ifesdjeen commented on code in PR #3274:
URL: https://github.com/apache/cassandra/pull/3274#discussion_r1598405048
##########
src/java/org/apache/cassandra/db/MutationVerbHandler.java:
##########
@@ -41,6 +43,13 @@ private void failed()
public void doVerb(Message<Mutation> message)
{
+ if (approxTime.now() > message.expiresAtNanos())
Review Comment:
My logic here was that:
a) approx time is _really_ bad only for metrics (but even then, only for
idle/very fast clusters)
b) relative time between replicas is controlled by something like NTP
anyways, and operators should know their drift
c) diff between approx and exact time is up to 2 micros, which means that
for cross-node it'll be up to 4 micros; if we time out request 4 micros too
early, I think it is more than fine, we should not operate that close to
timeouts anyways
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
