Re: I have a blaze of 353 page allocation failures, all alike

From: Peter Kruse
Date: Wed Apr 13 2011 - 03:14:11 EST


thanks for your replies, I appreciate that.

On 04/13/2011 03:34 AM, David Rientjes wrote:
On Tue, 12 Apr 2011, Christoph Lameter wrote:

The problem maybe that you have lots and lots of SCSI devices which
consume ZONE_DMA memory for their control structures. I guess that is
oversubscribing the 16M zone.

but there are only two devices:

# fdisk -l

Disk /dev/sda: 249.8 GB, 249865175040 bytes
255 heads, 63 sectors/track, 30377 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x4ac39024

Device Boot Start End Blocks Id System
/dev/sda1 1 13 104422 83 Linux
/dev/sda2 14 30377 243898830 8e Linux LVM

WARNING: GPT (GUID Partition Table) detected on '/dev/sdb'! The util fdisk doesn't support GPT. Use GNU Parted.

Disk /dev/sdb: 37946.0 GB, 37946025574400 bytes
255 heads, 63 sectors/track, 4613341 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Disk identifier: 0x00000000

Disk /dev/sdb doesn't contain a valid partition table

You can try to get more memory reserves specifically for lowmem in
ZONE_DMA by changing /proc/sys/vm/lowmem_reserve_ratio. The values are
ratios, so lowering the numbers will yield larger amounts of memory
reserves in ZONE_DMA for GFP_DMA allocations. Try lowering the non-zero
entries to 1 to reserve the entire zone for lowmem, assuming your system
has enough RAM for everything else you're running.

this server has 48GB - that should be enough.

This will verify if ZONE_DMA is being depleted from the larger number of
SCSI devices. If you don't get any additional page allocation failures,
then check how much memory in ZONE_DMA is used at peak and that would be a
sane reserve ratio to use next time you restart the system.

Could that still be a problem if there are only two scsi devices?


To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at
Please read the FAQ at