[ceph-users] Re: Cephfs IO halt on Node failure

If your pool has a min_size 2 and size 2 (always a bad idea) it will pause IO in case of a failure until the recovery has finished. So the described behaviour is expected. Zitat von Amudhan P : Hi, Crush rule is "replicated" and min_size 2 actually. I am trying to test multiple volume con

[ceph-users] Re: Ceph as a Fileserver for 3D Content Production

Den sön 17 maj 2020 kl 14:33 skrev Marc Roos : > outs, and you are more likely to shoot yourself in the foot. At least > ask first. Eg this bcache, I am not 100% sure what it is, but if it is > sitting between the osd process and the disk, it could be getting nasty > with a reset/power outage, whe

[ceph-users] Re: Ceph as a Fileserver for 3D Content Production

Is there any experience in using bcache with write back on ceph osds? Especially considering stability after a power outage? Holen Sie sich Outlook für Android ___ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send

[ceph-users] Re: Luminous to Nautilus mon upgrade oddity - failed to decode mgrstat state; luminous dev version? buffer::end_of_buffer

Hi Tom, Did you get past this? It looks like the mon is confused how to decode because of your non-standard release. (So I imaging that running all 14.2.9 mons would get past it, but if you're being cautious this should be reproduceable on your test cluster). -- Dan On Wed, May 13, 2020 at 12:0

[ceph-users] Re: Luminous to Nautilus mon upgrade oddity - failed to decode mgrstat state; luminous dev version? buffer::end_of_buffer

Hi Dan, We ended up upgrading all mons+mgrs to 14.2.9 and the message stopped and the PG stats reappeared, as expected. Marcello started the OSD restarts this morning. I think it would have been much less stressful to get the cluster onto 12.2.13 before the nautilus upgrade, and much easier to

[ceph-users] Re: Cephadm and rados gateways

This will be fixed in 15.2.2 https://tracker.ceph.com/issues/45215 ___ ceph-users mailing list -- ceph-users@ceph.io To unsubscribe send an email to ceph-users-le...@ceph.io

[ceph-users] Re: nfs migrate to rgw

Hi Wido, I did a research on the nfs files. I found that it contains much pictures about 50KB, and much video files around 30MB. The amount of the files is more than 1 million. Maybe I can find a way to seperate the files in more buckets so that there is no more than 1M objects in each bucket. But

[ceph-users] RGW issue with containerized ceph

Hi Gents, We are having strange problem with demo ceph container. Any write operation we do from the RGW fail with a NameResolutionFailure Read operation work fine. If we use the ceph UI we can create objects fine though. This is the env varibles we are using to configure it: This is the full com

[ceph-users] Re: how to restart daemons on 15.2 on Debian 10

Use the same pattern …. systemctl restart ceph-{fsid}@osd.{id}.service ~Sean > On May 18, 2020, at 7:16 AM, Ml Ml wrote: > > Thanks, > > The following seems to work for me on Debian 10 and 15.2.1: > > systemctl restart ceph-5436dd5d-83d4-4dc8-a93b-60ab5db145df@mon.ceph01.service > > How can

[ceph-users] Re: nfs migrate to rgw

On 5/18/20 1:51 PM, Zhenshi Zhou wrote: > Hi Wido, > > I did a research on the nfs files. I found that it contains much > pictures about  > 50KB, and much video files around 30MB. The amount of the files is more than > 1 million. Maybe I can find a way to seperate the files in more buckets > so

[ceph-users] Re: nfs migrate to rgw

On Mon, May 18, 2020 at 1:52 PM Zhenshi Zhou wrote: > > 50KB, and much video files around 30MB. The amount of the files is more > than > 1 million. Maybe I can find a way to seperate the files in more buckets so > that > there is no more than 1M objects in each bucket. But how about the small > f

[ceph-users] Dealing with non existing crush-root= after reclassify on ec pools

I have reclassified a CRUSH map, using the crushtool to a class based ruleset. I still have an ec pool with an older ec profile with a new non existing crush-root=hdd I already switched the pool’s ruleset over to a newer rule with a newer ec-profile with a correct crush-root But pool ls detail s

[ceph-users] feature mask: why not use HAVE_FEATURE macro in Connection::has_feature()?

