system get freeze while copying large files over network

[engin dumlu]

in the middle of the copying process, system gets freeze, syslog has no line about the issue
only way to takeover is restarting..
for any disk related problems, i checked the raid status from bios, which says OK
hw infos:

server: IBM eServer x226-[8488E4Y]

os: Debian etch 64bit

root@srv:~# uname -a
Linux srv 2.6.18-6-amd64 #1 SMP Fri Dec 12 05:49:32 UTC 2008 x86_64 GNU/Linux

root@srv:~# dmesg |grep eth
eth0: Tigon3 [partno(BCM95721) rev 4101 PHY(5750)] (PCI Express) 10/100/1000BaseT Ethernet 00:0d:60:17:2d:45
eth0: RXcsums[1] LinkChgREG[0] MIirq[0] ASF[0] Split[0] WireSpeed[1] TSOcap[1]
eth0: dma_rwctrl[76180000] dma_mask[64-bit]
tg3: eth0: Link is up at 1000 Mbps, full duplex.
tg3: eth0: Flow control is on for TX and on for RX.
eth0: no IPv6 routers present
ADDRCONF(NETDEV_UP): eth0: link is not ready
tg3: eth0: Link is up at 1000 Mbps, full duplex.
tg3: eth0: Flow control is on for TX and on for RX.
ADDRCONF(NETDEV_CHANGE): eth0: link becomes ready
eth0: no IPv6 routers present

thanks

[Don Vanco - Linux Ninja]

Is this NFS, SAMBA, FTP, HTTP, or SCP??

What are you copying to / from? (and does it freeze when you copy to / from, or both?)

dmesg is only going to show you boot-time data, not really useful....

We need to know the transport protocol to assist further.

If it's NFS - are you doing any tuning?

[engin dumlu]

hello Don,

it's a fresh debian install which was installed five days ago and nothing done yet about tuning

the data is 50 G mysql backup, which i tried to copy over scp to this frozen box

the purpose of this re-installing was the same: the system gets frozen in 1 or 2 month period

thanks

[Don Vanco - Linux Ninja]

Hmm - well, I would start with some basic troubleshooting methods.

When you way the box freezes, does this mean network and console? Can you ping the host? Are you capable of getting a serial console? It would be good to see if you can get into the system during the "event".

Do you get any kind of diagnostic lights? Not sure what the x226 had available... but I think it was only on the motherboard, and then not too informative.

Do you look over the logs on reboot? Again, DMESG isn't going to tell you anything, but you should be able to look in /var/log/messages and see when the host reboots, and what got logged just prior.

Regarding the x226 - are you on the latest firmware, especially the BIOS, NIC, and storage controller?
Taking a quick look here:
http://www-304.ibm.com/systems/support/supportsite.wss/selectproduct?taskind=2&brandind=5000008&familyind=5178870&typeind=5213128&modelind=5256765&osind=5366634&psid=sr&continue.x=1&matrix=Y
...shows some fairly critical updates to NIC and hard disk firmware.

After firmware, my first guess would be drivers, so I'd ask if you've updated the system to the current kernel? (sorry, I don't know Debian kernel revs). I have seen systems where the NIC becomes unresponsive, but never where it hands the whole box - that really sounds more IRQ or possibly RAM related.

Your kernel appears to have "amd64" in the name - yet the model you specify is an Intel-based server. Is that OK? Again, not familiar with Debian kernel nomenclature...

Some xSeries boxes use the tg3 driver and some the bnx2 driver - but since networking comes up initially that probably isn't an issue.

[Srimalik]

>Your kernel appears to have "amd64" in the name - yet the model you specify is an Intel-based server. Is that OK? Again, not familiar with Debian kernel nomenclature...


Should be OK, x86_64 run on amd64/intel64.