PowerEdge T110 II Network Issues

Servers

Servers
Information and ideas on Dell PowerEdge rack, tower and blade server solutions.

PowerEdge T110 II Network Issues

This question has suggested answer(s)

We have a brand new PowerEdge T110 II server running Server 2008 R2 and Exchange 2010 as well as BackupExec 2010 R3.  We have an issue were this machine will randomly disconnect from the network (no disconnect message is recorded in the event log and the switch lights don't go out).  During this, all the network drives connected using DFS will failover to the 2nd server, but the Exchange system will just go offline until the issue goes away (usually about 30 seconds to 1 minute).  We believe it is the network drivers, but we don't have a spare Intel NIC (We use on the onboard Broadcom one) to test with.

We know its not the switches because the internet works fine and the 2nd server is running Server 2003 R2 and it has ZERO issues.

Does anyone else have the same system with Server 08 R2 with this issue?

All Replies
  • You could try Broadcom drivers, latest ones here: www.broadcom.com/.../netxtreme_server.php

  • We did download them from that site, but when we try to install them (there is no exe file) Windows says it is already using the most up to date driver and stays on the windows default driver.

  • You have to install them manually, device manage > network adaptors. select device, right click and select update drivers. Select Browse my computer for driver software, Select 'Let me pick from a list, then click 'Have disk', Browse to the folder where the drivers are, select the inf and click next. The drivers should install/update

  • I'm just brain storming here, but perhaps unchecking the "Allow the computer to turn off this device to save power" in the power management tab of the nic properties will make a difference? I'm also curious as to what the troubleshooting app says if it is run on the nic while in this state, perhaps it will yield clues.

    Oh, and if you have a BMC or idrac express shared on the nic, try disabling them so they arent sharing the port with the nic.

  • We will try this but as of right now after the last reboot the issue hasn't reappeared.  We turned off most of the features (like Checksum Offload) and rebooted and it has been ok.  We unchecked allow this device to save power.

  • I think we might have the same problem on our Dell PowerEdge T110. It is a W2K8 server (x86) domain controller running DHCP and DNS servers amongst other things. The box has a single Broadcom Netxtreme Gigabit Ethernet adaptor connected into the company network via a gigabit switch and routers (all manufactured by NetGear). Intermittently, we would get a DNS problem on our clients. No event logs indicating the cause of the problem. Using Wireshark and some of my own tools, I tracked this down to an intermittent failure to reply to a DNS request on the server. This would cause a whole bunch of knock-on effects on NBNS which had me well confised for a while. Anyway, during these periods of outage, we found we could not connect to the server using RDP either suggesting that it was not so much a DNS problem but one to do with general networking. We turned off the power option to stop the system putting the network card to sleep, but this did not help. We temporarily disabled the firewall, but this did not help either.

    Yesterday, I was searching the web for something else and came across some people with similar problems using the same NIC. They seemed to have resolved the issue by updated the network card driver. So I downloaded the latest driver from Broadcom (14.8.0.5) and updated it. I have not had any problems since, but it's very early days.

  • Problem returned so still unresolved. Any ideas?

  • have you tried the latest nic firmware: www.dell.com/.../DriverFileFormats

  • OK, I downloaded the package NETW_FRMW_WIN_R319248.EXE (version 6.5.0.131) signed by Dell in November and labelled:

    Dell Update Package: Broadcom BCM95708C PCI-E NIC, BCM5708 Copper LOM, BCM5708 SerDes (Fiber) LOM, BCM5709C Dual Port NIC without iSCSI Offload, BCM5722 Based LOM, BCM95722 PCI-E NIC, BCM57710 10GBase-T Dual Port Rack Mezzanine Card, BCM57710 10GBase-T Single

    I installed the package with no problems and rebooted as requested. Let's see if that's any better. Thx.

  • Unfortunately, it went wrong again about 10 minutes ago, so updating the driver and the firmware do not appear to have helped. It's particularly frustrating because its hard to get logs that indicate what is actually going wrong. If anyone has any other ideas, I would be grateful.

  • Try changing your NIC settings to match this and see if it works.  We changed them to this and haven't had any issues since. 

  • I worked through each setting in turn on Friday. I don't have an ARP Offload property. However, I changed the rest as you have specified. Unfortunately, the error cropped up again this morning, initially on a DNS request and then on an RDP connection. I have reverted all the settings back to where they were. Maybe, the problem I am seeing is different to the one you have.

    One thing I did notice is that IPv6 is currently disabled on the adaptor as we don't use v6 in our organisation. Also, there is anti-virus on the box which inserts a network security filter driver into the stack. Maybe, I should try disabling that.

