DB2 is running out of space:
/dev/xvda3 48654148 41411196 4731592 90% /
Notice the regular increase since its inception:
https://admin.fedoraproject.org/cacti/graph.php?action=view&rra_id=all&local_graph_id=269
There are a few possible causes for this but in the worst case scenario, koji is simply growing that quickly and there's little that can be done about it. One thing to be noticed is the dump size is no where near the size of the running db. Its possible vacuum's are not taking place properly. I'm investigating this but if it is just a fact that the db is this big now, I'll create another virtual machine with more storage.
This went off almost without a hitch. At one point the throughput according to iptraf went down to 15kbps. Thats over a lan link. Something strange was going on with, I suspect, either the netapp or iscsi. iptraf would show very slow transfer rates at the same time iowait would go to about 100%. Additionally the xen guest itself was showing great throughput, even with all other hosts on the dom0 shut down. Something to keep an eye on. After I migrated the new db2 from xen1 (where I was staging the copy) to xen6 everything seemed fine.
Login to comment on this ticket.