Balzer
Sent: Friday, 21 November 2014 2:39 PM
To: 'ceph-users'
Cc: Bond, Darryl
Subject: Re: [ceph-users] Kernel memory allocation oops Centos 7
Hello,
On Fri, 21 Nov 2014 04:31:18 +0000 Bond, Darryl wrote:
> Using the standard Centos 3.10.0-123.9.3.el7.x86_64 kernel. The NIC is a
&
s'
Cc: Bond, Darryl
Subject: Re: [ceph-users] Kernel memory allocation oops Centos 7
Hello,
On Fri, 21 Nov 2014 04:31:18 +0000 Bond, Darryl wrote:
> Using the standard Centos 3.10.0-123.9.3.el7.x86_64 kernel. The NIC is a
> 10G Ethernet broadcom so not infiniband. Tried swappiness =
31249404
Darryl
From: Christian Balzer
Sent: Friday, 21 November 2014 10:06 AM
To: 'ceph-users'
Cc: Bond, Darryl
Subject: Re: [ceph-users] Kernel memory allocation oops Centos 7
On Thu, 20 Nov 2014 22:10:02 +0000 Bond, Darryl wrote:
> Brief outlin
and not ipoib which was suggested as a
potential troublemaker. Sorry, just clicked on the google invitation
while scratching my eye, please ignore it (blaming Gmail interface).
On Fri, Nov 21, 2014 at 1:10 AM, Bond, Darryl wrote:
> Brief outline:
>
> 6 Node production cluster. Each node D
Brief outline:
6 Node production cluster. Each node Dell R610, 8x1.4TB SAS Disks, Samsung M.2
PCIe SSD for journals, 32GB RAM, Broadcom 10G interfaces.
Ceph 0.80.7-0.el7.centos from the ceph repositories.
About 10 times per day, each node will oops with the following message:
An example:
No
Restart one of the mons. It will clean them up.
6.1.4 does not accumulate these files. Might be worth upgrading.
James Harper wrote:
My cluster is in HEALTH_WARN state because one of my monitors has low disk
space on /var/lib/ceph. Looking into this in more detail, there are a bunch of
.sst f
Cluster has gone into HEALTH_WARN because the mon filesystem is 12%
The cluster was upgraded to cuttlefish last week and had been running on
bobtail for a few months.
How big can I expect the /var/lib/ceph/mon to get, what influences it's size.
It is at 11G now, I'm not sure how fast it has been
That was it, pool 3 had a rule configured that I had removed. So the rule no
longer existed.
I hadn't used the pool for some time, it was just used for some testing. I set
the rule set to a real one and the cluster is rebalancing as I am typing.
I'll probably delete the pool once I am sure that I
I have a small 3 node ceph cluster with 6 OSDs on each node
I would like to re-export some rbd images via LIO.
Is it recommended to run RBD/LIO on one of the cluster nodes?
Preliminary tests show that it works fine. I have seen reports (that I can't
find) that it is not recommended to run the RBD
