[Openstack] xenserver and cinder
Hello, guys! We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to kilo) We have 2 servers, controller has glance, nova and keystone, and compute vms have only nova-compute installed. We use local storage on each server, not shared one (this is important) We want to implement cinder service to manage volumes easily, but i can't understand how can we do it.XenAPINFSDriver is deprecated and isn't supported any more. The only solution i found so far is to create storage VM, give it some space and use this space for cinder volumes, and then mount it in VMs as ISCSI targets. And in the same time nova uses xenAPI to create and manage volumes when it creates instances. And what i want to make is to make cinder use Xenserver volumes both when nova creates VM and when i create volume with cinder. Without second level of abstraction. Can you tell me what are the best practices to use cinder for Xenservers with local storage? Regards, Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] xenserver and cinder
So, do I understand it right - I can't get rid of additional layer for cinder volumes, if I want to use local storage? Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 12:34 PM To: Ivan Derbenev; openstack@lists.openstack.org Subject: RE: xenserver and cinder Hi Ivan, XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied on Nova aggregates) - however it's not the easiest way to consume Cinder volumes. The XenServer Nova integration supports BFV and volume attach for Cinder volumes presented over iSCSI, so however those are managed by Cinder is independent of how XenServer can consume them. Even when using the deprecated XenAPINFSDriver, the storage had to be remote storage (i.e. provided to the host through an NFS server) so that does not let you use the hypervisor-local storage as block storage provisionable by Cinder and you would have needed a separate VM if you actually need your Cinder volumes to be provisioned from local storage. Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 13 July 2015 18:53 To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: [Openstack] xenserver and cinder Hello, guys! We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to kilo) We have 2 servers, controller has glance, nova and keystone, and compute vms have only nova-compute installed. We use local storage on each server, not shared one (this is important) We want to implement cinder service to manage volumes easily, but i can't understand how can we do it.XenAPINFSDriver is deprecated and isn't supported any more. The only solution i found so far is to create storage VM, give it some space and use this space for cinder volumes, and then mount it in VMs as ISCSI targets. And in the same time nova uses xenAPI to create and manage volumes when it creates instances. And what i want to make is to make cinder use Xenserver volumes both when nova creates VM and when i create volume with cinder. Without second level of abstraction. Can you tell me what are the best practices to use cinder for Xenservers with local storage? Regards, Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] xenserver and cinder
Ok, well... I don't have the reason to use precisely Cinder. What I need is the ability to create, delete and attach data volumes to instances. As far as I know I can't do it without Cinder. If you can show me a path how to do it, and use local storages only - that would be really great Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 4:58 PM To: Ivan Derbenev; openstack@lists.openstack.org Subject: RE: xenserver and cinder As I understand it Cinder provides volumes that must be accessible to VMs on any of the hosts - i.e. must be accessed over the network. Therefore if you want to use Cinder to provide data volumes (or BFV) then you need a VM/physical machine/storage array providing the actual storage which is accessible over the network. I'm not aware of a way to use Cinder create volumes that are only accessible to VMs on a single host - which is what you would need to use storage that is entirely local to the hypervisor. Just to understand, why are you trying to use Cinder backed by local storage only? What's the use case here? Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 14 July 2015 14:50 To: Bob Ball; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder So, do I understand it right - I can't get rid of additional layer for cinder volumes, if I want to use local storage? Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 12:34 PM To: Ivan Derbenev; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder Hi Ivan, XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied on Nova aggregates) - however it's not the easiest way to consume Cinder volumes. The XenServer Nova integration supports BFV and volume attach for Cinder volumes presented over iSCSI, so however those are managed by Cinder is independent of how XenServer can consume them. Even when using the deprecated XenAPINFSDriver, the storage had to be remote storage (i.e. provided to the host through an NFS server) so that does not let you use the hypervisor-local storage as block storage provisionable by Cinder and you would have needed a separate VM if you actually need your Cinder volumes to be provisioned from local storage. Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 13 July 2015 18:53 To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: [Openstack] xenserver and cinder Hello, guys! We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to kilo) We have 2 servers, controller has glance, nova and keystone, and compute vms have only nova-compute installed. We use local storage on each server, not shared one (this is important) We want to implement cinder service to manage volumes easily, but i can't understand how can we do it.XenAPINFSDriver is deprecated and isn't supported any more. The only solution i found so far is to create storage VM, give it some space and use this space for cinder volumes, and then mount it in VMs as ISCSI targets. And in the same time nova uses xenAPI to create and manage volumes when it creates instances. And what i want to make is to make cinder use Xenserver volumes both when nova creates VM and when i create volume with cinder. Without second level of abstraction. Can you tell me what are the best practices to use cinder for Xenservers with local storage? Regards, Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] xenserver and cinder
Well, we are migrating from cloudstack now and it has this feature So, do I understand It right. In current situation I cannot use data disk's at local storages at all? The only solution is to make additional layer through cinder+lvm and share cinder vols through iscsi? Do you know what's the best practice for a system like mine? I mean XS with local storages Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 5:18 PM To: Ivan Derbenev; openstack@lists.openstack.org Subject: RE: xenserver and cinder Hi Ivan, In that case then no, I don't believe what you are trying to achieve is currently possible with OpenStack. My understanding is that detachable block storage is provided only by Cinder and Cinder must be network-attached storage as there is no concept of tying the storage to particular VMs. For example, you create the cinder volumes _before_ you attach it to a virtual machine - so there is no knowledge of which VM or which hypervisor will be attaching to the volume. Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 14 July 2015 15:10 To: Bob Ball; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder Ok, well... I don't have the reason to use precisely Cinder. What I need is the ability to create, delete and attach data volumes to instances. As far as I know I can't do it without Cinder. If you can show me a path how to do it, and use local storages only - that would be really great Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 4:58 PM To: Ivan Derbenev; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder As I understand it Cinder provides volumes that must be accessible to VMs on any of the hosts - i.e. must be accessed over the network. Therefore if you want to use Cinder to provide data volumes (or BFV) then you need a VM/physical machine/storage array providing the actual storage which is accessible over the network. I'm not aware of a way to use Cinder create volumes that are only accessible to VMs on a single host - which is what you would need to use storage that is entirely local to the hypervisor. Just to understand, why are you trying to use Cinder backed by local storage only? What's the use case here? Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 14 July 2015 14:50 To: Bob Ball; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder So, do I understand it right - I can't get rid of additional layer for cinder volumes, if I want to use local storage? Regards, IT engineer Farheap, Russia Ivan Derbenev From: Bob Ball [mailto:bob.b...@citrix.com] Sent: Tuesday, July 14, 2015 12:34 PM To: Ivan Derbenev; openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: RE: xenserver and cinder Hi Ivan, XenAPINFSDriver was primarily useful for pooled scenarios (which in turn relied on Nova aggregates) - however it's not the easiest way to consume Cinder volumes. The XenServer Nova integration supports BFV and volume attach for Cinder volumes presented over iSCSI, so however those are managed by Cinder is independent of how XenServer can consume them. Even when using the deprecated XenAPINFSDriver, the storage had to be remote storage (i.e. provided to the host through an NFS server) so that does not let you use the hypervisor-local storage as block storage provisionable by Cinder and you would have needed a separate VM if you actually need your Cinder volumes to be provisioned from local storage. Bob From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: 13 July 2015 18:53 To: openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Subject: [Openstack] xenserver and cinder Hello, guys! We are currently making an installation of XS6.5+OS Juno (maybe we'll switch to kilo) We have 2 servers, controller has glance, nova and keystone, and compute vms have only nova-compute installed. We use local storage on each server, not shared one (this is important) We want to implement cinder service to manage volumes easily, but i can't understand how can we do it.XenAPINFSDriver is deprecated and isn't supported any more. The only solution i found so far is to create storage VM, give it some space and use this space for cinder volumes, and then mount it in VMs as ISCSI targets. And in the same time nova uses xenAPI to create and manage volumes when it creates instances. And what i want to make is to make cinder use Xenserver volumes both when nova creates VM and when i create volume with cinder. Without second level of abstraction. Can you tell me what are the
