I also have this problem. Is it perhaps possible to block clients entirely if it is not using a specific version of Ceph? BTW, I often stumble upon the cephfs problem: "client failing to respond to capability release", which result in blocked requests aswell. But i'm not entirely sure if you run CephFS. On 09/13/2016 02:44 AM, Christian Balzer wrote: > > Hello, > > On Mon, 12 Sep 2016 19:28:50 -0500 shiva rkreddy wrote: > >> By saying "old clients" did you mean, (a) Client VMs running old Operating >> System (b) Client VMs/Volumes that are in-use for a long time and across >> ceph releases ? Was there any tuning done to fix it? >> > I'm pretty sure he means c) that is: > Clients (hosts) running older versions of Ceph. > > Which is a pretty common thing, both because the Ceph cluster operator may > not be controlling all the clients, there are version dependencies from > things like OpenStack, the Ceph version on the client was installed by > using an OS distro package (usually well aged). > > What I'd like to know from Wido is if he opened a tracker issue for that, > since this kind of regression should not happen [tm]. > > Christian > >> Thanks, >> >> On Mon, Sep 12, 2016 at 3:05 PM, Wido den Hollander <wido@xxxxxxxx> wrote: >> >>> >>>> Op 12 september 2016 om 18:47 schreef "WRIGHT, JON R (JON R)" < >>> jonrodwright@xxxxxxxxx>: >>>> >>>> >>>> Since upgrading to Jewel from Hammer, we're started to see HEALTH_WARN >>>> because of 'blocked requests > 32 sec'. Seems to be related to writes. >>>> >>>> Has anyone else seen this? Or can anyone suggest what the problem might >>> be? >>>> >>> >>> Do you by any chance have old clients connecting? I saw this after a Jewel >>> upgrade as well and it was because of very old clients still connecting to >>> the cluster. >>> >>> Wido >>> >>>> Thanks! >>>> _______________________________________________ >>>> ceph-users mailing list >>>> ceph-users@xxxxxxxxxxxxxx >>>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>> _______________________________________________ >>> ceph-users mailing list >>> ceph-users@xxxxxxxxxxxxxx >>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>> >> >> >> On Mon, Sep 12, 2016 at 3:05 PM, Wido den Hollander <wido@xxxxxxxx> wrote: >> >>> >>>> Op 12 september 2016 om 18:47 schreef "WRIGHT, JON R (JON R)" < >>> jonrodwright@xxxxxxxxx>: >>>> >>>> >>>> Since upgrading to Jewel from Hammer, we're started to see HEALTH_WARN >>>> because of 'blocked requests > 32 sec'. Seems to be related to writes. >>>> >>>> Has anyone else seen this? Or can anyone suggest what the problem might >>> be? >>>> >>> >>> Do you by any chance have old clients connecting? I saw this after a Jewel >>> upgrade as well and it was because of very old clients still connecting to >>> the cluster. >>> >>> Wido >>> >>>> Thanks! >>>> _______________________________________________ >>>> ceph-users mailing list >>>> ceph-users@xxxxxxxxxxxxxx >>>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>> _______________________________________________ >>> ceph-users mailing list >>> ceph-users@xxxxxxxxxxxxxx >>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com >>> > > _______________________________________________ ceph-users mailing list ceph-users@xxxxxxxxxxxxxx http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
 |