xfs
[Top] [All Lists]

Two problems with a new installation

To: linux-xfs@xxxxxxxxxxx
Subject: Two problems with a new installation
From: "Dave Sill" <ds-linux-xfs@xxxxxxxxxxxxxxxxx>
Date: 17 May 2001 16:04:26 -0400
Mail-followup-to: linux-xfs@xxxxxxxxxxx
Sender: owner-linux-xfs@xxxxxxxxxxx
User-agent: Gnus/5.0807 (Gnus v5.8.7) XEmacs/21.1 (Cuyahoga Valley)
I installed XFS 1.0 + Red Hat 7.1 via the SGI installer on a 2-proc
Dell server with 2 x 250GB AMI MegaRAID's. The installation went
smoothly, but I've been getting occasional messages like:

May 17 03:39:08 daacsnfs kernel: __alloc_pages: 2-order allocation failed.
May 17 03:39:08 daacsnfs kernel: __alloc_pages: 2-order allocation failed.
May 17 03:42:52 daacsnfs kernel: __alloc_pages: 0-order allocation failed.

What do these messages mean?

I'm migrating one of the RAID's to xfs and it's going rather slowly
(6 hrs to untar 18 GB so far). Watching the disk space used on the xfs
raid, I see lots of periods with no obvious activity. E.g., at
5-second intervals:

                     248533712  13932188 234601524   6% /smm1
                     248533712  13934712 234599000   6% /smm1
                     248533712  13936248 234597464   6% /smm1
                     248533712  13936248 234597464   6% /smm1
                     248533712  13936248 234597464   6% /smm1
                     248533712  13936780 234596932   6% /smm1
                     248533712  13951096 234582616   6% /smm1
                     248533712  13951096 234582616   6% /smm1
                     248533712  13962360 234571352   6% /smm1
                     248533712  13974288 234559424   6% /smm1
                     248533712  13974328 234559384   6% /smm1
                     248533712  13974328 234559384   6% /smm1
                     248533712  13974328 234559384   6% /smm1
                     248533712  13974328 234559384   6% /smm1
                     248533712  13974328 234559384   6% /smm1
                     248533712  13993080 234540632   6% /smm1

The __alloc_pages errors are much less frequent than the I/O pauses,
so they could be two separate problems.

hinv says:

Main memory size: 896 Mbytes
2 GenuineIntel Pentium III (Katmai) processors
1 vga+ graphics device
1 keyboard
PCI bus devices:
    ISA bridge: Intel Corporation 82371AB PIIX4 ISA (rev 2).
    IDE interface: Intel Corporation 82371AB PIIX4 IDE (rev 1).
    USB Controller: Intel Corporation 82371AB PIIX4 USB (rev 1).
    Bridge: Intel Corporation 82371AB PIIX4 ACPI (rev 2).
    VGA compatible controller: ATI Technologies Inc 3D Rage Pro (rev 92).
    PCI bridge: Intel Corporation 80960RP [i960 RP Microprocessor/Bridge] (rev 
3).
    PCI bridge: Intel Corporation 80960RP [i960 RP Microprocessor/Bridge] (#2) 
(rev 3).
    PCI bridge: Intel Corporation 80960RP [i960 RP Microprocessor/Bridge] (#3) 
(rev 3).
    I2O: Intel Corporation 80960RP [i960RP Microprocessor] (rev 3).
    I2O: Intel Corporation 80960RP [i960RP Microprocessor] (#2) (rev 3).
    I2O: Intel Corporation 80960RP [i960RP Microprocessor] (#3) (rev 3).
    Host bridge: Intel Corporation 450NX - 82451NX Memory & I/O Controller (rev 
3).
    SCSI storage controller: Adaptec AHA-2940U2/W / 7890 (rev 0).
    SCSI storage controller: Adaptec AHA-2940U2/W / 7890 (#2) (rev 0).
    SCSI storage controller: Adaptec AIC-7860 (rev 3).
    Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (rev 8).
    Ethernet controller: Intel Corporation 82557 [Ethernet Pro 100] (#2) (rev 
8).
    Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge 
(rev 4).
    Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge 
(#2) (rev 4).
    Host bridge: Intel Corporation 450NX - 82454NX/84460GX PCI Expander Bridge 
(#3) (rev 4).

-- 
Dave Sill <MaxFreedom@xxxxxxxxxxxxxxxxx>       <http://web.infoave.net/~dsill>
Oak Ridge National Lab, Workstation Support
<http://www.lifewithqmail.org>: almost everything you always wanted to know.

<Prev in Thread] Current Thread [Next in Thread>