Hi all,
I have a Ceph 17.2.5 cluster deployed via cephadm. After a few reboots it has
now entered a fairly broken state as shown below.
I am having trouble even beginning to diagnose this as a lot of the commands
just hang. For example “cephadm ps”, “ceph orch ls” just hang forever. Other
comm
Had to reduce the debug level back to normal. Debug Level 20 generated about
70GB log file in one hour. Of course there was no crash in that period.
-
Hi
I'm trying to enable Cephx on a cluster already running without Cephx.
Here is what I did.
1. I shutdown the cluster.
2. Enabled Cephx in ceph.conf, Mon and Mgr.
3. Brought the Monitor cluster up. No issue.
4. Tried to bring first Manager up, I'm getting following error:
=== mgr.a ===
S
> In Ceph/Libvirt docs only cachmodes writetrough and writeback are discussed.
> My clients's disks are all set to writeback in the libvirt client
> xml-definition.
>
> For a backup operation, I notice a severe lag on one of my VM's. Such a
> backup operation that takes 1 to 2 hours (on a same m
Hello,
I run a a hyperconverged pve cluster (V7.2) with 11 nodes. Each node has
8 4TB disks. pve and ceph are installed an running.
I wanted to create some ceph-pools with each 512 pgs. Since I want to
use erasure coding (5+3) when creating a pool one rbd pool for metadata
and the data pool
Dear list
Yesterday we updated our ceph cluster from 15.2.17 to 16.2.10 using
packages.
Our cluster is a mix of ubuntu 18 and ubuntu 20 with ceph coming from
packages in the ceph.com repo. All went well and we now have all nodes
running Pacific. However, there's something odd in the dashboar
Hi everyone,
In Rook, we have come up with `kubectl` based krew plugin tool name
`rook-ceph`, which will help users with the most commonly faces issues or
answer common questions users have like is my cluster is on the right
configuration or whether required resources are running and other things,
So it seems like a bucket still has objects listed in the bucket index but the
underlying data objects are no longer there. Since you made reference to a
customer, I’m guessing the customer does not have direct access to the cluster
via `rados` commands, so there’s no chance that they could have
if its this:
http://www.acmemicro.com/Product/17848/Kioxia-KCD6XLUL15T3---15-36TB-SSD-NVMe-2-5-inch-15mm-CD6-R-Series-SIE-PCIe-4-0-5500-MB-sec-Read-BiCS-FLASH-TLC-1-DWPD
its listed as 1 DWPD with a 5 year warranty. So should be ok.
Thanks,
Kevin
From: Rob
Hi,
you need to take into account the number of replicas as well. With 88
OSDs and the default max PGs per OSD of 250 you get the mentioned
22000 PGs (including replica): 88 x 250 = 22000.
With EC pools each chunk counts as one replica. So you should consider
shrinking your pools or let aut
Arguably the error should say something like
" TASK ERROR: error with 'osd pool create': mon_command failed - pg_num 512
size 8 would result in 251 cumulative PGs per OSD (22148 total PG replicas on
88 ‘in’ OSDs), which exceeds the mon_max_pg_per_osd value of 250.
IMHO this is sort of mixing
Hi Manuel,
I did the IO500 runs back in 2020 and wrote the cephfs aiori backend for
IOR/mdtest. Not sure about the segfault, it's been a while since I've
touched that code. It was working the last time I used it. :D Having
said that, I don't think that's your issue. The userland backend
Update on this, I have figured out what happened.
I had ceph packages installed on the node, since this was converted from
ceph-deploy to cephadm when we tested Octopus. When I upgraded Ubuntu,
it updated the packages and triggered a permission issue which is
already documented. This whole mes
13 matches
Mail list logo
