Search results

what is your cpu usage during the bench ? currently iothread use only 1core, so you could be cpu limited in rand4k. (I'm working on it to add support for the new multithreading feature ). It could explain the difference, if hyper-v is able to use multiple cores by disk.

It should come in the future with proxmox datacenter manager. (2 differents cluster + storage replication ). Not sure It'll be ready for 1.0 , but I'm pretty sure that the feature will be available in 2016. For now, you need to replicate the storage yourself (depending of the storage techno...

if your fleecing storage is full, sure, you'll got io-error. But it's helping too to avoid slowdown during backup if your pbs is slow. for example, production is on nvme, and pbs is on hdd. You'll have slowdown when backup is running if write occur on a block not yet saved. With fleecing, if...

The lock is managed by proxmox code directly. (through pmxcfs/corosync). You can't delete 2 lvm volumes at the same time.

you can also use a fleecing device in schedule backups advanced option, to define a temp buffer disk. (could be local storage)

I'm not sure, as it's using an ioctl ioctl(fd, BLKZEROOUT) , to tell to the storage to write 0 from a begin sector to a end sector . (Some kind of offloading, a little bit like discard). So, I don't think that O_DIRECT , CACHE_NONE,... apply here, as we don't really write zeroes through the...

Maybe because if the space is allocate again to a new vm, if the vm is doing a discard, it should remove the zeroed block yes, it's zeroing by security, to avoid to have old datas when you'll create a new vm on the old allocated space of previous vm.. but of course this zerroed space is...

for secure delete, the new blkdiscard here, is not using discard, but zeroing feature (blkdiscard -z). It's a little bit different, because it's really writing zeroes (telling to the storage to writing zeroes by range from this begin sector - end sectore), that t's it's faster than cstream...

I wonder yes, it should work. The only "problem" is that lvm is reservering blocks address space. The creation of snapshot itself is not writing zero. I wonder if it could be possible to declare a lun with with virtual size bigger than your san real storage space ? (I think it's depend of...

I'm still working on it ;)

Currently they are no thinprovisioning, so It'll not help for lvm+qcow2 to retreive space. (as the lvm under qcow2 is thick) But it could reduce the size of you backup as it's cleaning the disk blocks on file deletion.

@bbgeek17 zeroing optimisation has been merged in pve-storage >= 9.0.14 !

ah you can do it also on the pool, great :)

you don't need to do it manually in the guest (at least on linux), the virtio drivers are able to change value automatically.

@fstrankowski I'm looking to add the option in the proxmox gui, to be sure, how do you set the value ? "ceph config set client.admin rbd_read_from_replica_policy localize" ?

It's supported on pve9. (with lvm block volume formated with qcow2, and qcow2 snapshot chain of lvm volume) . Qcow2 is not only for file. (same for vhd on xcp-ng, their lvm volume are formated with vhd)

about snapshot && space usage, only qcow2 with external snapshot (so for shared block device), is currently still experimental and need indeed same space for each snapshots. but all others plugins can do snapshots && don't need same space for each snapshot. (including shared block like ceph...

Block jobs Non-active block-commit was optimized to keep sparseness blockdev-mirror was optimized to do less work with zero blocks blockdev-mirror and blockdev-backup gained new options, see QMP section interesting :)...

mmm , this is strange. for the intel, I have some of them in production,and I don't have theses result. does it solve the problem if you force the values manually ? (you can use "ceph config set osd.x osd_mclock_max_capacity_iops_ssd 30000" )

https://lore.proxmox.com/pve-devel/mailman.50.1727091601.332.pve-devel@lists.proxmox.com/t/ " * During backup, there is often a longer running connection open to our QMP socket of running VMs (/var/run/qemu-server/XXXX.qmp, where XXXX is the vmid). This blocks our management stack from...