[ https://issues.apache.org/jira/browse/CASSANDRA-9318?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14604771#comment-14604771 ]
Aleksey Yeschenko commented on CASSANDRA-9318: ---------------------------------------------- bq. I think you're [Aleksey] confusing replica-side and coordinator-side. There are no separate write and read queues on the coordinator and there probably shouldn't be. I don't think I am. I also don't think that the distinction between replica-side and cordinator-side is very useful here, given the most common case of collocating the two roles (with the exception of Rick). There are no separate queues, but there should eventually be something bounding both reads and writes. And arguably the two are different enough from each other (one is request heavy, the other is response heavy) to be treated differently, and in effect as two separate queues. Otherwise, this discussions sounds like fun, but I'm going to stay out of it until I form a 2.1.x-appropriate opinion on it. Just wanted to clarify some hints-related questions and CASSANDRA-6230 (which, being a 3.0 ticket, is somewhat not immediately relevant). > Bound the number of in-flight requests at the coordinator > --------------------------------------------------------- > > Key: CASSANDRA-9318 > URL: https://issues.apache.org/jira/browse/CASSANDRA-9318 > Project: Cassandra > Issue Type: Improvement > Reporter: Ariel Weisberg > Assignee: Ariel Weisberg > Fix For: 2.1.x, 2.2.x > > > It's possible to somewhat bound the amount of load accepted into the cluster > by bounding the number of in-flight requests and request bytes. > An implementation might do something like track the number of outstanding > bytes and requests and if it reaches a high watermark disable read on client > connections until it goes back below some low watermark. > Need to make sure that disabling read on the client connection won't > introduce other issues. -- This message was sent by Atlassian JIRA (v6.3.4#6332)