Re: [Openstack] XenServer 6.5 or KVM
We had a small deployment (2 nodes) of OS Juno+XS6.5 and after I understood how it should work with Xen everything worked great. Still, we had to move for KVM because XS with local storage lacked support of dynamic volumes via Cinder. If you plan to use shared storage for VMs, XS is a great choice. It’s MUCH easier to maintain then KVM, especially for a newbie in such technologies Regards, IT engineer Farheap, Russia Ivan Derbenev From: Leandro Mendes [mailto:theflock...@gmail.com] Sent: Friday, July 24, 2015 5:15 PM To: Mārtiņš Jakubovičs Cc: openstack@lists.openstack.org Subject: Re: [Openstack] XenServer 6.5 or KVM Thank you guys. One last question: This 3 node setup is just a proof of concept of a bigger architecture (with around 30 nodes). I personally would prefer use KVM, so we can use Kilo and Neutron, but for some political reasons, we are really pushed to use XenServer (no problem about XS itself, I think it is great), but we don't have enough time and "hands" to make it work. Thinking about it,i would ask (maybe Bob will give me the best answer) what should be the best choise for use Openstack with XenServer regarding their versions? Maybe an older XCP with Juno or Icehouse? And what about VMWare as it is as XenServer in Group B? Thank you again. On Fri, Jul 24, 2015 at 4:20 AM, Mārtiņš Jakubovičs mailto:mart...@vertigs.lv>> wrote: Hello Leandro, From my experience OpenStack + XenServer is really difficult to build, but it work. Newest features like neutron, advanced cinder drivers, XenServer pool's, etc is not working well. At least with XenServer 6.2 and Icehouse release. If you know both technologies and don't know which to choose then use KVM. If you want to use XenServer then community member Bob Ball (bob.b...@citrix.com<mailto:bob.b...@citrix.com>) can help you more with this. Best regards, Martins On 2015.07.24. 05:03, Leandro Mendes wrote: Hi Guys, I'm trying to setup a 3 node (controller, neutron and compute) Openstack setup, but i'm having a lot of trouble regarding the XenServer(6.5) setup. Looking around the internet for docs about XenServer as compute node, i've got the impression that the main effort of OpenStack is use KVM instead of XenServer, because the docs (and lots of another links outhere) are outdated. May i consider it true? Should i forget about XenServer and move to KVM? What do you guys think about that? Thank you. ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] Adding a volume to existing instance
Hello, guys! That can be rather simple question for you, but I can't quite understand. For example, we have 2 hardware machines with Ubuntu+KVM+Kilo Each machine has 2 TB of local storage Shared storage on freenas can be used ONLY for images, not for VMs We created a Debian instance on server1, and now we need to attach a 100gb disk to it. And we want this disk to be created on the same server1. It doesn't really matter what backend we use, lvm\file image\anything else Well... it's easily ca be done by KVM itself with virsh attach-device. But how can we do this via OpenStack services? Cinder will use iscsi even inside single machine (I don't know whether it will be slower or not) and we can't predict what storage server it will use to create disk. And I couldn't find any way to do it inside nova. Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] New instances booting time
Hello, guys, I have a question We now have OS Kilo + KVM+ Ubuntu 14.04 Nova-compute.conf: [libvirt] virt_type=kvm images_type = lvm images_volume_group =openstack-controller01-ky01-vg volume_clear = none the problem is when nova boots a new instance it's SUPER slow exactly this step: Running cmd (subprocess): sudo nova-rootwrap /etc/nova/rootwrap.conf qemu-img convert -O raw /var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a19b42e /dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a662327b608c_disk Well, I understand what this step is doing - it's copying raw image to lvm. How can we speed it up? I don't wanna have instance from 100GB image booted for 4 hours Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] cinder and nova volumes at the same place
Hello! We have an installation of OS Kilo+Ubunto14.04+KVM Now we use LVM backend for both Nova and Cinder. I do want to switch it to file backend and store VMs in qcow2 files on ext4 fs. But us there a way to store cinder volumes in the same way? Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] New instances booting time
