See https://issues.apache.org/jira/browse/CASSANDRA-10414 for an overview of why vnodes are currently less efficient for secondary index queries.
On Tue, Sep 29, 2015 at 12:45 PM, Robert Coli <rc...@eventbrite.com> wrote: > On Tue, Sep 15, 2015 at 7:44 AM, Tom van den Berge < > tom.vandenbe...@gmail.com> wrote: > >> Read queries on a secondary index are somehow causing an excessively high >> CPU load on all nodes in my DC. >> > ... > >> What really surprised me is that executing a single query on this >> secondary index makes the "Local read count" in the cfstats for the index >> go up with almost 200000! When doing the same query on one of my "good" >> nodes, it only increases with a small number, as I would expect. >> >> Could it be that the use of vnodes is causing these problems? >> > > I am not too surprised to hear of this performance degradation. > > Yes, it is relatively likely to be the use of vnodes which is causing this > problem. You could verify by having one of your nodes use 64 vnodes instead > of the default 256... you will get less even distribution with current > vnode random allocation, but you will pay less of a penalty for having > multiple ranges... > > =Rob > > > -- Tyler Hobbs DataStax <http://datastax.com/>
