Hello list. I admin a small network with a samba 1.9.18p10 server running under Linux 2.0.34. The server processes domain logins, WINS requests and is the LMB. Everthing had been running very smooth for months then all of a sudden, none of the client (win 9x) machines could login; the error was "No server found to processes your domain logon." Further investigation proved that I could indeed ping the Linux box and I could even do a 'net view \\server' and get a list of all shares. A thorough investigation turned up nothing out of the ordinary save for the fact that when I restarted samba, it did *not* force an election. The log files show that it sets up the interfaces, etc, but thats where nmbd stops. I finally broke down and rebooted the server and nmbd forced the election and domain logons worked as always. Now, the only thing I can think of is that this server had been up for approx. 144 days. Are the problems we experienced somehow related to that? Perhaps a problem with the 2.0.34 kernel? I am interested to hear other people's stories running samba under Linux 2.0.x with long uptimes or from anyone who can shed some light on my mini-mystery. Thanks in advance :-) -- Joel Knight jwknight@cyberlink.bc.ca PGP Key: hkp://keys.pgp.com/jwknight@cyberlink.bc.ca KeyID 2048/38C24864 Fingerprint 6D7D 1E4F 728B ACDA 6557 F3EC 85BB BA7C 38C2 4864
Hello, At 07.48 23/02/99 +1100, you wrote:>Now, the only thing I can think of is that this server had been up for >approx. 144 days. Are the problems we experienced somehow related to that? >Perhaps a problem with the 2.0.34 kernel? I am interested to hear other >people's stories running samba under Linux 2.0.x with long uptimes or from >anyone who can shed some light on my mini-mystery.never reched such a long uptime. Anyway I always had problems, even with much shorter uptimes, whenever I used servers without ECC memory modules. AFAIK 1 Mbit of DRAM (megabit *not* megabyte) looses about one bit a month on the average. In my experience uptimes longer then a week on non ECC system may be really dangerous. Hope it helps. Bye! Sbragion Denis InfoTecna Tel, Fax: +39 039 2324054 URL: http://space.tin.it/internet/dsbragio
I have Linux and Netware servers that have run well over 6 months at a time with no problems. Not only don't they have ECC, I don't even think the SIMMS have parity! I also have a customer with an SCO OpenServer system (with Samba) that ran continuously for about 8 months, and a few years back, I had an IX/386 system that ran on an old 386 supposedly for a year and a half without a reboot. When they finally DID shut it off to move it, the 5-1/4 ESDI hard drive seized and we had to wait till it cooled down before we could start it up again. After letting it cool down and cleaning out the dust, it came right up. They only used it for a few weeks after that so we didn't get a chance to try to break the record.> -----Original Message----- > From: Dick Balaska [SMTP:dick@buckosoft.com] > Sent: Tuesday, February 23, 1999 11:02 AM > To: Multiple recipients of list > Subject: Re: Large uptime > > Denis Sbragion wrote: > > > > Hello, > > > > At 07.48 23/02/99 +1100, you wrote: > > >Now, the only thing I can think of is that this server had been up for > > >approx. 144 days. Are the problems we experienced somehow related to > that? > > >Perhaps a problem with the 2.0.34 kernel? I am interested to hear other > > >people's stories running samba under Linux 2.0.x with long uptimes or > from > > >anyone who can shed some light on my mini-mystery. > > > > never reched such a long uptime. Anyway I always had problems, even with > > much shorter uptimes, whenever I used servers without ECC memory > modules. > > AFAIK 1 Mbit of DRAM (megabit *not* megabyte) looses about one bit a > month > > on the average. In my experience uptimes longer then a week on non ECC > > system may be really dangerous. > > http://www.buckosoft.com/linux/ > sarah.buckosoft.com Linux 2.0.13 Samba 1.9.15p8 current uptime > sarah up 390+19:16, 2 users, load 0.43, 0.28, 0.17 > > I had to patch ruptime when she went over 365 days. (I'd like to find > a way to get that back into distribution...I surf and see a .no dude > made the > same fix 2-3 years ago) > > I really lucked onto this rock solid machine. My other two Linux boxen > are nowhere near this stable. > > dik