Failing XFS memory allocation

[Nikolay Borisov]

On 03/24/2016 11:33 AM, Christoph Hellwig wrote:
> Hi Nikolay,
> 
> can you give the patch below a spin?  While it doesn't solve the root
> cause it makes many typical uses of kmem_realloc behave less badly,
> so it should help with at least some of the less dramatic cases of very
> fragmented files:
> 

Sure, however I just checked some other servers with analogical setup
and there are files which even larger extents (the largest I saw was 2
millions) so I guess in this particular case the memory was fragmented
and compaction as invoked from the page allocator couldn't satisfy it.
So I don't know if it will help in my particular case but in any case I
will give it a go.

Thanks