[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Qemu-devel] fstrim and cache=none
From: |
Paolo Bonzini |
Subject: |
Re: [Qemu-devel] fstrim and cache=none |
Date: |
Tue, 04 Jun 2013 08:53:14 +0200 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130514 Thunderbird/17.0.6 |
Il 04/06/2013 03:02, Dusty Mabe ha scritto:
> On Mon, Jun 3, 2013 at 5:39 PM, Dusty Mabe <address@hidden> wrote:
>>
>> It is possible that I was mistaken. Now when I switch back to
>> cache=none I am still able to execute trim operations. I'll
>> investigate a little more into my exact steps that I took and see if I
>> can come up with a reproducer for the issue I observed.
>>
>
> Paolo,
>
> I was mistaken. I had added a disk to the guest and then created an
> ext4 FS on the disk. I then noticed that I couldn't perform fstrim
> operations. I then examined the xml and the only difference between
> this disk and the other disk in the system was the cache=none. So I
> removed cache=none and then shutdown/booted the guest. After the
> reboot all worked fine so I assumed it was the cache=none line. This
> was wrong.
>
> Actually it is the formatting of the filesystem. When i create an FS
> in the guest it disables discard. There are also some errors printed
> in /var/log/messages.
>
> Please see the output in file at the following location:
>
> http://dustymabe.com/content/qemu-devel-mail.txt
Please attach the QEMU command line too. I'll try to reproduce.
Paolo
- [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/03
- Re: [Qemu-devel] fstrim and cache=none, Paolo Bonzini, 2013/06/03
- Re: [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/03
- Re: [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/03
- Re: [Qemu-devel] fstrim and cache=none,
Paolo Bonzini <=
- Re: [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/04
- Re: [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/11
- Re: [Qemu-devel] fstrim and cache=none, Paolo Bonzini, 2013/06/12
- Re: [Qemu-devel] fstrim and cache=none, Dusty Mabe, 2013/06/26