Well grepping the logs for a file resulted in this: During startup that file was opened OK. During runtime the compactionexecutor tried to compact more files that included this one but threw and exception for another one that is first in that list. So I do not have more info. In one particular day I cleared the logs. But the file I am talking about opened ok during startup.(I have that in the log BATCHOPEN section) So not much more info on this. Any ideeas on what to do ? Decommission , delete the /data dir and readd to cluster? For that keyspace I have a replication factor of 3 so data should be available on other nodes.
On Wednesday, March 26, 2014 12:36 AM, Robert Coli <rc...@eventbrite.com> wrote: On Tue, Mar 25, 2014 at 3:24 PM, James Rothering <jrother...@codojo.me> wrote: What message would be on the log, or not on the log, what would you be grep-ing for? You'd be looking for the broken SSTable to appear in either a MemtableFlusher line or a compaction line. Also, why were you interested in startup procedure? Is there some "best practices: for that which could impact this? Startup procedure? I was asking details about shutdown procedure... ? =Rob
