I upgrade two small converged ovirt clusters last night, both running 3.10.x on
top of centos 7.4. Neither was heavily loaded to begin with, but I’m not seeing
significant change. I wasn’t heavily instrumented, but it looks only 1-2%
increase in average utilization on the 1G interconnected pair, less on the 10G
connected trio, maybe 4-5% more than I’m used to seeing during the heal after
rebooting them but hard to tell on those systems.
My dev cluster of 3 older systems running centos 7.4 w/ gluster 3.12.4 (up from
3.12.3) in an ovirt cluster with no load other than the management engine saw
no noticeable change in utilization. Not doing much at the moment, of course,
but there you go.
Haven’t done my 3 larger servers yet, but looking good for the weekend…
-D
> From: Shyam Ranganathan <srang...@redhat.com>
> Subject: Re: [Gluster-users] So how badly will Gluster be affected by the
> Intel 'fix'
> Date: January 4, 2018 at 1:48:36 PM CST
> To: WK; gluster-users@gluster.org
>
> On 01/04/2018 12:58 PM, WK wrote:
>> I'm reading that the new kernel will slow down context switches. That is
>> of course a big deal with FUSE mounts.
>>
>> Has anybody installed the new kernels yet and observed any performance
>> degradation?
>
> We are in the process of testing the same out. Hopefully later next week
> we would be able to post the numbers that we observe.
>
> Other results in the interim are welcome as well!
>
>>
>> -wk
>>
>>
>> ___
>> Gluster-users mailing list
>> Gluster-users@gluster.org
>> http://lists.gluster.org/mailman/listinfo/gluster-users
> ___
> Gluster-users mailing list
> Gluster-users@gluster.org
> http://lists.gluster.org/mailman/listinfo/gluster-users
___
Gluster-users mailing list
Gluster-users@gluster.org
http://lists.gluster.org/mailman/listinfo/gluster-users
