riak hangs after after a lot of map reduce queries
jan.metzner at gmail.com
Thu Nov 18 05:49:35 EST 2010
we had a testrun last night which looked ok until the morning, after
12 hours of different map reduce queries
the server cpu load abrupt rose and the riak server did not respond to
riak-admin status said that the node was down but the processes were
running. The beam process consumed
most of the cpu. There were no error in the logs so far.
Our setup is one node on debian with the normal riak 0.13 debain
package installed. Backend is bitcask and
we only disabled the js query cache.
Any suggestions why this can happen and how to avoid this?
More information about the riak-users