Hi, everyone: Why don't we use HAVE_FEATURE macro in Connection::has_feature()? Do the features in a Connection not need to care about incarnation things? Missing the macro in Connection is really confusing. Would anyone like to give some explains on this? Thanks! __

[ceph-users] Re: how to restart daemons on 15.2 on Debian 10

Thanks, The following seems to work for me on Debian 10 and 15.2.1: systemctl restart ceph-5436dd5d-83d4-4dc8-a93b-60ab5db145df@mon.ceph01.service How can i restart a single OSD? Cheers, Michael On Sun, May 17, 2020 at 5:10 PM Sean Johnson wrote: > > I have OSD’s on the brain … that line shou

[ceph-users] Re: how to restart daemons on 15.2 on Debian 10

Is there no official link/docs how to manage the services on Debian 10 and 15.2.1? I have seach and seeked: systemctl restart ceph-5436dd5d-83d4-4dc8-a93b-60ab5db145df@mon.ceph01.service journalctl -u ceph-5436dd5d-83d4-4dc8-a93b-60ab5db145df@mon.ceph01.service docker logs Back in the days we ha

[ceph-users] Re: Dealing with non existing crush-root= after reclassify on ec pools

I think I did a bad job explaining my issue: I have a fairly old cluster which had a crush map with two trees, one for hdds and one for ssd, like root hdd {..} and root ssd {...} now with the newer class based rules I used crushtool —reclassify to merge those two trees into root default {...} So

[ceph-users] Re: Dealing with non existing crush-root= after reclassify on ec pools

that part of erasure profiles are only used when a crush rule is created when creating a pool without explicitly specifying a crush rule Paul -- Paul Emmerich Looking for help with your Ceph cluster? Contact us at https://croit.io croit GmbH Freseniusstr. 31h 81247 München www.croit.io Tel:

[ceph-users] v15.2.2 Octopus released

We're happy to announce the second bugfix release of Ceph Octopus stable release series, we recommend that all Octopus users upgrade. This release has a range of fixes across all components and a security fix. Notable Changes --- * CVE-2020-10736: Fixed an authorization bypass in mons

[ceph-users] Reweighting OSD while down results in undersized+degraded PGs

In a recent cluster reorganization, we ended up with a lot of undersized/degraded PGs and a day of recovery from them, when all we expected was moving some data around.  After retracing my steps, I found something odd.  If I crush reweight an OSD to  0 while it is down - it results in the PGs o

[ceph-users] Mismatched object counts between "rados df" and "rados ls" after rbd images removal

Hi, The numbers of object counts from "rados df" and "rados ls" are different in my testing environment. I think it maybe some zero bytes or unclean objects since I removed all rbd images on top of it few days ago. How can I make it right / found out where are those ghost objects? Or i should igno

[ceph-users] Re: nfs migrate to rgw

Awesome, thanks a lot ! I'll try it. Paul Emmerich 于2020年5月18日周一 下午8:53写道： > > On Mon, May 18, 2020 at 1:52 PM Zhenshi Zhou wrote: > >> >> 50KB, and much video files around 30MB. The amount of the files is more >> than >> 1 million. Maybe I can find a way to seperate the files in more buckets s

[ceph-users] Re: v15.2.2 Octopus released

I am getting the following error when trying to upgrade via cephadm ceph orch upgrade status {     "target_image": "docker.io/ceph/ceph:v15.2.2",     "in_progress": true,     "services_complete": [],     "message": "Error: UPGRADE_FAILED_PULL: Upgrade: failed to pull target image" } Are

[ceph-users] Re: Cephfs IO halt on Node failure

Behaviour is same even after setting min_size 2. On Mon 18 May, 2020, 12:34 PM Eugen Block, wrote: > If your pool has a min_size 2 and size 2 (always a bad idea) it will > pause IO in case of a failure until the recovery has finished. So the > described behaviour is expected. > > > Zitat von Amu

[ceph-users] Re: Cephfs IO halt on Node failure

Was that a typo and you mean you changed min_size to 1? I/O paus with min_size 1 and size 2 is unexpected, can you share more details like your crushmap and your osd tree? Zitat von Amudhan P : Behaviour is same even after setting min_size 2. On Mon 18 May, 2020, 12:34 PM Eugen Block, wr

[ceph-users] Re: Mismatched object counts between "rados df" and "rados ls" after rbd images removal

That's not wrong, those are expected objects that contain information about your rbd images. If you take a look into the rbd_directory (while you have images in there) you'll find something like this: host:~ $ rados -p pool listomapkeys rbd_directory id_fe976bcfb968bf id_ffc37728edbdab name_