[ https://issues.apache.org/jira/browse/CASSANDRA-14459?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16551064#comment-16551064 ]
Joseph Lynch commented on CASSANDRA-14459: ------------------------------------------ Ok based on the feedback I've moved the latency probes into a scheduled executor that runs every second and sends a single message to one node needing latency probes. If no probes are needed then they are not sent. I've gone a step further and posted a follow up commit which replaces all the uses of histograms with a simpler (and 10x more performant) EMA implementation. The second patch is very optional but I know that some folks have mentioned they don't like all the garbage that the Histograms create so I fixed it. ||trunk||trunk+EMA|| |[patch|https://github.com/apache/cassandra/compare/trunk...jolynch:CASSANDRA-14459]|[patch|https://github.com/apache/cassandra/compare/trunk...jolynch:CASSANDRA-14459-EMA]| |[unit tests|https://github.com/apache/cassandra/compare/trunk...jolynch:CASSANDRA-14459#diff-f1baf62fd510947bf60e9a9def69810b] [!https://circleci.com/gh/jolynch/cassandra/tree/CASSANDRA-14459.png?circle-token= 1102a59698d04899ec971dd36e925928f7b521f5!|https://circleci.com/gh/jolynch/cassandra/tree/CASSANDRA-14459]|[unit tests|https://github.com/apache/cassandra/compare/trunk...jolynch:CASSANDRA-14459-EMA#diff-f1baf62fd510947bf60e9a9def69810b] [microbenchmark|https://github.com/apache/cassandra/compare/trunk...jolynch:CASSANDRA-14459-EMA#diff-f48455469e3cd6a2e225e3cb2b389d93] [!https://circleci.com/gh/jolynch/cassandra/tree/CASSANDRA-14459.png?circle-token= 1102a59698d04899ec971dd36e925928f7b521f5!|https://circleci.com/gh/jolynch/cassandra/tree/CASSANDRA-14459-EMA]| > DynamicEndpointSnitch should never prefer latent nodes > ------------------------------------------------------ > > Key: CASSANDRA-14459 > URL: https://issues.apache.org/jira/browse/CASSANDRA-14459 > Project: Cassandra > Issue Type: Improvement > Components: Coordination > Reporter: Joseph Lynch > Assignee: Joseph Lynch > Priority: Minor > Fix For: 4.x > > > The DynamicEndpointSnitch has two unfortunate behaviors that allow it to > provide latent hosts as replicas: > # Loses all latency information when Cassandra restarts > # Clears latency information entirely every ten minutes (by default), > allowing global queries to be routed to _other datacenters_ (and local > queries cross racks/azs) > This means that the first few queries after restart/reset could be quite slow > compared to average latencies. I propose we solve this by resetting to the > minimum observed latency instead of completely clearing the samples and > extending the {{isLatencyForSnitch}} idea to a three state variable instead > of two, in particular {{YES}}, {{NO}}, {{MAYBE}}. This extension allows > {{EchoMessages}} and {{PingMessages}} to send {{MAYBE}} indicating that the > DS should use those measurements if it only has one or fewer samples for a > host. This fixes both problems because on process restart we send out > {{PingMessages}} / {{EchoMessages}} as part of startup, and we would reset to > effectively the RTT of the hosts (also at that point normal gossip > {{EchoMessages}} have an opportunity to add an additional latency > measurement). > This strategy also nicely deals with the "a host got slow but now it's fine" > problem that the DS resets were (afaik) designed to stop because the > {{EchoMessage}} ping latency will count only after the reset for that host. > Ping latency is a more reasonable lower bound on host latency (as opposed to > status quo of zero). -- This message was sent by Atlassian JIRA (v7.6.3#76005) --------------------------------------------------------------------- To unsubscribe, e-mail: commits-unsubscr...@cassandra.apache.org For additional commands, e-mail: commits-h...@cassandra.apache.org