Github user joshelser commented on a diff in the pull request:
https://github.com/apache/accumulo/pull/220#discussion_r102779143
--- Diff:
server/master/src/main/java/org/apache/accumulo/master/replication/RemoveCompleteReplicationRecords.java
---
@@ -216,4 +220,13 @@ protected long removeRowIfNecessary(BatchWriter bw,
SortedMap<Key,Value> columns
return recordsRemoved;
}
+
+ /*
+ * Metrics calls snapshot. Largest time stamp will be found for each
snapshot then reset.
+ */
+ private void collectLatency(Long createdTime) {
+ long latency = System.currentTimeMillis() - createdTime;
--- End diff --
> I am not sure, but I think this createdTime may come from a persisted
source
I believe createdTime comes from the TabletServer, likely calling
currentTimeMillis(). As long as we're reporting non-negative latencies, I think
it will be fine (e.g. take the max of 0 and the computed latency).
The createdTime is coming from a remote JVM (the tserver), so I dont' think
there's a need to use nanoTime here for the monotonically increasing time. We
can just fudge it.
---
If your project is set up for it, you can reply to this email and have your
reply appear on GitHub as well. If your project does not have this feature
enabled and wishes so, or if the feature is enabled but not working, please
contact infrastructure at [email protected] or file a JIRA ticket
with INFRA.
---
