Args so this was the reason definitively: bash -c ip route - does
not work, and then CS agent deletes 169.254.0.0 route completely...
why - have no damn clue...
I made simple cron, creating route, sleep 30sec, then create route again...
- so at least VR creation will be prolonged for max
I think I located the problem, sort of:
it seems like ACS agent is deleting route to 168.254.0.0 - so after the
patchviasocket - agent can not connect to private interface on VM
(169.254.x.x)
if I try to provision VPC - it waits, and waits, and waits... until I add
route manually on the particual
Yes I understadn that - more of a problem, qemu-img from CentOS 6.5 doesnt
work snapshots (redhat removed -S flag, etc.), then we need RBD support
(for CEPH), which RH does not support in normal RHEL/CentOS, etc.
Next cloud deployement - dump the the RHEL/CentOS, and go with Ubuntu
14.04, which d
Yeah, I don't remember all of the discussion around the minimum
version and why, but each libvirt and qemu has such different behavior
(not all good), that we decided to target a minimum regardless of what
was supplied by the distros. It's just too hard to support them all
with their differing beha
Marcus - QEMU is 0.12.x (CentOS 6.4)
but libvirt is compiled 1.2.3
On 14 March 2015 at 16:58, Marcus wrote:
> Oh, this is an unsupported version. should be using libvirt 0.9.4+ and
> Qemu 1.0+ per the install docs.
>
> On Sat, Mar 14, 2015 at 8:56 AM, Andrija Panic
> wrote:
> > Tried the propo
Oh, this is an unsupported version. should be using libvirt 0.9.4+ and
Qemu 1.0+ per the install docs.
On Sat, Mar 14, 2015 at 8:56 AM, Andrija Panic wrote:
> Tried the proposed solution - but it still don't work - BTW my qemu is
> 0.12.x - from CentOS 6.4...
>
> When I manually start the script
People are mostly using upgraded 'virt' repos.
On Sat, Mar 14, 2015 at 8:58 AM, Marcus wrote:
> Oh, this is an unsupported version. should be using libvirt 0.9.4+ and
> Qemu 1.0+ per the install docs.
>
> On Sat, Mar 14, 2015 at 8:56 AM, Andrija Panic
> wrote:
>> Tried the proposed solution - b
Tried the proposed solution - but it still don't work - BTW my qemu is
0.12.x - from CentOS 6.4...
When I manually start the script with all parameters, it simply hangs, no
output at all, until ctrl + c
BTW this is with now upgraded ACS to 4.3.2...
:(
On 12 March 2015 at 09:24, Andrija Panic w
Hi Marcus, will do - I'm just unconfortable changing live things, but that
is my problem. Will let you know of the output.
Thanks
On 12 March 2015 at 01:34, ilya musayev
wrote:
> Andrija,
>
> Have you tried the solution Loic mentioned? Its seems like an easy one to
> try..
>
> Thanks
> ilya
>
>
Andrija,
Have you tried the solution Loic mentioned? Its seems like an easy one
to try..
Thanks
ilya
On 3/11/15 3:54 AM, Andrija Panic wrote:
Oh well, I will then update to 4.3.2 anyway, and hope to see the problem go
away...
THanks guys
On 10 March 2015 at 22:45, Loic Lambiel wrote:
Hi
Oh well, I will then update to 4.3.2 anyway, and hope to see the problem go
away...
THanks guys
On 10 March 2015 at 22:45, Loic Lambiel wrote:
> Hi,
>
> Maybe the same issue we faced here: https://github.com/exoscale/
> cloudstack/commit/df3e7533a78c3594cdd39b8b18e6f4a3abcb8cfd
>
> The patch ha
Hi,
Maybe the same issue we faced here:
https://github.com/exoscale/cloudstack/commit/df3e7533a78c3594cdd39b8b18e6f4a3abcb8cfd
The patch has not been yet submitted upstream. I'll do it once it is
running on our production.
Anyway, you may easily test it by performing the small change direct
I think I've seen some bugfixes for this in newer versions of 4.3+
On Tue, Mar 10, 2015 at 11:40 AM, Andrija Panic wrote:
> Hi guys,
>
> Kernel is Elrepo 3.10.67 (CentOS 6.6)
> Libvirt manually compiled 1.2.3 with RBD support.
> Qemu 0.12.1.2 - Inktank patched original RHEL 6.4 rpms with RBD supp
Hi guys,
Kernel is Elrepo 3.10.67 (CentOS 6.6)
Libvirt manually compiled 1.2.3 with RBD support.
Qemu 0.12.1.2 - Inktank patched original RHEL 6.4 rpms with RBD support.
And here is the output, with debug mode, from Agent.
At the begining I see agent deleting routes for 169.254.x.x network...:
What kernel are you running? Please include Libvirt and Qemu version.
On 3/10/15 8:21 AM, Andrija Panic wrote:
Hi,
we are having issues from time to time with starting VR - while creating
VPC.
Here is the few lines from MGMT log:
2015-03-10 16:11:50,018 WARN [c.c.n.v.VpcManagerImpl]
(Job-Exe
k.apache.org, us...@cloudstack.apache.org
> Sent: Tuesday, 10 March, 2015 15:21:24
> Subject: [HELP] VR sometimes fail to start
> Hi,
>
> we are having issues from time to time with starting VR - while creating
> VPC.
>
> Here is the few lines from MGMT log:
>
> 2015-03
Hi,
we are having issues from time to time with starting VR - while creating
VPC.
Here is the few lines from MGMT log:
2015-03-10 16:11:50,018 WARN [c.c.n.v.VpcManagerImpl]
(Job-Executor-27:ctx-84f4fba4 ctx-970acb35) Failed to start vpc [VPC
[621-VPC-delete] due to
com.cloud.exception.AgentUnav
17 matches
Mail list logo
