friends ask me to help. They have a linux 10.3 running as file server (with samba). Now since some weeks the client computers show messages that indicate that the drive mappings have been lost. This could be fixed with a new login. At first this was once a week, than it became more and more, now it is about every 60 minutes.
In the samba log file I found today every hour an entry saying samba daemon has been started. So I wonder why samba restarts every time. The server is several 100 km far from me so I can´t sit down at the server and check what´s going on. But I can call the server via webmin. Within webmin I found in the list of current running processes that all of the processes are running only since several minutes (the server is intended to run 24/7 and there is no one that shuts it down and restarts). I compared the time with the time of last start of smbd and it was the same. So I think that there is a server reboot every 60 minutes since about 1 week or more.
I looked for other log files but until now I did not find any log file that has any information about why the server shuts down, there is never any information about going down anyway. If I compare with my own suse 10.3 box I find for example in /var/log/messages some lines every time I shut down the PC. But not so in this server.
So I assume that there is something that causes the server to restart without a correct shutdown before. Where do you think I have to search for? Hardware? Maybe power supply? Mainboard? Memory? Or can I hope to find some hints somewhere in some log files? Where?
Thank you for every help.
Klemens
If you see NOTHING about a shutdown each time I’d definitely suspect a
power problem of some sort whether it be power supply, power strip, or
idiot unplugging it. Either way, not something you’re likely to solve
easily without watching things in person.
Good luck.
Klemens Lichter wrote:
| Hi,
|
| friends ask me to help. They have a linux 10.3 running as file server
(with samba). Now since some weeks the client computers show messages
that indicate that the drive mappings have been lost. This could be
fixed with a new login. At first this was once a week, than it became
more and more, now it is about every 60 minutes.
| In the samba log file I found today every hour an entry saying samba
daemon has been started. So I wonder why samba restarts every time. The
server is several 100 km far from me so I can´t sit down at the server
and check what´s going on. But I can call the server via webmin. Within
webmin I found in the list of current running processes that all of the
processes are running only since several minutes (the server is intended
to run 24/7 and there is no one that shuts it down and restarts). I
compared the time with the time of last start of smbd and it was the
same. So I think that there is a server reboot every 60 minutes since
about 1 week or more.
| I looked for other log files but until now I did not find any log file
that has any information about why the server shuts down, there is never
any information about going down anyway. If I compare with my own suse
10.3 box I find for example in /var/log/messages some lines every time I
shut down the PC. But not so in this server.
| So I assume that there is something that causes the server to restart
without a correct shutdown before. Where do you think I have to search
for? Hardware? Maybe power supply? Mainboard? Memory? Or can I hope to
find some hints somewhere in some log files? Where?
| Thank you for every help.
| Klemens
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
I agree, it sounds like hardware. Unfortunately, you will not be able to see that in any logs. Being remote, you are very limited on how much you can diagnose the hardware. You can remotely check cpu and mobo temps, and run smart on the disks. You can also stress test the cpu and ram with Prime, and test the ram alone with memtest. There are diagnostic checks that can be made on the hardware, but unfortunately you have to be physically at the machine.
Also, if the server is running solo on a UPS that might be the problem (UPS resetting due to a defect or attempting calibration).
Letting the server run directly on the mains will tell you enough.
Wishing you luck,
Wj
p.s. What brand is the server… does it have some kind of internal BIOS logging? Maybe that will show the reason…