[pve-devel] Sheepdog problems

[Stefan Priebe]

I think the same but under heavy 10g and ssd load you'll run into even heavier problems like absolutely slow sequential write speed 100mb/s so to me it seems no solution sheepdog and or ceph are ready for any productional use case.

So let's talk about them in a year ;-) and see what has happened until then.

Greets
Stefan

Am 01.08.2012 um 07:49 schrieb Dietmar Maurer <dietmar at proxmox.com>:

> The sheepdog developers thinks that it is a good idea to unconditionally start recovery/rebalance cluster data as soon as one nodes joins/leave.
>  
> That basically means that you fully utilize your network during a node reboot, doing
> totally unnecessary data movements.
>  
> Second, they think they do not need automatic cleanup until you run ‘collie cluster cleanup’ manually.
>  
> One the sheepdog lists it seems that I am the only one concerned about that.
>  
> For me this make sheepdog quite unusable for pve?
>  
> Any opinions on that?
>  
> - Dietmar
> _______________________________________________
> pve-devel mailing list
> pve-devel at pve.proxmox.com
> http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.proxmox.com/pipermail/pve-devel/attachments/20120801/b28510e7/attachment.htm>