We started seeing for the first time errors that look like this last night after things were running smoothly for a couple days after switching to 0.6-rc1:
ERROR [pool-1-thread-60] 2010-04-13 04:58:11,806 Cassandra.java (line 1492) Internal error processing insert We had not made any other changes to cassandra server or client config in days. Operationally we did try to bootstrap two new nodes, one of which failed after timing out. The other thing we did was a cleanup operation. Is it possible somehow any of those operations/events would somehow get things into a bad state? If not, any other ideas of what to look for as the root cause? Overall, the cluster seemed relatively healthy with the exception that one of our nodes which happens to be the seed node has been having excessively high network delays, hovering around 30-40 ms although it's been in that state for more than a week. Also a side question I was curious about, is the state of a boostrap or cleanup operation maintained after restart? thanks, Scott