bump Regards, IT engineer Farheap, Russia Ivan Derbenev From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: Wednesday, August 5, 2015 1:21 PM To: openstack@lists.openstack.org Subject: [Openstack] New instances booting time Hello, guys, I have a question We now have OS Kilo + KVM+ Ubuntu 14.04 Nova-compute.conf: [libvirt] virt_type=kvm images_type = lvm images_volume_group =openstack-controller01-ky01-vg volume_clear = none the problem is when nova boots a new instance it's SUPER slow exactly this step: Running cmd (subprocess): sudo nova-rootwrap /etc/nova/rootwrap.conf qemu-img convert -O raw /var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a19b42e /dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a662327b608c_disk Well, I understand what this step is doing - it's copying raw image to lvm. How can we speed it up? I don't wanna have instance from 100GB image booted for 4 hours Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] New instances booting time
Well, thanks. Actually, wanted to know, if this behavior is ok for every installation, or I has misconfigured smthing Regards, IT engineer Farheap, Russia Ivan Derbenev -Original Message- From: Daniel P. Berrange [mailto:berra...@redhat.com] Sent: Friday, August 14, 2015 5:56 PM To: Jay Pipes Cc: openstack@lists.openstack.org Subject: Re: [Openstack] New instances booting time On Fri, Aug 14, 2015 at 03:34:09PM +0100, Daniel P. Berrange wrote: > On Fri, Aug 14, 2015 at 03:10:13PM +0100, Daniel P. Berrange wrote: > > On Fri, Aug 14, 2015 at 09:55:42AM -0400, Jay Pipes wrote: > > > On 08/13/2015 11:37 PM, Ivan Derbenev wrote: > > > >*From:*Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] > > > >*Sent:* Wednesday, August 5, 2015 1:21 PM > > > >*To:* openstack@lists.openstack.org > > > >*Subject:* [Openstack] New instances booting time > > > > > > > >Hello, guys, I have a question > > > > > > > >We now have OS Kilo + KVM+ Ubuntu 14.04 > > > > > > > >Nova-compute.conf: > > > > > > > >[libvirt] > > > > > > > >virt_type=kvm > > > > > > > >images_type = lvm > > > > > > > >images_volume_group =openstack-controller01-ky01-vg > > > > > > > >volume_clear = none > > > > > > > >the problem is when nova boots a new instance it’s SUPER slow > > > > > > > >exactly this step: > > > > > > > >Running cmd (subprocess): sudo nova-rootwrap > > > >/etc/nova/rootwrap.conf qemu-img convert -O raw > > > >/var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a1 > > > >9b42e > > > >/dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a6623 > > > >27b608c_disk > > > > > > > >Well, I understand what this step is doing – it’s copying raw image to > > > >lvm. > > > > > > > >How can we speed it up? > > > > > > > >I don’t wanna have instance from 100GB image booted for 4 hours > > > > > > Don't use base images that are 100G in size. Quite simple, really. > > > > Presumably the actual image is only a few 100 MB in size, but the > > virtual qcow2 size is 100 GB ? > > > > I guess qemu-img probably has to still write out 100 GB of zeros as > > it can't assume that the LVM target is pre-zeroed. I do wonder > > though if there's a way to optimize this so that qemu-img only has > > to write out the 100 MB of actual data, and not al the zeros. > > > > Perhaps there's scope for a flag to qemu-img to tell it to seek in > > the target when it sees holes in the source image. If OpenStack then > > used sparse LVM images, there would be no need to rwite out 100's of > > GB of zeros. > > Speaking with the KVM block layer maintainers, the behaviour seen is > not entirely expected. qemu-img wants to ensure no existing data is > left behind as that could cause data corruption if the original qcow2 > file had dropped empty sectors. So it has to make sure it can write > out the 100 GB even if that is mostly zeros. > > That said, it has support for using discard, and it will use that in > preference to writing zeros, which should be really fast. For it to > work with block devices though, it seems that we need to disable the > I/O cache by passing '-t none' to qemu-img when doing this conversion. > This also assumes the underlying storage that the user has configured > behind their LVM volume supports discard which may not always be the > case. > > So it would be interesting to know if changing > > $ qemu-img convert -O raw \ > /var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a19b42e \ > > /dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a662327b60 > 8c_disk > > to > > $ qemu-img convert -O raw -t none \ > /var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a19b42e \ > > /dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a662327b60 > 8c_disk > > makes any significant difference to the speed. I filed a bug to track this possible enhancement https://bugs.launchpad.net/nova/+bug/1484992 Regards, Daniel -- |: http://berrange.com -o-http://www.flickr.com/photos/dberrange/ :| |: http://libvirt.org -o- http://virt-manager.org :| |: http://autobuild.org -o- http://search.cpan.org/~danberr/ :| |: http://entangle-photo.org -o- http://live.gnome.org/gtk-vnc :| ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] New instances booting time
