various full sync status metrics are not correctly updated in Riak EE 1.4.8

[cipy]

When starting a full sync from prompt via sudo riak-repl fullsync start, last_fullsync_started is correctly updated, but not if the full sync is enabled and triggered internally by riak

Easiest way to replicate:

• configure FS+RT replication between cluster A and B
• restart cluster A, wait for FS to start running
• check riak-repl status

[cipy]

last_fullsync_duration is set to undefined for internally triggered runs, will be set correctly only for manually triggered starts

[cipy]

fullsyncs_completed doesn't seem to be updated / incremented either

[cipy]

apparently this code path is never hit in 1.4.8, FS+AAE as I cannot find "Fullsync complete" in logs

        MyClusterName = riak_core_connection:symbolic_clustername(),
        lager:info("Fullsync complete from ~s to ~s",
                   [MyClusterName, State#state.other_cluster]),

[cmeiklejohn]

Filed against 1.4.9 milestone.

[tburghart]

As of the time of writing, the string "Fullsync complete from ... to ..." appears in the logs in every scenario I've tried on the 2.0 branch.

[tburghart]

Regarding the status metrics, last_fullsync_started shows a valid timestamp while a full sync is running, but is deliberately reset to undefined when it completes.

A related metric, last_fullsync_completed, is requested in issue #634 explicitly in response to this behavior, and if we want to change the last_fullsync_started behavior it should probably be done in concert with that issue.

As such, there do not appear to be any unresolved portions of this issue remaining in the 2.0 branch.