Linux - NewbieThis Linux forum is for members that are new to Linux.
Just starting out and have a question?
If it is not in the man pages or the how-to's this is the place!
Welcome to LinuxQuestions.org, a friendly and active Linux Community.
You are currently viewing LQ as a guest. By joining our community you will have the ability to post topics, receive our newsletter, use the advanced search, subscribe to threads and access many other special features. Registration is quick, simple and absolutely free. Join our community today!
Note that registered members see fewer ads, and ContentLink is completely disabled once you log in.
Installed mdk 9 on a Pentium 200 mhz machine. Network card detected during install and woked fine on first boot. It's goint to be used as a company intranet server so it will need to run 24 x 7 x 365. After the install, the network connection died the following day. As soon as I rebooted it started working again, then died about a day and a half later again. Rebooting always gets the network going but it always dies after 20 to 30 hours. My network card is a 3com 3c905B-TX. There are no power management options enabled in the bios. I replaced the card with another one - same kind, still have same problem. Is there a problem with these 3com cards or should I look elsewhere ?
Any pci card recommendations ?
Distribution: Red Hat (8.0, RHEL5,6), CentOS, SuSE (10.x, 11.x, 12.2, 13.2), Solaris (8-10), Tru64, MacOS, Raspian
Re: Network connection dies every 20 to 30 hours
Originally posted by explorer ... It's goint to be used as a company intranet server so it will need to run 24 x 7 x 365. After the install, the network connection died the following day. As soon as I rebooted it started working again, then died about a day and a half later again. Rebooting always gets the network going but it always dies after 20 to 30 hours. ...
If this is in an office environment, see if there's any correlation to when the cleaning crew comes in. Perhaps they're unplugging something so they can vacuum. Perhaps a hub that your machine is plugged into and the power cycle causes the card to latch up. Stranger things have happened. (Years ago, we used to leave the PCs in our offices on all night. Some of us were running long simulations, etc., and many times we'd come in next morning to find that the systems had rebooted. Turns out that the cleaning crew had been told to turn off half the hallway lights, and since they were unable to find the right breaker in the panel, they'd randomly flip them until they got the right one, whacking the power in many of the offices in the process. Until we figured out what was going on and labeled the breakers for them. :-) )
I'm not sure which log files to check or exactly what to look for, but the ones I found showed nothing unusual. It is an office environment but it's not the cleaning crews. The machine is currently running in my office and it will be working fine when I come in in the morning then later in the day with no one there but me it will die, or it may run all day and die the next day at noon. I have seen the cleaning crews do some nasty power offs on computer equipment in the past though - good thing to consider. I't inhouse so there's no external isp involved. I was originally using dhcp to get it's address and suspected something was up with dhcp address lease renewals so I changed that and gave it a static IP and still have the problem.
I noticed one thing today though. I took the cover off to see if it had any ISA slots on the board or not. I was considering trying an ISA network card. While I had the cover I noticed the PCI Video card was barely seated in the slot. I shut it down and pulled the video card out so I could see if the network card was also not fully seated, and what I saw was the the network card was farther out of the PCI slot than the video card was - some of the pins must have been barely making contact. Since the mainboard fits in the case kind of funky, I had bend the case a little to make the video card seat right, and I had to move the network card to another PCI slot. This might possibly have been the problem all along but I won't know for sure until I try it this coming week. I'll keep you posted.....
Nope that wasn't the problem. The connection to it died about 7 oclock this morning. This time, powering down the hub for a minute and powering it back up made the conneciton live again without having to reboot. I have 2 other computers on the same hub and the network was still good on both of them ( w2k boxes ) only the mdk connections dropped.
I found this in my /var/log/messages file. I don't understand everything in the file, but it looks like the server became a master browser for the workgroup at about 10:15, and about 9 minutes later started getting eth0 timeout message which continued every 10 seconds until I reset the hub this morning....I wonder if this is just coincidence or if something happenned when it became a master browser ....