When I said "which is slower", I meant that it makes that running Instance
slower later... With preallocatoin=metadata, the running instances will be
a bit faster without losing storage space with lots of zeros...
My /var/lib/nova/insntances subdir in a XFS File System on top of a
hardware RAID...
Guys,
How can I configure the Compute Node, to create its qcow2 images, using
preallocation?
Currently, OpenStack Nova Compute creates the instance image, using
"qemu-img create" without passing the option "-o preallocation=metadata" to
it, which is slower.
So, how can I reconfigure nova comp
Recently, I figured out that, when using Ubuntu 14.04 + QEmu 2.0 (KVM), you
can not use the KSM (Kernel Samepage Merging) and run W2k8 R2 at the same
time, take a look:
Ubuntu 14.04 + QEmu 2.0 + KSM = 1, makes Windows 2008 R2 guests to crash
(BSOD)
https://bugs.launchpad.net/qemu/+bug/1338277
Thi
Hi All,
I am kind of new to the openstack.
I was looking to work on the nova component.
My aim is to be able to perform some tasks as soon as a VM is created on a
particular compute node.
I found there can be 2 ways:
1) Listen to the notification queue of nova (not sure which one)
2) Use a nova
So Stephen (presented the perf work at the last hackathon) had an idea about
maybe saving on some connections via batching up groups of container updates
before making the connection to the container server in async_update().
Clearly this would delay updates by some amount but they're not immed
On Thu, Jul 10, 2014 at 04:48:02PM +0100, Michael Hearn wrote:
>Want to use Trusts but when creating I am being told I do not have the
>correct authorization e.g
>You are not authorized to perform the requested action.", "code": 403,
>"title": "Forbidden
>
>Distro: Icehouse
>
As Pete mentioned, since Swift can use a lot of sockets and fds when the system
is under load. Take a look at
http://docs.openstack.org/developer/swift/deployment_guide.html#general-system-tuning
for some sysctl settings that can help. Also, note that if you start Swift as
root (it can drop per
Thanks for your inputs Edward and Pete. I'll set sysctl net.ipv4.tcp_tw_reuse.
On Fri, Jul 11, 2014 at 8:05 AM, Pete Zaitcev wrote:
> On Tue, 8 Jul 2014 16:26:10 -0700
> Shrinand Javadekar wrote:
>
>> I see that these servers do not use a persistent http connection
>> between them. So every blo
The first error is probably triggered because the Mysql user you created for
Cinder doesn't have write permissions on the cinder db.
The second error is because you passed the "logs" argument (string) to the
"cinder-manage db sync" command which can only accept an integer (for example
3, if ver
hi all
in cinder configuration from installlation guide there is a database setting
and when I try cinder-manage db sync got this error
cinder-manage db sync
2014-07-11 23:32:29.907 10032 CRITICAL cinder [-] OperationalError:
(OperationalError) attempt to write a readonly database u'\nCREATE
TAB
On Tue, 8 Jul 2014 16:26:10 -0700
Shrinand Javadekar wrote:
> I see that these servers do not use a persistent http connection
> between them. So every blob get/put/delete request will create a new
> connection, use it and tear it down. In a highly concurrent
> environment with thousands of such
Hi
I got a problem with Ceph as a backend for ephemeral volumes.
Creation of the instance completes as supposed, and instance is functional,
but when I try to delete it - instance stuck in "Deleting" state forever
Part of nova.conf on compute node:
...
[libvirt]
inject_password=false
inject_key=f
we change /etc/security/limits.conf (actually add an entry to
/etc/security/limits.conf.d) for the rabbit user.
We just do that for the non-compute nodes.
Gavin
From: Jens-Christian Fischer [mailto:jens-christian.fisc...@switch.ch]
Sent: Friday, July 11, 2014 3:41 PM
To: Brebn
There is just one queue that is building up: notifications.info - there are no
consumers for it.
I have upped the ulimit for the rabbitmq server (by changing
/etc/default/rabbitmq-server) but that hasn’t made any difference.
Are you also upping the limits on the compute nodes?
cheers
jc
--
Delete issues in my experience are typically linked to messaging getting stuck
(or dead hosts). I would have a look at the
RMQ queue lengths and see if any are building up.
Have you adjusted your open file descriptor limits ? The default Linux limit
(1024) is NOT enough in our experience for any
Did you ever come to the bottom of this?
We are seeing the same thing with a brand new Icehouse / Ceph installation
(using the stock distributions)
Restarting nova-compute on the compute nodes usually makes the instances go
away (or resetting the state, and then restarting nova-compute)
We see
Hi, guys.
I read a lot of OpenStack blueprints about SR-IOV.
It looks like actively proceeding to be done in Juno.
However, I'd like to having some clues of SR-IOV.
Does anyone know how SR-IOV is planning in more detail than the current
bluprints?
Is there something what I can test for it?
A
17 matches
Mail list logo
