Thanks, Can I ignore this warning then?
health HEALTH_WARN
crush map has legacy tunables (require bobtail, min is firefly)
Cheers,
Mike
> On Jul 12, 2016, at 9:57 AM, Sean Redmond <[email protected]> wrote:
>
> Hi,
>
> Take a look at the docs here
> (http://docs.ceph.com/docs/jewel/rados/operations/crush-map/#tunables
> <http://docs.ceph.com/docs/jewel/rados/operations/crush-map/#tunables>) for
> details on what tunables work with which kernel version. and how you can
> change them e.g:
>
> 'ceph osd crush tunables bobtail'
>
> You could use a main line kernel from here (http://elrepo.org/tiki/kernel-ml
> <http://elrepo.org/tiki/kernel-ml>) as 4.6 is available for CentOS7
>
> Thanks
>
> On Tue, Jul 12, 2016 at 5:45 PM, Mike Jacobacci <[email protected]
> <mailto:[email protected]>> wrote:
> Hi Sean,
>
> Thanks for the quick response, this is what I see in dimes:
>
> set mismatch, my 102b84a842a42 < server's 40102b84a842a42, missing
> 400000000000000
>
> How can I set the tunable low enough? And what does that mean for
> performance?
>
> Cheers,
> Mike
>
>> On Jul 12, 2016, at 9:43 AM, Sean Redmond <[email protected]
>> <mailto:[email protected]>> wrote:
>>
>> Hi,
>>
>> It should work for you with kernel 3.10 as long as turntables are set low
>> enough - Do you see anything in 'dmesg'?
>>
>> Thanks
>>
>> On Tue, Jul 12, 2016 at 5:37 PM, Mike Jacobacci <[email protected]
>> <mailto:[email protected]>> wrote:
>> Hi All,
>>
>> Is mounting rbd only really supported in Ubuntu? All of our servers are
>> CentOS 7 or RedHat 7 and since they are at Kernel 3.10, I can’t get rbd
>> working. Even if I use legacy tunables or set tunables to 1/2/3… Nothing
>> will mount the block device.
>>
>> The only answer I can find to get this working under RHEL or CentOS is to
>> upgrade the kernel… Doesn’t that mean RBD isn’t support for a production
>> environment unless it’s Ubuntu or whatever OS supports a Kernel over 3.18?
>>
>> Thanks for everyone’s help so far!
>>
>> Cheers,
>> Mike
>> _______________________________________________
>> ceph-users mailing list
>> [email protected] <mailto:[email protected]>
>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>> <http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com>
>>
>
>
_______________________________________________
ceph-users mailing list
[email protected]
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
