>>Has anybody tested whether creating the filesystem with various block >>sizes circumvent this bug?
Hi, I'm not sure the problem come from the filesystem block size. I think if more about sectors alignement betwwen physical drive (4K) and virtual drive (qemu default is 512bytes) some interresting patch http://comments.gmane.org/gmane.comp.emulators.qemu/126267 " block: add support for 4k logical blocks" " 1. This patch aims to add 4k logical support for qemu running on a host > with 4k logical block size, right? No, it adds support for 512b logical block sizes running on a host with 4k logical block size and cache=none. This is suboptimal as it requires bounce buffers, but it can happen with migration and until libvirt provides a knob for the guest's logical block size. > For guest, we can use > logical_block_size=4096 to achieve that even on a host with the > logical_block_size of 512. Am I right? " So this patch seem to get default qemu drive of 512b working on 4K drive. (but suboptimal). To be perfect, we need to pass logical_block_size=4096 to qemu command line qemu -device virtio-blk-pci,logical_block_size=4096,physical_block_size=4096,.... But user on forum seem to say that it wasn't working :( ----- Mail original ----- De: "Michael Rasmussen" <[email protected]> À: [email protected] Envoyé: Jeudi 8 Novembre 2012 18:40:17 Objet: Re: [pve-devel] new cache benchmark results On Thu, 08 Nov 2012 18:06:47 +0100 (CET) Alexandre DERUMIER <[email protected]> wrote: > Yes, I remember the post in the forum. It doesn't work with cache=none or > cache=writeback > Has anybody tested whether creating the filesystem with various block sizes circumvent this bug? -- Hilsen/Regards Michael Rasmussen Get my public GnuPG keys: michael <at> rasmussen <dot> cc http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xD3C9A00E mir <at> datanom <dot> net http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE501F51C mir <at> miras <dot> org http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xE3E80917 -------------------------------------------------------------- _______________________________________________ pve-devel mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel _______________________________________________ pve-devel mailing list [email protected] http://pve.proxmox.com/cgi-bin/mailman/listinfo/pve-devel
