Re: [pve-devel] lxc 1.1 : qcow2 support ?
not sure if we should use that? One advantage, It's possible to do live migration with qcow2 on top of nfs. (Note that this need to be tested with nbd, i'm not sure it's work fine with cache flush,...) What other possibilities do we have to create snapshots on normal file systems (for example nfs). with nfs, the only way is that the storage support snapshot api (like netapp for example) Maybe using a loop mounted image with btrfs? It should work. But with btrfs maybe it is better to use sub volume ? https://btrfs.wiki.kernel.org/index.php/Manpage/btrfs-subvolume - Mail original - De: dietmar diet...@proxmox.com À: aderumier aderum...@odiso.com Cc: pve-devel pve-devel@pve.proxmox.com, Kamil Trzciński ayu...@ayufan.eu Envoyé: Samedi 25 Avril 2015 07:50:15 Objet: Re: [pve-devel] lxc 1.1 : qcow2 support ? It's seem also possible to use lxc.rootfs.mount = nbd:file:myfile.qcow2 I finally implemented qemu-nbd using our own create code, see https://git.proxmox.com/?p=pve-container.git;a=commitdiff;h=6ed8c6dd2b4307cd077b623bee49fca13d77d62e not sure if we should use that? What other possibilities do we have to create snapshots on normal file systems (for example nfs). Maybe using a loop mounted image with btrfs? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
On Sun, 26 Apr 2015 19:14:12 +0200 (CEST) Dietmar Maurer diet...@proxmox.com wrote: My current thinking is that we should only use native storage feature, i.e. simply no snapshots for NFS. For btrfs, zfs, lvm (thin), sheepdog, ceph, drbd we can use native snapshots. Those storage types also provides disk quotas. We can use a loop mount to provide disk quota for NFS, ext3, ext4, xfs. What do you think? I think this is a reasonable decision. I users wants to be able to make snapshots the should choose a proper storage and file system. Implementing self-constructed hacks is just a waiting logical bomb. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael at rasmussen dot cc http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xD3C9A00E mir at datanom dot net http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xE501F51C mir at miras dot org http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xE3E80917 -- /usr/games/fortune -es says: For adult education nothing beats children. pgpr8K1uoIKbB.pgp Description: OpenPGP digital signature ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
not sure if we should use that? One advantage, It's possible to do live migration with qcow2 on top of nfs. (Note that this need to be tested with nbd, i'm not sure it's work fine with cache flush,...) I just notice that it is not even possible to make s snapshot with qemu-nbd. That makes using qcow2 files quite useless ... What other possibilities do we have to create snapshots on normal file systems (for example nfs). with nfs, the only way is that the storage support snapshot api (like netapp for example) Maybe using a loop mounted image with btrfs? It should work. But with btrfs maybe it is better to use sub volume ? https://btrfs.wiki.kernel.org/index.php/Manpage/btrfs-subvolume Sure, if the file system is local. But what if we use NFS? My current thinking is that we should only use native storage feature, i.e. simply no snapshots for NFS. For btrfs, zfs, lvm (thin), sheepdog, ceph, drbd we can use native snapshots. Those storage types also provides disk quotas. We can use a loop mount to provide disk quota for NFS, ext3, ext4, xfs. What do you think? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
I am unsure if this is a good idea, because lxc-create does more things :) for example? Most things I see are sooo much simpler with perl. ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
I am unsure if this is a good idea, because lxc-create does more things :) Looking at source code https://github.com/lxc/lxc/blob/master/src/lxc/lxc_create.c It doesn't do too much thing. I agree with Diemar, I can be easy to implement this in perl. (and possible to use qemu-img to create qcow2 file for example, ceph rbd, ... maybe reuse pve-storage ? ) - Mail original - De: Kamil Trzciński ayu...@ayufan.eu À: dietmar diet...@proxmox.com Cc: aderumier aderum...@odiso.com, pve-devel pve-devel@pve.proxmox.com Envoyé: Vendredi 24 Avril 2015 10:54:54 Objet: Re: [pve-devel] lxc 1.1 : qcow2 support ? I am unsure if this is a good idea, because lxc-create does more things :) Maybe it will be better to implement this method (or even add own storage): https://github.com/lxc/lxc/blob/master/src/lxc/bdev.c#L3090 On Fri, Apr 24, 2015 at 9:18 AM, Dietmar Maurer diet...@proxmox.com wrote: I am quit unsure if this lxc-create binary is what we want to use. I try to write my own create script now, hoping I can handle different storage types easier. On April 24, 2015 at 8:45 AM Alexandre DERUMIER aderum...@odiso.com wrote: Yes, but this is about 'lxc-start', not 'lxc-create'. mmm, I'm not familiar with lxc yet. what does exactly lxc-create ? from this doc https://linuxcontainers.org/fr/lxc/manpages/man5/lxc.container.conf.5.html It's seem also possible to use lxc.rootfs.mount = nbd:file:myfile.qcow2 does it help ? -- Kamil Trzciński ayu...@ayufan.eu www.ayufan.eu ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
