Re: keep qcow2 images in sync

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

 



On 2019-06-29 7:26 a.m., Marcus Schopen wrote:
Hi,

i am looking for a solution to keep qcow2 images of two KVM hosts sync.
(about 20 guests, images size max 100 GB). Unfortunately, shared
storage is not available to me. My first look was on Gluster, but after
what I've read so far, there might be performance issues here. I don't
need real HA. If the first host fails, it would be sufficient for me to
manually start the KVM guests on the second host. Ultimately, I just
want to avoid having to restore the images from the nightly backup and
have a maximum "data loss" of 24 hours. Is drdb a alternative here? And
what should I pay particular attention to when setting up, split brain
etc.?

Cheers
Marcus


Hi Marcus,

    I use GlusterFS for VM image storage.  Most of them are RAW files and QCOW2 is fine too.  There are things to watch for performance.  In a setup we have for a client, the Windows File server (8TB + 5TB) is now screaming compared to the old setup: 120 MBps!  And the overtaxed mail server (CentOS 7 Linux VM running CommuniGate) on the same cluster is doing fine too (low IOwait considering the 1TB+ Public Folders, 140 Outlook users and a total of 2.5 TB of emails).

    This setup is configured in three way replication.  The 3 servers are Supermicro using LSI RAID 6 arrays.  GlusterFS bricks are ontop thin provisioned LVM logical volumes.  Watch out for lvm / filesystem optimal alignment VS your physical RAID arrays.

    Just keep in mind that with 2 servers, you are exposed to split brain issues.  The other way of doing it in the above setup would be 2+1 replication with an arbiter: It would saves space and still enforce quorum.  But many things i read are pointing to 3 way replication for simplicity and stability.


Here are the infos about this Gluster cluster:

3 x Supermicro 3U servers with each:

1 x Xeon E5-2609 CPU

32 Gig RAM

Integrated LSI SAS 2208

32 TB Array consisting of 8 x HGST NAS 7200 RPMs drives in RAID 6

2 x Integrated Intel X540-AT2 10 Gbps ethernet

We are considering of adding SSD caching, still looking for the best way of doing it... (LVM Cache  / Gluster Tiering / etc)


There are 2 KVM Servers running VMs with dual E5-2630v2 CPU, 96 Gigs RAM and dual 10 Gbps ether

The next part will be to add OVirt.


Be sure to simulate what you want to achieve before going in prod.  Test disconnection of a node, etc.  I like Gluster but there are situations that you must know what you're doing.  I paid a premium for the Red Hat Gluster Video self training and it is not a good course IMHO (more of an intro...), i didn't learn what was of interest to me (highly technical) and i taught i could ask questions: They direct you to tech support and you must have a contract.... They seem to want to direct you to their "design" team.

I didn't find any "best practices", not real good guidelines and explanations...

A real important thing to consider: The network part.  Look for bonding + dual attach switch and for ECM Routing + OSPF on the servers / network.  I'm still searching for the best network setup to get performance and full redundancy.


Hope this helps.


Guy

--
Guy Boisvert, ing.
IngTegration inc.
http://www.ingtegration.com
https://www.linkedin.com/pub/guy-boisvert/7/48/899/fr

AVIS DE CONFIDENTIALITE : ce message peut contenir des
renseignements confidentiels appartenant exclusivement a
IngTegration Inc. ou a ses filiales. Si vous n'etes pas
le destinataire indique ou prevu dans ce  message (ou
responsable de livrer ce message a la personne indiquee ou
prevue) ou si vous pensez que ce message vous a ete adresse
par erreur, vous ne pouvez pas utiliser ou reproduire ce
message, ni le livrer a quelqu'un d'autre. Dans ce cas, vous
devez le detruire et vous etes prie d'avertir l'expediteur
en repondant au courriel.

CONFIDENTIALITY NOTICE : Proprietary/Confidential Information
belonging to IngTegration Inc. and its affiliates may be
contained in this message. If you are not a recipient
indicated or intended in this message (or responsible for
delivery of this message to such person), or you think for
any reason that this message may have been addressed to you
in error, you may not use or copy or deliver this message to
anyone else. In such case, you should destroy this message
and are asked to notify the sender by reply email.

_______________________________________________
Gluster-users mailing list
Gluster-users@xxxxxxxxxxx
https://lists.gluster.org/mailman/listinfo/gluster-users




[Index of Archives]     [Gluster Development]     [Linux Filesytems Development]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [eCos]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux