HP Proliant DL380 w Tao 4 (RHEL4 clone) crash under heavy disk usage

[Jerker Nyberg]

Hello, I've got a HP Proliant DL380 running Tao Linux 4 (RHEL4 clone). Its a new machine with a single CPU, 1 GB ram, 6*150 15 krpm internal SCSI-disk (RAID0+1), 128 MB battery backuped raidcache, an Intel gigabit ethernet NIC and an external 500 GB USB-disk (maxtor III onetouch),

When the machine crashed it was copying data från ethernet (the intel card) to the internal disks (smart array 6i) with rsync (8 parallel rsyncs each limited to 1 MByte/s throughput) and an rsync from the internal disks to the external USB-attached disk with no limit in bandwidth.

The kernel message was some ext3 not syncing,

Kernel 2.6.9-22.0.2.ELsmp. Filesystem ext3. LVM is not used.

The interesting thing is that i managed to crash a similar configured machine (with an extra Smart Array 6400 card with 256 MB raidcache) in december. This machine was serving files over NFS, no USB involved. It was a magificent crash with currupt files and disks that died, so I don't think Linux was the problem that time. But it makes me wonder.

So. Have anyone else had these kind of problems ? Any solutions?

Are there any mailinglists regarding Proliant and Linux? www.van-dijk.net seem to be dead (for the moment at least is no webserver responding).

[Jerker Nyberg]

The kernel panic message was something about "ext3 not syncing" but I do not know anything more. It wasnt me that rebooted it.