On Mon, 20 Feb 2012 14:34:15 -0700 Stephen John Smoogen <smooge@xxxxxxxxx> wrote: > My one concern is that both backup02 and fedorapeople can have a large > IO needs and we can have the starvation issues we had last year with > backup02... where it would mysteriously crash when it was doing lvm > items and the other ibiblio boxes were having a load. > > How do the other ibiblio boxes look like for space or reallocation of > disks? > Not fantastic, I'm afraid. Do we know that lvm issue was related to disk io bandwidth starvation? I only ask b/c processor and mem are not really constricted here. > > Kevin suggested possibly replicating this people## over to the old > > location using rsync so, in the event of unplanned calamity, we > > could continue to offer the service. > > That sounds good.. though slooooow. Sounds like change from day to day on people is only1-3gb. So not ridiculous to rsync -sv _______________________________________________ infrastructure mailing list infrastructure@xxxxxxxxxxxxxxxxxxxxxxx https://admin.fedoraproject.org/mailman/listinfo/infrastructure
