On Tue, 18 Sep 2012, Tren Blackburn wrote: > Hi List; > > I've been working with ceph 0.51 lately, and have noticed this for a > while now, but it hasn't been a big enough issue for me to report. > However today I'm turning up a 192 OSD cluster, and 30 seconds per OSD > adds up pretty quick. For some reason it's taking 30 seconds between > checking the OSD for a pre-existing store: > > 2012-09-18 13:53:28.400590 7fe895d25780 -1 filestore(/var/ceph/disk11) > could not find 23c2fcde/osd_superblock/0//-1 in index: (2) No such > file or directory > > And then actually creating the new store: > > 2012-09-18 13:53:57.371396 7fe895d25780 -1 created object store > /var/ceph/disk11 journal /dev/mapper/vg-journal.disk11 for osd.34 fsid > bca82801-04d7-402e-917f-8023a4b161a8 > 2012-09-18 13:53:57.371449 7fe895d25780 -1 auth: error reading file: > /var/ceph/disk11/keyring: can't open /var/ceph/disk11/keyring: (2) No > such file or directory > 2012-09-18 13:53:57.371527 7fe895d25780 -1 created new key in keyring > /var/ceph/disk11/keyring > > I can provide many examples as I'm watching it slowly plod through > currently. Also, the horse power of the server makes no difference. > The servers in question here are dual E5-2600's with 96GB ram and > 12x2TB drives. > > What information can I provide to help debug this? Or is this an > already known issue? If you add -x to the #!/bin/bash line in mkcephfs we can see exactly which command(s) are going slow... sage > > Thanks in advance! > > t. > -- > To unsubscribe from this list: send the line "unsubscribe ceph-devel" in > the body of a message to majordomo@xxxxxxxxxxxxxxx > More majordomo info at http://vger.kernel.org/majordomo-info.html > > -- To unsubscribe from this list: send the line "unsubscribe ceph-devel" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
