[SOLVED] debian jessie kvm installation

[RRJ]

not to discuss here, but there is no anything like gluster or ceph production ready. These technologies are young. So at this point I would say, that gluster is mature enough if you exactly know, what you need.
May be ceph got some advantages, but it is definitely much more slower. So just have to stick to one of them and wait for another to die RH won't develop both of them.

 

[badji]

I have abandoned glusterfs entirely. Despite what the gluster people claim it is simply not production ready!

Hi mir, perhaps this news will make you change your mind

http://blog.nixpanic.net/2015/05/glusterfs-370-has-been-released.html

I use glusterfs.

Merci.

Moula.

 

[RRJ]

Okay,
It seems that only nodes with I210 are effected. But why only with D8? And there is lots of stuff about this NIC and igb driver. Going to investigate it further.
Any comments?

 

[RRJ]

No, it is not I210 issue.
I'm still not able to install jessie as VM on proxmox. Upgraded glusterfs to 3.6.4, still same problem on every proxmox node and does not matter really is it HA or DISTR volume. Today installation of Ubuntu 14.04 LTS with Mate went just fine. So it seems like Debian8 problem...
Could you guys from devs team just create a simple lab and give it a test?

Use GlusterFS as storage backend and try to install debian8 as VM guest. I really want to solve this problem.

 

[RRJ]

here is one of the errors example. its like files that debian installer copies to the virtual disk that is located on glusterfs storage are getting corrupted.
in-target is /dev/vda1

 

[RRJ]

It seems it is proxmox or debian fault!

http://www.gluster.org/pipermail/gluster-users/2015-July/022804.html

other (ovirt ie) VEs run D8 without problems on glusterfs.

 

[RRJ]

devs, please consider creating this type of lab, it seems it proxmox's fault.

 

[liska_]

I tried it now on my systems and I noticed some delays when installing debian8 on glusterfs when using virtio. After using iscsi it seems ok to me. Maybe it could help to change cache or drive settings? Anyway, I can only use vm on glusterfs when using cache write through.

 

[RRJ]

How your glusterfs is configured? you add storage via promox GUI or from console? I do not use caching for any VM and they run well.
But thanks for sharing your experience. Could you describe you setup please?

 

[RRJ]

filled a bug on glusterfs

https://bugzilla.redhat.com/show_bug.cgi?id=1242913

 

[liska_]

I added storage through GUI. I have two servers mirroring data for other hosts via gluster and ceph. My config follows some howtos on proxmox wiki and other sources:

Volume Name: virtual
Type: Replicate
Volume ID: c18748c4-b9f6-45a5-8f51-e7a4c1e32cb0
Status: Started
Number of Bricks: 1 x 2 = 2
Transport-type: tcp
Bricks:
Brick1: cl1.ucl:/big/gluster/virtual
Brick2: cl2.ucl:/big/gluster/virtual
Options Reconfigured:
performance.write-behind: off
server.allow-insecure: on
diagnostics.brick-log-level: WARNING
diagnostics.client-log-level: WARNING
nfs.enable-ino32: on
nfs.addr-namelookup: off
nfs.disable: on
performance.cache-max-file-size: 2MB
performance.cache-refresh-timeout: 4
performance.cache-size: 256MB
performance.write-behind-window-size: 4MB
performance.io-thread-count: 32

 

[RRJ]

thanx.
these two seem to me weird:

performance.write-behind: off
performance.write-behind-window-size: 4MB

if its off why you specify its window-size?
but well, it does not matter anyway. Doesn't seem like some of these options could really affect something