It may not have actually finished at that point. Though, according to JMX, both compactions of each CF had completed, so I assumed it was done.
On Mon, May 31, 2010 at 11:29 AM, Stu Hood <stu.h...@rackspace.com> wrote: > Did you watch in the logs to confirm that repair had actually finished? The > `nodetool repair` call is not blocking before 0.6.3 (unreleased): see > https://issues.apache.org/jira/browse/CASSANDRA-1090 > > -----Original Message----- > From: "James Golick" <jamesgol...@gmail.com> > Sent: Sunday, May 30, 2010 3:43am > To: cassandra-u...@incubator.apache.org > Subject: Inconsistency even after nodetool repair? > > I was getting a ton of these in my logs: > > INFO [pool-1-thread-77] 2010-05-30 03:39:18,263 StorageProxy.java (line > 499) DigestMismatchException: Mismatch for key 142667/everything > (5546784193bbe4a4e066dc7fc142f589 vs 03edb87cd131a614d8256ddfdc50dd17) > > The nodes were all getting really overloaded, presumably from the > read-repair, since everything else looked healthy. So, I ran nodetool > repair > on both of our nodes, but even after running the repairs, I'm still seeing > the same thing in the logs and high load on the nodes. > > Does that sound right? > > >
