[Ocfs2-users] ocfs2 hangs during webserver usage

[Bret Baptist]

On Tuesday 27 January 2009 12:52:45 pm David Johle wrote:
> Even if the reads are scattered about the disk, I do find it hard to
> believe that the SAN can't keep up with 100 IOPs.
>
> I will look into the blktrace for sure ... expect a reply on that soon.
>
>
>
> And FWIW, today it did more than just slow down or lag for a bit, it
> actually hung for real.  Tons of processes in D state making load
> average go up to 280+ and one process spinning at 100% (but not
> actually doing anything according to strace).
>
> I did get some interesting output from syslog which might provide
> some more insight:
>
> Jan 27 11:12:08 wolf kernel: BUG: soft lockup - CPU#7 stuck for 10s!
> [vsftpd:9868]
> Jan 27 11:12:08 wolf kernel: CPU 7:
> Jan 27 11:12:08 wolf kernel: Modules linked in: mptctl mptbase
> ipmi_devintf ipmi_si ipmi_msghandler dell_rbu ocfs2(U) dm_round_robin
> ocfs2_dlmfs(U) ocfs2_dlm(U) ocfs2_nodemanager(U) configfs ipv6
> xfrm_nalgo crypto_api ib_iser rdma_cm ib_cm iw_cm ib_sa ib_mad ib_co
> re ib_addr iscsi_tcp libiscsi scsi_transport_iscsi dm_multipath video
> sbs backlight i2c_ec i2c_core button battery asus_acpi acpi_me
> mhotplug ac parport_pc lp parport joydev ata_piix libata sg
> i5000_edac ide_cd serio_raw bnx2 e1000e shpchp edac_mc cdrom pcspkr
> dm_s napshot dm_zero dm_mirror dm_mod megaraid_sas sd_mod scsi_mod
> ext3 jbd uhci_hcd ohci_hcd ehci_hcd

This is just a shot in the dark, but I have heard about issues with the e1000e 
module with certain kernel versions.

The forum thread that I saw it on is here:
http://www.proxmox.com/forum/showthread.php?t=906

Hope that helps.
-- 
Bret Baptist
Senior Network Administrator
bbaptist at iexposure.com
Internet Exposure, Inc.
http://www.iexposure.com
(612) 676-1946 x17

Providing Internet Services since 1995
Web Development ~ Search Engine Marketing ~ Web Analytics
Network Security ~ On Demand Tech Support ~ E-Mail Marketing