Il giorno mar 17 mar 2020 alle ore 01:30 David Sterba
<dsterba@xxxxxxx> ha scritto:
> without tuning of the rate I can't say if the buffered vs dio happens
> all the time or just once. For the 'once' case I would not be worried.
Thanks a lot David for your interest.
Well, with latest kernel (5.5.13), I still see it, but a lot less than
before (same setup, hardware and
software versions).
After a few hours of work I just got this bouch of warnings (btw, no
boot of virtual machine guest, it was in status saved/suspended).
[gio mar 26 05:22:58 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 05:22:58 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 25176 Comm: kworker/4:3
[gio mar 26 05:28:41 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 05:28:41 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 35075 Comm: kworker/4:1
[gio mar 26 05:44:00 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 05:44:00 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 38801 Comm: kworker/4:4
[gio mar 26 05:54:48 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 05:54:48 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 44366 Comm: kworker/4:1
[gio mar 26 06:02:43 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 06:02:43 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 44366 Comm: kworker/4:1
[gio mar 26 06:02:43 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 06:02:43 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 53354 Comm: kworker/4:2
[gio mar 26 06:02:43 2020] Page cache invalidation failure on direct
I/O. Possible data corruption due to collision with buffered I/O!
[gio mar 26 06:02:43 2020] File:
/mnt/4TB/piastrelli/home/virtual/VirtualBox
VMs/Zuccotti/Snapshots/{23d5aff0-4514-46b9-ab38-76ae59b3acbf}.vdi P
ID: 55712 Comm: kworker/4:6
I have no problem at the moment (this is happening on the test setup
to evaluate BTRFS), but maybe it's important to spot it for you all
devs.
If it help you, I can test the same setup with ext4 or another fs.
Thanks again,
Gelma
Ciao,
Gelma
