On Tue, Aug 19, 2014 at 3:36 PM, Gregory Farnum <greg at inktank.com> wrote: > [Re-adding the list] > > On Tue, Aug 19, 2014 at 2:24 PM, Randy Smith <rbsmith at adams.edu> wrote: > > Gregory, > > > > # ceph osd tree > > # id weight type name up/down reweight > > -1 0.2 root default > > -2 0.2 host cs00 > > 0 0.09999 osd.0 up 1 > > 1 0.09999 osd.1 up 1 > > > > The min_size on the ruleset is 1. Here's the relevant portion of my > > crushmap. > > Sorry, by "min_size" I mean the member associated with the pool, not > with the rule. You should be able to see it by looking at the pools > section of "ceph osd dump", among others. I believe that by default it > should be 2, and the full size is 3...which if you have 2 OSDs and > aren't segregating across hosts should be good, Ah, yes, of course. All of the pools were set to size 3 and min_size 2. I've reset those to 2 and 1 since I only have two osds in this test. > but I'm guessing your > pools are using the default rules that will want to segregate across > hosts. > I changed the crushmap to use osd rather than host in the 'rule_replicated' rule which I though would take care of that. I guess I missed something since I'm still seeing the incomplete pgs. > -Greg > Software Engineer #42 @ http://inktank.com | http://ceph.com > > > > > > # buckets > > host cs00 { > > id -2 # do not change unnecessarily > > # weight 0.200 > > alg straw > > hash 0 # rjenkins1 > > item osd.0 weight 0.100 > > item osd.1 weight 0.100 > > } > > root default { > > id -1 # do not change unnecessarily > > # weight 0.200 > > alg straw > > hash 0 # rjenkins1 > > item cs00 weight 0.200 > > } > > > > # rules > > rule replicated_ruleset { > > ruleset 0 > > type replicated > > min_size 1 > > max_size 10 > > step take default > > step chooseleaf firstn 0 type osd > > step emit > > } > > > > This is a single host "cluster" as it's just for testing. > > > > > > > > On Tue, Aug 19, 2014 at 2:44 PM, Gregory Farnum <greg at inktank.com> > wrote: > >> > >> On Tue, Aug 19, 2014 at 1:37 PM, Randy Smith <rbsmith at adams.edu> wrote: > >> > Greetings, > >> > > >> > I'm creating a new ceph cluster for testing and it's reporting "192 > >> > stale+incomplete" pgs. > >> > > >> > `ceph health detail` lists all of the pgs that are stuck. Here's a > >> > representative line. > >> > > >> > pg 2.2c is stuck stale for 3076.510998, current state > >> > stale+incomplete, > >> > last acting [0] > >> > > >> > But when I run `ceph pg 2.2c query`, I get > >> > > >> > Error ENOENT: i don't have pgid 2.2c > >> > > >> > What can I do to clean up the cluster? I have no problems wiping the > >> > system > >> > because it's a new test cluster. > >> > >> Incomplete generally means the PGs aren't mapped to enough OSDs. > >> What's the output of "ceph osd tree" and what is the min_size set to > >> for your pools? > >> -Greg > >> Software Engineer #42 @ http://inktank.com | http://ceph.com > > > > > > > > > > -- > > Randall Smith > > Computing Services > > Adams State University > > http://www.adams.edu/ > > 719-587-7741 > -- Randall Smith Computing Services Adams State University http://www.adams.edu/ 719-587-7741 -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.ceph.com/pipermail/ceph-users-ceph.com/attachments/20140819/078192a6/attachment.htm>
 |