I am unsure if this is a good idea, because lxc-create does more things :) Maybe it will be better to implement this method (or even add own storage): https://github.com/lxc/lxc/blob/master/src/lxc/bdev.c#L3090 On Fri, Apr 24, 2015 at 9:18 AM, Dietmar Maurer diet...@proxmox.com wrote: I am quit unsure if this lxc-create binary is what we want to use. I try to write my own create script now, hoping I can handle different storage types easier. On April 24, 2015 at 8:45 AM Alexandre DERUMIER aderum...@odiso.com wrote: Yes, but this is about 'lxc-start', not 'lxc-create'. mmm, I'm not familiar with lxc yet. what does exactly lxc-create ? from this doc https://linuxcontainers.org/fr/lxc/manpages/man5/lxc.container.conf.5.html It's seem also possible to use lxc.rootfs.mount = nbd:file:myfile.qcow2 does it help ? -- Kamil Trzciński ayu...@ayufan.eu www.ayufan.eu ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
Maybe ship modprobe.d with options loop max_loop=? On Fri, Apr 24, 2015 at 5:05 PM, Dietmar Maurer diet...@proxmox.com wrote: Looking at source code https://github.com/lxc/lxc/blob/master/src/lxc/lxc_create.c It doesn't do too much thing. I agree with Diemar, I can be easy to implement this in perl. (and possible to use qemu-img to create qcow2 file for example, ceph rbd, ... maybe reuse pve-storage ? ) That is exactly the point - I want to reuse code from pve-storage. I already have a working prototype, but I need to do more tests, especially with unprivileged containers ... OK, here is the final code: https://git.proxmox.com/?p=pve-container.git;a=blob;f=src/PVE/LXCCreate.pm;h=7366be460488fe0cf1266802a43c2dbe332fc31d;hb=5b4657d05eab58d04a41e502c03b7f8f23b60bab I implemented the loop mount case as example. Not sure if I miss something, but this also works with unprivileged containers. But loop devices are limited - 7 by default, so that is not really usable. Is there a way to use more loop devices dynamically? -- Kamil Trzciński ayu...@ayufan.eu www.ayufan.eu ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
Looking at source code https://github.com/lxc/lxc/blob/master/src/lxc/lxc_create.c It doesn't do too much thing. I agree with Diemar, I can be easy to implement this in perl. (and possible to use qemu-img to create qcow2 file for example, ceph rbd, ... maybe reuse pve-storage ? ) That is exactly the point - I want to reuse code from pve-storage. I already have a working prototype, but I need to do more tests, especially with unprivileged containers ... OK, here is the final code: https://git.proxmox.com/?p=pve-container.git;a=blob;f=src/PVE/LXCCreate.pm;h=7366be460488fe0cf1266802a43c2dbe332fc31d;hb=5b4657d05eab58d04a41e502c03b7f8f23b60bab I implemented the loop mount case as example. Not sure if I miss something, but this also works with unprivileged containers. But loop devices are limited - 7 by default, so that is not really usable. Is there a way to use more loop devices dynamically? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
It's seem also possible to use lxc.rootfs.mount = nbd:file:myfile.qcow2 I finally implemented qemu-nbd using our own create code, see https://git.proxmox.com/?p=pve-container.git;a=commitdiff;h=6ed8c6dd2b4307cd077b623bee49fca13d77d62e not sure if we should use that? What other possibilities do we have to create snapshots on normal file systems (for example nfs). Maybe using a loop mounted image with btrfs? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
On Fri, 24 Apr 2015 18:25:20 +0200 (CEST) Dietmar Maurer diet...@proxmox.com wrote: how many loop devices are possible? If I remember correct the maximum number of loop devices is 256. You asked in another post if loop devices can be dynamically configured. The answer is yes. search for /dev/loop-control. -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael at rasmussen dot cc http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xD3C9A00E mir at datanom dot net http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xE501F51C mir at miras dot org http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0xE3E80917 -- /usr/games/fortune -es says: We come to bury DOS, not to praise it. (Paul Vojta, vo...@math.berkeley.edu, paraphrasing a quote of Shakespeare) pgpRnD1zmUlm7.pgp Description: OpenPGP digital signature ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
