Re: Cores/Memory/GHz recommendation for SSD based OSD servers

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



We have the model with 25 disks per node.

We have just two 10G network interfaces per node. Does that not limit
the thgouthput and hence the load on the CPUs?

-Sreenath

On 4/2/15, Alexandre DERUMIER <aderumier@xxxxxxxxx> wrote:
> Hi,
>
>>>with HP SL4540 server?
>
> this model http://www8.hp.com/h20195/v2/GetDocument.aspx?docname=c04128155
> ?
>
> (60 drives ?)
>
> I think for a full ssd node, it'll be impossible to reach max performance,
> you'll be cpu bound.
>
>
> I think a small node with 6-8 ssd osd  for 20cores should be ok.
>
>
> ----- Mail original -----
> De: "Sreenath BH" <bhsreenath@xxxxxxxxx>
> À: "Christian Balzer" <chibi@xxxxxxx>
> Cc: "ceph-users" <ceph-users@xxxxxxxxxxxxxx>
> Envoyé: Jeudi 2 Avril 2015 11:55:52
> Objet: Re:  Cores/Memory/GHz recommendation for SSD based
> OSD	servers
>
> Thanks for the tips.
> Could anyone share their experience building a SSD pool or a SSD cache
> tier with HP SL4540 server?
>
> rgds,
> Sreenath
>
> On 4/2/15, Christian Balzer <chibi@xxxxxxx> wrote:
>>
>> Hello,
>>
>> On Wed, 1 Apr 2015 18:40:10 +0530 Sreenath BH wrote:
>>
>>> Hi all,
>>>
>>> we are considering building all SSD OSD servers for RBD pool.
>>>
>> I'd advise you to spend significant time reading the various threads in
>> this ML about SSD based pools.
>> Both about the current shortcomings and limitations of SSD pools and which
>>
>> SSDs to (not) use.
>>
>>> Couple of questions:
>>>
>>> Does Ceph have any recommendation for number of cores/memory/ghz per
>>> SSD drive, similar to what is usually followed for hard drives(1
>>> core/1 GB Ram/1Ghz speed)?
>>>
>> Note that that 1GHz core per OSD only applies with pure HDD OSDs, once a
>> journal SSD enters the picture you're likely to want 2-3 times that.
>>
>> You probably don't want to try this with anything less than the
>> upcoming Hammer release, but even with that the current rule for SSD based
>>
>> pools is "the fastest cores you can afford and as many as possible".
>> And given the right loads, small write IOPS basically, you're probably
>> still going to be CPU bound.
>>
>> RAM is the same as with HDD based OSDs, but given how much more RAM helps
>>
>> I would advise at least 2GB per OSD and as much as you can afford.
>>
>> Regards,
>>
>> Christian
>>> thanks,
>>> Sreenath
>>> _______________________________________________
>>> ceph-users mailing list
>>> ceph-users@xxxxxxxxxxxxxx
>>> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>>>
>>
>>
>> --
>> Christian Balzer Network/Systems Engineer
>> chibi@xxxxxxx Global OnLine Japan/Fusion Communications
>> http://www.gol.com/
>>
> _______________________________________________
> ceph-users mailing list
> ceph-users@xxxxxxxxxxxxxx
> http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com
>
_______________________________________________
ceph-users mailing list
ceph-users@xxxxxxxxxxxxxx
http://lists.ceph.com/listinfo.cgi/ceph-users-ceph.com





[Index of Archives]     [Information on CEPH]     [Linux Filesystem Development]     [Ceph Development]     [Ceph Large]     [Linux USB Development]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]     [xfs]


  Powered by Linux