8919fc-27fb-4ca7-5837-5e91ac43cac5' not currently known on server '2a1f51b88403779'

[prisamuel]

comments from @lukemarsden
this probably has to do with the asynchronous replication not working
or the fact that API calls to list dots should look at the master's snapshot list

[prisamuel]

Only occurs on a multi node cluster on doing a dm log on a branch which has no commits on it.

[prisamuel]

A very happy backup run just took place from the dothub into a local two node cluster AND into the two node backup cluster. Nice work @lukemarsden @binocarlos A lot of dots that failed to push have now completed successfully.

There was an instance of #336 on backing up zfs-kernel-modules.

[lukemarsden]

This is important enough that we should do a quick 0.3.2 release

[lukemarsden]

Done