Well, I have 4gb qcow2 image But it contains a 100gb filesystem And the rest 96 gbs are still copied Regards, IT engineer Farheap, Russia Ivan Derbenev -Original Message- From: Jay Pipes [mailto:jaypi...@gmail.com] Sent: Friday, August 14, 2015 4:56 PM To: openstack@lists.openstack.org Subject: Re: [Openstack] New instances booting time On 08/13/2015 11:37 PM, Ivan Derbenev wrote: > *From:*Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] > *Sent:* Wednesday, August 5, 2015 1:21 PM > *To:* openstack@lists.openstack.org > *Subject:* [Openstack] New instances booting time > > Hello, guys, I have a question > > We now have OS Kilo + KVM+ Ubuntu 14.04 > > Nova-compute.conf: > > [libvirt] > > virt_type=kvm > > images_type = lvm > > images_volume_group =openstack-controller01-ky01-vg > > volume_clear = none > > the problem is when nova boots a new instance it's SUPER slow > > exactly this step: > > Running cmd (subprocess): sudo nova-rootwrap /etc/nova/rootwrap.conf > qemu-img convert -O raw > /var/lib/nova/instances/_base/999f7fff2521e4a7243c9e1d21599fd64a19b42e > /dev/openstack-controller01-ky01-vg/5f831046-435c-4636-8b71-a662327b60 > 8c_disk > > Well, I understand what this step is doing - it's copying raw image to lvm. > > How can we speed it up? > > I don't wanna have instance from 100GB image booted for 4 hours Don't use base images that are 100G in size. Quite simple, really. Best, -jay ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] Problems with OpenStack and LDAP
You need to create service users in ldap ADMIN_TOKEN should work for assigning roles Regards, IT engineer Farheap, Russia Ivan Derbenev -Original Message- From: Marc Pape [mailto:marc.p...@gmail.com] Sent: Monday, August 17, 2015 10:32 AM To: openstack@lists.openstack.org Subject: [Openstack] Problems with OpenStack and LDAP Hello everybody, i've got some problems with our OpenStack (Juno) and the Integrate Identity Service over LDAP. The LDAP connection is read only, so i configured the [identity], [ldap] and [assignment] parts in keystone conf. The identity part use "driver = keystone.identity.backends.ldap.Identity" and assignment "driver = keystone.assignment.backends.sql.Assignment" Our goal is a user authentication via LDAP and project assignment in the internal SQL . It would be great if the service users of OpenStack are also stored in SQL, but they are also currently in the LDAP deposited. After restarting the Keystone Service authentication via LDAP is possible. The user get the message that no projects assigned to him. Now there are wto problems. How can you log in as admin to assign projects and keystone said that it couldn't find the service user like ceilometer, neutron and so on. I've followed the instructions on docs.openstack.org for Identity management, but i didn't find any notices about that problems. Many greetings and thanks for a possible answer Marc ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] resize lvm-backed instance
Hello! Is it possible to resize (and, possibly, migrate) lvm-backed instances in kilo? When I try to do so, I am getting an error Instance rollback performed due to: Migration pre-check error: Migration is not supported for LVM backed instances What's the workaround? Yes, I can go into DB manually every time, but I don't like this Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] resize lvm-backed instance
bump Regards, IT engineer Farheap, Russia Ivan Derbenev From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: Tuesday, October 13, 2015 1:53 PM To: openstack@lists.openstack.org Subject: [Openstack] resize lvm-backed instance Hello! Is it possible to resize (and, possibly, migrate) lvm-backed instances in kilo? When I try to do so, I am getting an error Instance rollback performed due to: Migration pre-check error: Migration is not supported for LVM backed instances What's the workaround? Yes, I can go into DB manually every time, but I don't like this Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] work from home 19.10.2015
Ввиду проведения санехнических работы вынужден поработать из дома первую половину дня Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] no dhcp record for new vm
Hello, guys, we have the newest kilo release (2015.1.2) with nova-network And we've got a problem with new vms - they can't get new ip until nova-network is restarted /var/lib/nova/networks/nova-br1004.conf file isn't reloaded, so there is no new record. This happens only on compute nodes, on controller nova-br1004.conf updates at the same time when the instance is booted I tried to debug the code to find when this should happen and why it doesn't, but no result Help is deeply appreciated, thanks Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] single dnsmasq instance in cloud
Hello, guys! We do have installation of kilo + nova-network Is it possible to have only one dnsmasq dhcp server on controller node, instead of every-node installation? Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] public snapshots via horizon/nova
Hello! Is there a way to make a public snapshot via horizon? Kilo, Ubuntu 14.04 Not to change metadata for already-created snapshot, but to make it public immediately after creation? Or at least how can I create public snapshot via #nova image-create (again, not change property in glance, but via nova) Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] [Nova] [Glance] Errors Produced by Nova and Glance Interaction