  • Saxonite, I am having the same problem with a brand new Dell PowerEdge T110 II, SBS2011. Here is my situation: I've setup the server and three new systems in my shop, with no problem. I have a Netgear Gigabit Router. Took the server and the systems to replace an old Dell PowerEdge 400 SC SBS 2003 and three old desktops for a client (with another 7 newer desktops in the office). Switched out the servers, and discovered that I was losing Internet connection. Started to look into it, and discovered that I was getting Intermitent response from the gateway (Cisco E2000 non-gigabit router) when pinging continously. I thought the router went bad, as a bad coincidence. Plugged the server straight into the router, bypassing all network switches, and nothing else but the router and server connected, same problem. Because I was under a deadline to connect with a software vendor, for some additional configuration, I had to run to the store, bought a new Gigabit DLink 655 router and solved (temporarily) the problem. You can read more details in the link below. However, the story doesn't end here:

    http://www.experts-exchange.com/Software/System_Utilities/Remote_Access/VPN/Q_27596062.html

    Once I completed the setup, everybody was happy, I went back because I wanted to switch the new router with another new router, with DUAL Band features, to provide better coverage for the office. The new router, a DLINK 825 Gigabit. I was stunned to discover that I have the same problem, exactly the same problem, intermittent pinging response to the gateway, from the T110 II server. Again, bypassed all switches, plugged server directly into the router, same problem. I asked a friend of mine who's a networking expert to come by with his protocol analyzer, to look into the issue, just last night (March 3). He plugged the server into a 10mpbs Hub, plugged the router too, plugged his laptop with the software in the hub too, and watched the packets going from the server to the router and back. He was able to see how the packets are going out the server, to hub, into the router, back to the hub, out the hub to server. However, the server would drop some of them, while others would be processed just fine. So, it was clear that the problem was somewhere at the server level. Right after he left, I called Dell Server Support.

    The Support guy didn't have any good ideas that could fix the problem, and was constantly talking to his peers, with no good feedback whatsoever either. As I was talking to him, I ran multiple diagnostics on the NIC, using the Broadcom software. After getting 10-11 successful loops on all parameters, I went back to the configuration tab, and looked at things, for the millionth time. I saw Priority & VLAN Enabled. Since I don't have any VLAN, I decided to disable the VLAN part and leave just Priority Enabled. The pinging became continous, and the problem disappeared. I jumped up and down, but that was too soon. Ended the Dell call, and decided to install a Windows update that popped up.

    Noticed upon reboot a BMC Communication Failure to for the BMC utilitiy. The area in the boot process where we see the IP Configuration, etc, was no longer available. Went back to Windows, and was able to ping continously the gateway. However, I was concerned with that BMC communication failure looked it up on Google and saw that I can clear it by shutting the server down, unplugging the power, everything else, pushing the power button as I would turn on the server, and waiting for 30 seconds. Did that, and cleared the error.
     
    However, when I came back to Windows, the intermitent pinging came back. So, after all it was not the Priority & VLAN Enabled setting on the Broadcom driver. I put that back, and then remembered that prior to changing this setting, I ran a diagnostic on the network connector using the Broadcom software for about 10-11 cycles, while I was talking to the guy at Dell, who was clueless in this area. So, I tried again the diagnostic for about 8-9 times, rebooted, and then again had the BMC communication failure in the boot process. Went back to Windows, was able to ping with no problems.
     
    Cleared again the BMC communication failure, back to Windows, the intermitent pinging re-appeared.
     
    All this with the server pinging the router through the entire network setup, not directly with the server plugged into the router.
     
    So now, I was able to duplicate something.

     
    Called Dell Support, talked to a different guy. He asked me about BIOS, for the Broadcom, for the Server. He wanted to flash them, to make sure that there are no issues, even though they were current. Did that, and no success.
     
    He told me that there are two possibilities:
     
    a. The system board may be bad and may have a problem
    b. Windows could be the culprit.
     
    He asked me to create a boot disk for another operating system, and sent me to the following link (http://linux.dell.com/files/openmanage-contributions/omsa-65-live/) to do one. I will work on this over the weekend, and at some point, will come back and see if I can ping continously the gateway from it, if possible.
     
    If pinging continously from another OS would work, we know it is Windows that is having a problem. If it's not working from another OS either, we know it's the system board, and more specific the network chip.
     
    For the hardware, I would call Dell back and they would set up a system board replacement, which would hopefully resolve the problem. If it is Windows, we would need to remove the TCPIP stack, and re-install it, as maybe something it is corrupted in there.
     
    At this point I placed back in use the DLink 655 router which works fine for now.
     
    I will try to work on the tasks above sometimes next, to see what can we eliminate for this server. Will keep you posted.

  • Hey Saxonite, have you figured out what was the causing the problem? I am having the same problem and I've tried all the suggestions here, but no luck. I am thinking I may try installing a different NIC.

  • No. I still have the problem. I was waiting for a few days to see if btm2003sf came up with anything.