[ovirt-users] Q: 2-Node Failover Setup - NFS or GlusterFS ?

[Yaniv Kaul]

On Mon, Jan 1, 2018 at 4:00 PM, Andrei V <andreil1 at starlett.lv> wrote:

> On 01/01/2018 10:10 AM, Yaniv Kaul wrote:
>
>
> On Mon, Jan 1, 2018 at 12:50 AM, Andrei V <andreil1 at starlett.lv> wrote:
>
>> Hi !
>>
>> I'm installing 2-node failover cluster (2 x Xeon servers with local RAID
>> 5 / ext4 for oVirt storage domains).
>> Now I have a dilemma - use either GlusterFS replica 2 or stick with NFS?
>>
>
> Replica 2 is not good enough, as it can leave you with split brain. It's
> been discussed in the mailing list several times.
> How do you plan to achieve HA with NFS? With drbd?
>
> Hi, Yaniv,
> Thanks a lot for detailed explanation!
>
> I know Replica 2 is not optimal solution.
> Right now I have only 2 servers with internal RAIDs for nodes, and till
> end of this week system had to be running in whatever condition.
> May be its better to use local storage domain on each node, set export
> domain on backup node, and backup VMs to 2nd backup node in timed interval?
> Its not highly-available yet workable solution.
>
> 4.2 Engine is running on separate hardware.
>>
>
> Is the Engine also highly available?
>
>
> Its KVM appliance, could be launched on 2 SuSE servers.
>
> Each node have its own storage domain (on internal RAID).
>>
>
> So some sort of replica 1 with geo-replication between them?
>
>
> Could it be the following?
> 1) Local storage domain on each node
> 2) GlusterFS geo-replication or over these directories? Not sure this will
> work.
>
>
>> All VMs must be highly available.
>>
>
> Without shared storage, it may be tricky.
>
>
> Seems to be timely VM backup to 2nd node is enough for this time.
> With current hardware anything above is too cumbersome to setup.
>

Agreed.
Y.


>
>
>
> One of the VMs is an accounting/stock control system with FireBird SQL
>> server on CentOS is speed-critical.
>>
>
> But is IO the bottleneck? Are you using SSDs / NVMe drives?
> I'm not familiar enough with FireBird SQL server - does it have an
> application layer replication you might opt to use?
> In such case, you could pass-through a NVM disk and have the application
> layer perform the replication between the nodes.
>
>
>> No load balancing between nodes necessary. 2nd is just for backup if 1st
>> for whatever reason goes up in smoke. All VM disks must be replicated to
>> backup node in near real-time or in worst case each 1 - 2 hour.
>> GlusterFS solves this issue yet at high performance penalty.
>>
>
> The problem with a passive backup is that you never know it'll really work
> when needed. This is why active-active is many time preferred.
> It's also more cost effective usually - instead of some HW lying around.
>
>
>>
>> >From what I read here
>> http://lists.ovirt.org/pipermail/users/2017-July/083144.html
>> GlusterFS performance with oVirt is not very good right now because QEMU
>> uses FUSE instead of libgfapi.
>>
>> What is optimal way to go on ?
>>
>
> It's hard to answer without additional details.
> Y.
>
>
>> Thanks in advance.
>> Andrei
>>
>> _______________________________________________
>> Users mailing list
>> Users at ovirt.org
>> http://lists.ovirt.org/mailman/listinfo/users
>>
>
>
>
> _______________________________________________
> Users mailing list
> Users at ovirt.org
> http://lists.ovirt.org/mailman/listinfo/users
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ovirt.org/pipermail/users/attachments/20180101/279a0b6f/attachment.html>