Looks like this is your issue https://bugs.launchpad.net/glance/+bug/1476770 Regards, IT engineer Farheap, Russia Ivan Derbenev From: Richard Raseley [mailto:rich...@raseley.com] Sent: Friday, December 4, 2015 5:50 AM To: openstack@lists.openstack.org Subject: [Openstack] [Nova] [Glance] Errors Produced by Nova and Glance Interaction I am running OpenStack Juno installed from RDO packages, all version 2015.1.1-1.el7, on CentOS 7. I am experiencing an issue whenever Nova has to interact with Glance wherein it is unable to do anything other than retrieve the list of available images. Here is an example: ``` (openstack)➜ openstack glance image-list +--+-+ | ID | Name| +--+-+ | 70bcde32-dcfc-47f6-8596-9d968fae2011 | cirros-0.3.4-x86_64 | +--+-+ (openstack)➜ openstack glance image-show 70bcde32-dcfc-47f6-8596-9d968fae2011 +--+--+ | Property | Value| +--+--+ | architecture | x86_64 | | checksum | ee1eca47dc88f4879d8a229cc70a07c6 | | container_format | bare | | created_at | 2015-12-03T15:10:22Z | | description | CirrOS 0.3.4 x86_64 | | disk_format | qcow2| | id | 70bcde32-dcfc-47f6-8596-9d968fae2011 | | min_disk | 1| | min_ram | 512 | | name | cirros-0.3.4-x86_64 | | owner| bd712506c2ba41f2b1d661f97c927f8a | | protected| False| | size | 13287936 | | status | active | | tags | [] | | updated_at | 2015-12-03T15:10:26Z | | virtual_size | None | | visibility | private | +--+--+ (openstack)➜ openstack nova image-list +--+-+++ | ID | Name| Status | Server | +--+-+++ | 70bcde32-dcfc-47f6-8596-9d968fae2011 | cirros-0.3.4-x86_64 | ACTIVE || +--+-+++ (openstack)➜ openstack nova image-show 70bcde32-dcfc-47f6-8596-9d968fae2011 ERROR (ClientException): The server has either erred or is incapable of performing the requested operation. (HTTP 500) (Request-ID: req-f27a5253-dc65-4689-9d89-f80ddc873e37) ``` This is causing the inability to launch instances. Whenever this occurs, the Glance API logs look clean and I see a 200 response to each request, however the Nova API logs show an ‘AttributeError’ (http://paste.openstack.org/show/480853/) for each request. I had originally thought this was related to this bug (https://bugs.launchpad.net/glance/+bug/1476770), but am still experiencing it with keystonemiddleware 1.5.3, requests 2.8.1, and urllib3 1.12. Any assistance would be greatly appreciated, as this is currently a blocking issue for me. Regards, Richard ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] change IP of instance
Hello! Openstack kilo+Ubuntu 14.04 Nova-network, not neutron, vlanDHCPmanager Is there a possibility to change ip of instance, without database hacking? Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] resize lvm-backed instance
Wow, I completely missed this answer, and now I do have the same problem I looked at code and found that LVM migration is not implemented, however, I don't wanna migrate, I want to resize Here's log I'm getting 2016-02-14 14:38:45.016 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Total usable vcpus: 8, total allocated vcpus: 13 2016-02-14 14:38:45.017 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Final resource view: name=openstack04 phys_ram=32175MB used_ram=32256MB phys_disk=837GB used_disk=104GB total_vcpus=8 used_vcpus=13 pci_stats=None 2016-02-14 14:38:45.060 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Compute_service record updated for openstack04:openstack04 2016-02-14 14:38:45.061 25601 DEBUG oslo_concurrency.lockutils [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Lock "compute_resources" released by "nova.compute.resource_tracker._update_available_resource" :: held 0.197s inner /usr/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:265 2016-02-14 14:38:51.528 25601 DEBUG oslo_service.periodic_task [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Running periodic task ComputeManager._heal_instance_info_cache run_periodic_tasks /usr/lib/python2.7/dist-packages/oslo_service/periodic_task.py:213 2016-02-14 14:38:51.529 25601 DEBUG nova.compute.manager [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Starting heal instance info cache _heal_instance_info_cache /usr/lib/python2.7/dist-packages/nova/compute/manager.py:5478 2016-02-14 14:38:51.612 25601 DEBUG oslo_concurrency.lockutils [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Acquired semaphore "refresh_cache-3e2836b6-602a-44de-a835-9d1936a53785" lock /usr/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:198 2016-02-14 14:38:51.613 25601 DEBUG nova.objects.instance [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Lazy-loading `flavor' on Instance uuid 3e2836b6-602a-44de-a835-9d1936a53785 obj_load_attr /usr/lib/python2.7/dist-packages/nova/objects/instance.py:860 2016-02-14 14:38:51.754 25601 DEBUG nova.network.base_api [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] [instance: 3e2836b6-602a-44de-a835-9d1936a53785] Updating instance_info_cache with network_info: [VIF({'profile': None, 'ovs_interfaceid': None, 'preserve_on_delete': False, 'network': Network({'bridge': u'br1004', 'subnets': [Subnet({'ips': [FixedIP({'meta': {}, 'version': 4, 'type': u'fixed', 'floating_ips': [], 'address': u'10.1.10.49'})], 'version': 4, 'meta': {u'dhcp_server': u'10.1.10.3'}, 'dns': [IP({'meta': {}, 'version': 4, 'type': u'dns', 