Issues with high node load and very slow response

Chaim Solomon chaim at
Sat Jul 26 23:54:22 EDT 2014


I've been having issues for some days with the cluster I am running (or now
- trying to run). The last major operation was going through all items of a
bucket and doing a read/update operation on them from Python code - nothing
fancy, just modifying some data.

I got issues with one node being VERY busy - 2 CPUs fully loaded - and the
requests were timing out or at least VERY slow.
I took the node out of the cluster and added another node - then after
clearing the node brought it back. After some days of waiting for AAE and
the migration of data to finish, I am still having one node at 2 CPUs at
100% and one node with one CPU at 100% and the requests are still VERY
slow. I am getting a lot of these in the console.log:

2014-07-26 23:37:33.459 [error] <0.7427.266>@yz_kv:index:206 failed to
index object {<<"linkmeta">>,<<"...">>} with error {"Failed to index
docs",{error,req_timedout}} because

And on almost all other nodex I am getting a lot of these:
2014-07-26 23:50:58.796 [info] <0.12268.221>@yz_kv:should_handoff:157
waiting for bucket types prefix and indexes to agree between '
riak at' and 'riak at'

How can I get the cluster to work again?

It is a 2.0.0rc1 cluster.

Chaim Solomon
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>

More information about the riak-users mailing list