how many loop devices are possible? If I remember correct the maximum number of loop devices is 256. OK You asked in another post if loop devices can be dynamically configured. The answer is yes. search for /dev/loop-control. Yes, I just found that in the kernel docs. But lxc-start does not support it ... ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
On April 24, 2015 at 5:32 PM Kamil Trzciński ayu...@ayufan.eu wrote: Maybe ship modprobe.d with options loop max_loop=? how many loop devices are possible? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
I am quit unsure if this lxc-create binary is what we want to use. I try to write my own create script now, hoping I can handle different storage types easier. On April 24, 2015 at 8:45 AM Alexandre DERUMIER aderum...@odiso.com wrote: Yes, but this is about 'lxc-start', not 'lxc-create'. mmm, I'm not familiar with lxc yet. what does exactly lxc-create ? from this doc https://linuxcontainers.org/fr/lxc/manpages/man5/lxc.container.conf.5.html It's seem also possible to use lxc.rootfs.mount = nbd:file:myfile.qcow2 does it help ? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
Just found a new here about lxc 1.1: https://linuxcontainers.org/fr/lxc/news/ * core: Add support for qcow2 images (through qemu-img) so, maybe disk quota is possible ? (size of qcow2) Does somebody have an example howto use that? I tried: # qemu-image create test.img 4G # mkfs.ext4 test.img but what arguments do we need to pass to lxc-create? # lxc-create -n test -t debian any hints? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
from : https://github.com/hmalphettes/ubuntu-uncloud-recloud/blob/master/lxc/lxc-example.conf sudo qemu-nbd -c /dev/nbd0 /path/to/vmdk sudo lxc-start -n myvminlxc -f lxc-example.conf with lxc-example.conf lxc.rootfs = /dev/nbd0p1 - Mail original - De: dietmar diet...@proxmox.com À: Kamil Trzciński ayu...@ayufan.eu Cc: aderumier aderum...@odiso.com, pve-devel pve-devel@pve.proxmox.com Envoyé: Jeudi 23 Avril 2015 19:05:53 Objet: Re: [pve-devel] lxc 1.1 : qcow2 support ? lxc-create -t debian -n test-loop-5 -B loop --fstype ext3 --fssize 1G It should create: /var/lib/lxc/test-loop-5/rootdev This is loop device. But how does it work using nbd? ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
https://github.com/hmalphettes/ubuntu-uncloud-recloud/blob/master/lxc/lxc-example.conf sudo qemu-nbd -c /dev/nbd0 /path/to/vmdk sudo lxc-start -n myvminlxc -f lxc-example.conf with lxc-example.conf lxc.rootfs = /dev/nbd0p1 Yes, but this is about 'lxc-start', not 'lxc-create'. ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
[pve-devel] lxc 1.1 : qcow2 support ?
Just found a new here about lxc 1.1: https://linuxcontainers.org/fr/lxc/news/ * core: Add support for qcow2 images (through qemu-img) so, maybe disk quota is possible ? (size of qcow2) other news: * core: Add support for Openvswitch bridges * init: Add identical support of systemd, upstart and sysvinit scripts * core: Add support for checkpoint and restore of containers using CRIU Alexandre Derumier Ingénieur système et stockage Fixe : 03 20 68 90 88 Fax : 03 20 68 90 81 45 Bvd du Général Leclerc 59100 Roubaix 12 rue Marivaux 75002 Paris MonSiteEstLent.com - Blog dédié à la webperformance et la gestion de pics de trafic ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
It seem that lxc support also other backend like zfs for example. qcow2,raw file seem also great. (convert kvm - lxc with qemu-img seem easy). - Mail original - De: aderumier aderum...@odiso.com À: pve-devel pve-devel@pve.proxmox.com Envoyé: Lundi 30 Mars 2015 09:08:45 Objet: [pve-devel] lxc 1.1 : qcow2 support ? Just found a new here about lxc 1.1: https://linuxcontainers.org/fr/lxc/news/ * core: Add support for qcow2 images (through qemu-img) so, maybe disk quota is possible ? (size of qcow2) other news: * core: Add support for Openvswitch bridges * init: Add identical support of systemd, upstart and sysvinit scripts * core: Add support for checkpoint and restore of containers using CRIU Alexandre Derumier Ingénieur système et stockage Fixe : 03 20 68 90 88 Fax : 03 20 68 90 81 45 Bvd du Général Leclerc 59100 Roubaix 12 rue Marivaux 75002 Paris MonSiteEstLent.com - Blog dédié à la webperformance et la gestion de pics de trafic ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
Re: [pve-devel] lxc 1.1 : qcow2 support ?
On March 30, 2015 at 8:00 PM Alexandre DERUMIER aderum...@odiso.com wrote: It seem that lxc support also other backend like zfs for example. qcow2,raw file seem also great. (convert kvm - lxc with qemu-img seem easy). yes, looks good. ___ pve-devel mailing list pve-devel@pve.proxmox.com http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