'address': u'8.8.4.4'}), IP({'meta': {}, 'version': 4, 'type': u'dns', 'address': u'10.1.2.100'})], 'routes': [], 'cidr': u'10.1.10.0/24', 'gateway': IP({'meta': {}, 'version': 4, 'type': u'gateway', 'address': u'10.1.10.3'})}), Subnet({'ips': [], 'version': None, 'meta': {u'dhcp_server': u'10.1.10.3'}, 'dns': [], 'routes': [], 'cidr': None, 'gateway': IP({'meta': {}, 'version': None, 'type': u'gateway', 'address': None})})], 'meta': {u'multi_host': True, u'should_create_bridge': True, u'should_create_vlan': True, u'bridge_interface': u'eth1', u'tenant_id': u'2b1b8aa01abf4db8ab9aa8c22cf861f0', u'vlan': 1004}, 'id': u'9540bfd0-e620-4302-bb1e-dfe69ad9bcda', 'label': u'tcdefault-1004'}), 'devname': None, 'vnic_type': u'normal', 'qbh_params': None, 'meta': {}, 'details': {}, 'address': u'fa:16:3e:eb:41:69', 'active': False, 'type': u'bridge', 'id': u'd9171467-de05-4a99-aaab-4d0d57343658', 'qbg_params': None})] update_instance_cache_with_nw_info /usr/lib/python2.7/dist-packages/nova/network/base_api.py:43 2016-02-14 14:38:51.769 25601 DEBUG oslo_concurrency.lockutils [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Releasing semaphore "refresh_cache-3e2836b6-602a-44de-a835-9d1936a53785" lock /usr/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:211 2016-02-14 14:38:51.770 25601 DEBUG nova.compute.manager [req-73be063c-
Re: [Openstack] resize lvm-backed instance
bump Regards, IT engineer Farheap, Russia Ivan Derbenev -Original Message- From: Ivan Derbenev [mailto:ivan.derbe...@tech-corps.com] Sent: Sunday, February 14, 2016 10:54 PM To: Chris Friesen ; openstack@lists.openstack.org Subject: Re: [Openstack] resize lvm-backed instance Wow, I completely missed this answer, and now I do have the same problem I looked at code and found that LVM migration is not implemented, however, I don't wanna migrate, I want to resize Here's log I'm getting 2016-02-14 14:38:45.016 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Total usable vcpus: 8, total allocated vcpus: 13 2016-02-14 14:38:45.017 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Final resource view: name=openstack04 phys_ram=32175MB used_ram=32256MB phys_disk=837GB used_disk=104GB total_vcpus=8 used_vcpus=13 pci_stats=None 2016-02-14 14:38:45.060 25601 INFO nova.compute.resource_tracker [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Compute_service record updated for openstack04:openstack04 2016-02-14 14:38:45.061 25601 DEBUG oslo_concurrency.lockutils [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Lock "compute_resources" released by "nova.compute.resource_tracker._update_available_resource" :: held 0.197s inner /usr/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:265 2016-02-14 14:38:51.528 25601 DEBUG oslo_service.periodic_task [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Running periodic task ComputeManager._heal_instance_info_cache run_periodic_tasks /usr/lib/python2.7/dist-packages/oslo_service/periodic_task.py:213 2016-02-14 14:38:51.529 25601 DEBUG nova.compute.manager [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Starting heal instance info cache _heal_instance_info_cache /usr/lib/python2.7/dist-packages/nova/compute/manager.py:5478 2016-02-14 14:38:51.612 25601 DEBUG oslo_concurrency.lockutils [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Acquired semaphore "refresh_cache-3e2836b6-602a-44de-a835-9d1936a53785" lock /usr/lib/python2.7/dist-packages/oslo_concurrency/lockutils.py:198 2016-02-14 14:38:51.613 25601 DEBUG nova.objects.instance [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] Lazy-loading `flavor' on Instance uuid 3e2836b6-602a-44de-a835-9d1936a53785 obj_load_attr /usr/lib/python2.7/dist-packages/nova/objects/instance.py:860 2016-02-14 14:38:51.754 25601 DEBUG nova.network.base_api [req-73be063c-747a-4933-bef3-31a19c10f278 - - - - -] [instance: 3e2836b6-602a-44de-a835-9d1936a53785] Updating instance_info_cache with network_info: [VIF({'profile': None, 'ovs_interfaceid': None, 'preserve_on_delete': False, 'network': Network({'bridge': u'br1004', 'subnets': [Subnet({'ips': [FixedIP({'meta': {}, 'version': 4, 'type': u'fixed', 'floating_ips': [], 'address': u'10.1.10.49'})], 'version': 4, 'meta': {u'dhcp_server': u'10.1.10.3'}, 'dns': [IP({'meta': {}, 'version': 4, 'type': u'dns', 'address': u'8.8.4.4'}), IP({'meta': {}, 'version': 4, 'type': u'dns', 'address': u'10.1.2.100'})], 'routes': [], 'cidr': u'10.1.10.0/24', 'gateway': IP({'meta': {}, 'version': 4, 'type': u'gateway', 'address': u'10.1.10.3'})}), Subnet({'ips': [], 'version': None, 'meta': {u'dhcp_server': u'10.1.10.3'}, 'dns': [], 'routes': [], 'cidr': None, 'gateway': IP({'meta': {}, 'version': None, 'type': u'gateway', 'address': None})})], 'meta': {u'multi_host': True, u'should_create_bridge': True, u'should_create_vlan': True, u'bridge_interface': u'eth1', u'tenant_id': u'2b1b8aa01abf4db8ab9aa8c22cf861f0', u'vlan': 1004}, 'id': u'9540bfd0-e620-4302-bb1e-dfe69ad9bcda', 'label': u'tcdefault-1004'}), 'devname': None, 'vnic_type': u'normal', 'qbh_params': None, 'meta': {}, 'details': {}, 'address': u'fa:16:3e:eb:41:69', 'active': False, 'type': u'bridge', 'id': u'd9171467-de05-4a99-aaab-4d0d57343658', 'qbg_params': None})] update_instance_cache_with_nw_info /usr/lib/python2.7/dist-packages/nova/network/base_api.py:43 2016-02-14 14:38:51.769 25601 DEBUG oslo_concurrency.locku
