[DRBD-user] The old question: drbd on lvm or vice versa?

Lee Riemer lriemer at bestline.net
Wed May 26 19:56:23 CEST 2010

Note: "permalinks" may not be as permanent as we would like,
direct links of old sources may well be a few messages off.


So you do this per node?

[ LVM <-> DRBD <-> LVM ]

On 5/26/2010 12:55 PM, Bart Coninckx wrote:
> On Wednesday 26 May 2010 19:36:33 Oliver Hoffmann wrote:
>    
>> Hi list!
>>
>> I searched the web but I only found lvm on drbd and many problems
>> concerning kvm or xen. I need lvm and drbd to have a flexible
>> file server (without xen or the like).
>>
>> If I put lvm on top of drbd I run into problems. Such as a complex
>> failover situation and I am limited to the physical space of a HD or
>> partitions.
>> Thus I think drbd on lvm would be better. In the end I want a
>> drbd- and pacemaker- based file server with nfs and cifs and iSCSI.
>> Plus I want to add one or more HD easily whenever needed or provide
>> more or less disk space for a client or a share.
>>
>> What do you think?
>>
>> Thanks for suggestions!
>>
>> Cheers,
>>
>> Oliver
>> _______________________________________________
>> drbd-user mailing list
>> drbd-user at lists.linbit.com
>> http://lists.linbit.com/mailman/listinfo/drbd-user
>>
>>      
> Oliver,
>
> I put DRBD on top of LVM for one sole reason: being able to resize DRBD
> resources if I would ever need to. Without LVM below it, it's much riskier
> (see DRBD manual).
> I also put LVM at the same time on top of DRBD (see "nested LVM" in the DRBD
> manual) which works great, provided you change /etc/lvm/lvm.conf
> About the complex failover situation: if you have cluster software taking care
> of this, it should not be referred to any more as being "complex". Setting
> that software up might be however.
>
> Hope this helps,
>
>
> Bart
> _______________________________________________
> drbd-user mailing list
> drbd-user at lists.linbit.com
> http://lists.linbit.com/mailman/listinfo/drbd-user
>    



More information about the drbd-user mailing list