|
|
|
|
|
|
|
|
|
|
xen-devel
Re: [Xen-devel] Problems with dom0 and low memory
Keir Fraser wrote:
If under tight memory conditions you get errors from the blkdev layer
in DOM0 then the backend driver will simply send those erros back to
the frontend, which will pass the error up into xenU's blkdev layer.
What ext3 then does with that error is not really Xen's problem -- a
similar low-memory condition running directly on the hardware (i.e.,
without Xen) would likely have the same outcome.
I don't know if it's worth throwing my $0.02 in, since I have such a
vague error report, but I've had a similar thing happen.
I had 2.0.1 running on a machine at home doing benchmarks and load
testing and whatnot with four VMs and Dom0 (using 2.6.9 for both) all
simultaneously doing heavy disk I/o. The Dom0 was running against a
"normal" partition while the VMs were all on LVs. Dom0 reported
"attempt to access beyond end of device" with some really outrageous
sector numbers (well beyond the physical device) and all the VM's (using
XFS) generated XFS_shutdown errors and unmounted their filesystems.
I rebooted everything, xfs_check'ed everything, and it was all fine. I
never saw that error again, even under identical test circumstances.
(And I really realy tried to break it.)
I doubt it was related to memory since my Dom0 had plenty of RAM
allocated, and none of the VMs were doing anything especially
memory-intesive, but I guess there's a possibility of a really rare
error condition in the block handler somewhere?
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://productguide.itmanagersjournal.com/
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxxx
https://lists.sourceforge.net/lists/listinfo/xen-devel
|
|
|
|
|