[Openstack] automatic role assignment
Hello! We are using OS liberty + Ubuntu LTS. I configured keystone to use ldap as an identity backend and sql as an assignment backend. Is there an automatic built-in way to assign some roles to all people from specific LDAP group? Regards, IT engineer Farheap, Russia Ivan Derbenev ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] instances with nfs-backed cinder volumes hanges after network fail
Hello guys! We have currently an installation of Openstack Liberty+KVM+Ububntu 14.04, mostly default Cinder and Nova are backed with qcow2 over NFS, cinder volumes resides on the same servers running nova-compute The problem is - we do have power issues in our building, and not always we can turn off servers properly, when this happens So, when the power is down, and network is down, all instances that has cinder volumes connected hangs. Simply nfs share becomes unavailable, even if the share is mounted locally (from 192.168.1.29 onto 192.168.1.29, for example), I can't even ls there, it hangs. And it never goes up, until the server reboot, even if the network is back. I can't even restart instance from KVM locally, it says that device is busy So, every instance which has cinder volume mounted hangs and wouldn't go up until hypervisor rebooted Is there a way to fix this issue? At least make instances with local cinder volumes (the same host where nova runs this instance) not hang? ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] allow user to run single api call
Hello all! Is it possible to give a user (or role) an ability to run specific api call? It's monitoring - only user, and I want to give it permissions for all %servicename% %itemname%-list calls And changing specific policies in policy.json seems to work, but not for things like nova/cinder service-list. So I can run service-list only when the user is admin (or after I changed context is_admin in policy.json) Can I somehow allow user to run ONLY nova service-list? ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
[Openstack] Openstack+KVM+overcommit, VM priority
Hello, guys! Imagine we have a compute node with kvm hypervisor installed It has 64gb RAM and quad core processor We create 2 machines in nova on this host - both with 64gb and 4 VCPUs if both vms start using all 64gb memory, both of them start using swap same for cpu - they use it equally So, the question is - is it possible to prioritize 1st vm above 2nd? so the second one will fail before the 1st, to leave maximum possible perfomance to the most importan one? like production and secondary services running on the same node. ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] Openstack+KVM+overcommit, VM priority
no, that's not my question. I'm already overcommiting, but now need to prioritize one instance above others in terms of perfomance. От: Eugene Nikanorov Отправлено: 12 января 2017 г. 1:13:27 Кому: Ivan Derbenev; openstack@lists.openstack.org Тема: Re: [Openstack] Openstack+KVM+overcommit, VM priority Ivan, see if it provides an answer: https://ask.openstack.org/en/question/55307/overcommitting-value-in-novaconf/ Regards, Eugene. On Wed, Jan 11, 2017 at 1:55 PM, James Downs mailto:e...@egon.cc>> wrote: On Wed, Jan 11, 2017 at 09:34:32PM +0000, Ivan Derbenev wrote: > if both vms start using all 64gb memory, both of them start using swap Don't overcommit RAM. > So, the question is - is it possible to prioritize 1st vm above 2nd? so the > second one will fail before the 1st, to leave maximum possible perfomance to > the most importan one? Do you mean CPU prioritization? There are facilities to allow one VM or another to have CPU priority, but what, if a high priority VM wants RAM, you want to OOM the other? That doesn't exist, AFAIK. Cheers, -j ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org<mailto:openstack@lists.openstack.org> Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack
Re: [Openstack] Openstack+KVM+overcommit, VM priority
What are the facilities for cpu? That's the initial question - how can I do it if this is possible Best Regards Tech-corps IT Engineer Ivan Derbenev Phone: +79633431774 -Original Message- From: James Downs [mailto:e...@egon.cc] Sent: Thursday, January 12, 2017 12:56 AM To: Ivan Derbenev Cc: openstack@lists.openstack.org Subject: Re: [Openstack] Openstack+KVM+overcommit, VM priority On Wed, Jan 11, 2017 at 09:34:32PM +0000, Ivan Derbenev wrote: > if both vms start using all 64gb memory, both of them start using swap Don't overcommit RAM. > So, the question is - is it possible to prioritize 1st vm above 2nd? so the > second one will fail before the 1st, to leave maximum possible perfomance to > the most importan one? Do you mean CPU prioritization? There are facilities to allow one VM or another to have CPU priority, but what, if a high priority VM wants RAM, you want to OOM the other? That doesn't exist, AFAIK. Cheers, -j ___ Mailing list: http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack Post to : openstack@lists.openstack.org Unsubscribe : http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack