When exercising btrfs via nfs4, system hangs and claims no disk space

[gpk]

I have an underlying btrfs filesystem which is mounted across 3 Ubuntu machines via NFSv4. I was running an application that caches some intermediate results on the disk; it produces about 5 files/sec (combined total across the three machines), each file is about 100k. The files are deleted randomly in groups of 30, and there is some directory creation/deletion mixed in. So the filesystem is moderately busy (but with some visible pauses).

After several hours of this, the application fails, claiming "no space". When I looked, shortly afterwards, though, there was still 25% free space. Also, the system was very slow, because things were wedged and timing out in the kernel. Dmesg had some interesting messages (attached as dmesg.txt).

A reboot fixes the immediate problems. This has now happened twice in a row, with a reboot in between.

ProblemType: Bug
DistroRelease: Ubuntu 11.04
Package: linux-image-2.6.38-9-generic 2.6.38-9.43
ProcVersionSignature: Ubuntu 2.6.38-9.43-generic 2.6.38.4
Uname: Linux 2.6.38-9-generic x86_64
AlsaVersion: Advanced Linux Sound Architecture Driver Version 1.0.23.
Architecture: amd64
ArecordDevices:
 **** List of CAPTURE Hardware Devices ****
 card 0: Intel [HDA Intel], device 0: VT1708B Analog [VT1708B Analog]
   Subdevices: 2/2
   Subdevice #0: subdevice #0
   Subdevice #1: subdevice #1
CRDA: Error: [Errno 2] No such file or directory
Card0.Amixer.info:
 Card hw:0 'Intel'/'HDA Intel at 0xfe878000 irq 44'
   Mixer name : 'VIA VT1708B 8-Ch'
   Components : 'HDA:1106e721,104382ea,00100100'
   Controls : 35
   Simple ctrls : 19
Date: Sun May 15 16:47:40 2011
InstallationMedia: Ubuntu 10.04.1 LTS "Lucid Lynx" - Release amd64 (20100816.1)
MachineType: System manufacturer System Product Name
ProcEnviron:
 LANGUAGE=en_GB:en
 PATH=(custom, user)
 LANG=en_GB.UTF-8
 SHELL=/bin/bash
ProcKernelCmdLine: BOOT_IMAGE=/vmlinuz-2.6.38-9-generic root=UUID=57bf6acd-3f72-4d1c-a530-2d3d36756168 ro quiet splash vt.handoff=7
RelatedPackageVersions:
 linux-restricted-modules-2.6.38-9-generic N/A
 linux-backports-modules-2.6.38-9-generic N/A
 linux-firmware 1.52
RfKill:

SourcePackage: linux
StagingDrivers: zram
UpgradeStatus: Upgraded to natty on 2011-05-12 (3 days ago)
dmi.bios.date: 06/08/2009
dmi.bios.vendor: American Megatrends Inc.
dmi.bios.version: 0516
dmi.board.asset.tag: To Be Filled By O.E.M.
dmi.board.name: P5KPL-AM
dmi.board.vendor: ASUSTeK Computer INC.
dmi.board.version: x.xx
dmi.chassis.asset.tag: Asset-1234567890
dmi.chassis.type: 3
dmi.chassis.vendor: Chassis Manufacture
dmi.chassis.version: Chassis Version
dmi.modalias: dmi:bvnAmericanMegatrendsInc.:bvr0516:bd06/08/2009:svnSystemmanufacturer:pnSystemProductName:pvrSystemVersion:rvnASUSTeKComputerINC.:rnP5KPL-AM:rvrx.xx:cvnChassisManufacture:ct3:cvrChassisVersion:
dmi.product.name: System Product Name
dmi.product.version: System Version
dmi.sys.vendor: System manufacturer

[gpk (gpk-kochanski)]

FYI, the relevant system calls (the ones that happen in the active loop) are stat(), statvfs(), open(), write(), close(), rmdir(), mkdir().

[Rolf Leggewie (r0lf)]

I wonder if this is a dupe of bug 874684? What does "df -h" say when you are running out of space?