What is your GC_GRACE_SECONDS ? What kind repair option do you use for nodetool repair on a keyspace ? Did you start the repair on one node? did you use nodetool repair -pr ? or just "nodetool repair keyspace" ? How many nodetool repair processes do you use on the nodes?
On Sun, Jul 30, 2017 at 10:53 PM, Jeff Jirsa <jji...@apache.org> wrote: > > > On 2017-07-27 21:36 (-0700), Mitch Gitman <mgit...@gmail.com> wrote: > > Now, the particular symptom to which that response refers is not what I > was > > seeing, but the response got me thinking that perhaps the failures I was > > getting were on account of attempting to run "nodetool repair > > --partitioner-range" simultaneously on all the nodes in my cluster. These > > are only three-node dev clusters, and what I would see is that the repair > > would pass on one node but fail on the other two. > > > Running nodetool repair --partitioner-range simultaneously on all nodes in > the cluster will indeed be a problem, and the symptoms will vary widely > based on node state / write load / compaction load. This is one of the > times when the right answer is "don't do that" until the project comes up > with a way to prevent you from doing it in order to protect you from > yourself. > > > > > --------------------------------------------------------------------- > To unsubscribe, e-mail: user-unsubscr...@cassandra.apache.org > For additional commands, e-mail: user-h...@cassandra.apache.org > >
