Re: "md_raid5: page allocation failure" when resyncing on 3.2

[Konstantinos Skarlatos <k.skarlatos@gmail.com>]

On Δευτέρα, 9 Ιανουάριος 2012 8:31:07 μμ, Roman Mamedov wrote:
> On Mon, 09 Jan 2012 19:25:23 +0200
> Konstantinos Skarlatos<k.skarlatos@gmail.com>  wrote:
>
>> Hello, i got this kernel message when i was resyncing my md raid 5 array
>> on a linux 3.2 machine.
>
> This probably doesn't have much to do with md, it is a generic Linux kernel failure in the VM subsystem.
>
> If the search on bugzilla.kernel.org was up, I'd point you to bug numbers (there was quite a lot of reports), but it isn't, so just see:
> http://www.google.com/search?ie=UTF-8&q=page+allocation+failure
> https://bugzilla.redhat.com/buglist.cgi?quicksearch=page+allocation+failure
> e.g.: "they are happening during heavy network and disk activity when the system has plenty of memory free"
> https://bugzilla.redhat.com/show_bug.cgi?id=551937
> and: "This is very common.  e1000 attempts to do large memory allocations from within interrupt context and the page allocator cannot satisfy the allocation and is not allowed to do the necessary work to make the allocation attempt succeed.  It's the same with all net drivers"
> http://kerneltrap.org/mailarchive/linux-netdev/2009/4/13/5473414
>
Thanks for the links, i will try some of the solutions offered there.
> In your backtrace I see the system was also handling a network transfer with the r8169 driver; do you use jumbo frames there (MTU over 1500)?

Yes, i use jumbo frames. Will disabling them make things better or 
worse?
Kind regards,
--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html