On Wed, Mar 28, 2012 at 7:49 AM, Whitney Sorenson <[email protected]> wrote: > This happens reliably when we run large jobs against our cluster which > perform many reads and writes, but it does not always happen on the > same keys. >
Interesting. Anything you can figure about a particular key if you dig in some? For example, could the key be at a region boundary or was the region moving/splitting at the time? Once you've found a bad key, does it stay bad? If so, that'd be useful trying to debug. It may have been fixed already but I've not done the research. St.Ack
