[
https://issues.apache.org/jira/browse/CASSANDRA-9206?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14500227#comment-14500227
]
Brandon Williams commented on CASSANDRA-9206:
---------------------------------------------
I did some digging as to why the probability code exists, and wasn't able to
find much. It came over the wall with facebook, and there's no mention of it
in the scuttlebutt paper (nor seeds at all, nor probabilistically gossiping
with unreachable members) so I'm not sure what the original reasoning was for
it.
> Remove seed gossip probability
> ------------------------------
>
> Key: CASSANDRA-9206
> URL: https://issues.apache.org/jira/browse/CASSANDRA-9206
> Project: Cassandra
> Issue Type: Improvement
> Reporter: Brandon Williams
> Assignee: Brandon Williams
> Fix For: 2.1.5
>
> Attachments: 9206.txt
>
>
> Currently, we use probability to determine whether a node will gossip with a
> seed:
> {noformat}
> double probability = seeds.size() / (double)
> (liveEndpoints.size() + unreachableEndpoints.size());
> double randDbl = random.nextDouble();
> if (randDbl <= probability)
> sendGossip(prod, seeds);
> {noformat}
> I propose that we remove this probability, and instead *always* gossip with a
> seed. This of course means increased traffic and processing on the seed(s),
> but even a 1000 node cluster with a single seed will only put ~1000 messages
> per second on the seed, which is virtually nothing. Should it become a
> problem, the solution is simple: add more seeds. Since seeds will also
> always gossip with each other, this effectively gives us a poor man's
> spanning tree, with the only cost being removing a few lines of code, and
> should greatly improve our gossip convergence time, especially in large
> clusters.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
