wewantrice Posted July 3, 2015 Share Posted July 3, 2015 Ever since upgrading to V6 and now with 6.01 my server has been unexpectedly dropping off the network. I have tried replacing the ethernet cable and switching to a different ethernet port as my Mobo has 2 onboard. No dice. Oddly if I disconnect the cable and plug it back in the server comes back online. I've attached my syslog. No problems with any other devices connected to the same switch. Any sage advice would be much appreciated. tower-syslog-20150703-0632.zip Quote Link to comment
RobJ Posted July 3, 2015 Share Posted July 3, 2015 Wow, that's one confusing networking setup! I don't blame the networking for being confused, as I sure was! * You have 2 networking chipsets, an Intel PRO/1000 onboard but not connected, and a Realtek that appears to be an addon card, connected. * You have the Realtek configured for jumbo frames, at 9200; the Intel does not indicate jumbo frames (may be set up for it, but it's not reporting it) * It looks like you have bonding set up, joining the two? Only one is connected, and it is set with a MTU of 9200 (other unknown) which may be too big? * You have bridging set up * You have it set to use DHCP, router is providing 192.168.0.102; router only gives one day leases, so it's renewing twice a day * Network line to Realtek stayed up most of the time, but dropped once 2 minutes before 192.168.0.145 connected, unknown reason * Bridge developed a problem, repeatedly reporting "br0: sending REQUEST (xid 0x736f6c0e)", once per minute for 7.5 hours, cannot tell if problem on your end or the other end but does look like it may have been the fault of the router; you disconnected and reconnected a couple of times and it started working again; I could not associate that xid with anything * Twice, the error "br0: IP address 192.168.0.102/24 already exists" was reported, no idea how serious, but cannot possibly be good! * Unrelated, but a USB subsystem dropped shortly after boot, taking down the attached Samsung SSD * Probably not related, but DMAR errors reported for PCI device 00:14.0, one of the USB controllers, probably the one for the SSD above All I can advise is to simplify simplify! I would remove all of the DHCP confusion by setting it to use a static IP, probably 192.168.0.102 but might want to pick something more unusual like 192.168.0.120 in case there's another .102. Turn off the bonding, since you aren't using it anyway, and it *may* be problematic matching up a Realtek with an Intel, especially if jumbo frames aren't matched everywhere (both NIC's, any switches, the router, and the other stations). Any reason why you aren't using the Intel? It's usually much more reliable than Realteks are. You might also be safer to forget about jumbo frames for now. Quote Link to comment
wewantrice Posted July 3, 2015 Author Share Posted July 3, 2015 Thanks for the insight, I'll try to implement some changes. My motherboard has 2 networking chipsets, when I started having network errors I thought maybe switching rm the Intel to the Realtek. Will report back if any of these changes solve the issue. I just find it odd that I never experienced this issue before upgrading from a v6 beta to 6.0 and 6.01 Thanks again! Quote Link to comment
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.