absolutvk
asked on
Help find cause of random server reboots (linux)
My Debian (Lenny) server is restarting multiple times a day at random times. I've looked though syslog and other any logs but can't find anything. I have a feeling its the web spider script I'm using. It starts 20 processes at once making lots of connections a second. A lot of connections also fail leaving them open. Bind is also put under pressure resolving lots of different domains, some which don't exist. If this or something else is causing a problem how can I diagnose the fault and possibly fix it?
There's lots of ram is available, load is low (<0.5) and all software is up-to-date.
There's lots of ram is available, load is low (<0.5) and all software is up-to-date.
ASKER CERTIFIED SOLUTION
membership
This solution is only available to members.
To access this solution, you must be a member of Experts Exchange.
There are three cause for this issue:
1. Power Supply Problem
2. Hard Disk has Bad Sector
3. RAM Faulty
I have come across this issue many time, every time I changed Power Supply of Server.
1. Power Supply Problem
2. Hard Disk has Bad Sector
3. RAM Faulty
I have come across this issue many time, every time I changed Power Supply of Server.
ASKER
I'm running a memtest inside Linux after reading your first comment. There is a limit to how much I can test the hardware as I can only access it remotely via ssh, it's a leased server. It would be possible to get the hardware tested properly but they would charge alot and it may be for nothing.
What I would like to do is stress test the hardware. If it reboots then I know its the hardware. Do you recommend something for this task? It has to be ran from terminal.
What I would like to do is stress test the hardware. If it reboots then I know its the hardware. Do you recommend something for this task? It has to be ran from terminal.
you could give the user space utility a try:
http://pyropus.ca/software/memtester/
and you could use some sort of shell script (to generate dma traffic and the like too):
http://people.redhat.com/dledford/memtest.html
http://pyropus.ca/software/memtester/
and you could use some sort of shell script (to generate dma traffic and the like too):
http://people.redhat.com/dledford/memtest.html
Verify temperatures on mobo/cpu/ram, verify power supply stability(replace to check).
Is it server motherboard? I encountered such mobos which worked with one kind of memory(recommended by vendor), occasionally hang/reboot with others.