Take the 2-minute tour ×
Server Fault is a question and answer site for professional system and network administrators. It's 100% free, no registration required.

I recently deployed ProCurve 1800-24G switches in place of some older ProCurve 2424M switches in my network. However, I'm having a serious problem with the switch connected to the router. It seems, every night when our Windows 2008 R2 server (off site) runs a backup to a iSCSI target (on site) [facilitated through a PPTP tunnel] the LAN loses connectivity with the router. To clarify, there is only one router which is connected to the switch affected by this problem. The only way to resolve the issue is to either reboot the router or pull the ethernet cable that goes to the router and plug it back in.

During the outage, clients cannot receive DHCP requests, DNS requests, ping, or do anything else with the router in this state. Now, neither the switch or router are configured extensively and the issue only seems to have surfaced with the new switch in place. I have tried a number of things including replacing cables, rebooting and checking the switch configuration (it is literally as basic as you can get at this point-- flat LAN, no trunking). Interestingly, the router shows (accessed externally) no changes in configuration or status during this state but similarly cannot ping or access other hosts on the network.

This issue occurs in different stages of backup (ie, different amounts transferred). I've also dumped packets from the switch into WireShark but cannot seem to find any anomaly yet (I'm looking at packets around the time the issue appeared and at the time when I reset the NIC). Any suggestions for what to look for? Ideas on what could be causing this? I'm seeing some transmit/receive errors on the NIC from both the router and switch side but nothing serious when compared to the total packet counts. I'm seriously doubting hardware at this point, as I have tried another switch, different cables, and a different NIC on the router.

share|improve this question
    
Are you using any dynamic routing associated with the PPTP tunnel? If the tunnel becomes a route to its own endpoint, this kind of thing will happen. Is the tunnel always up or does it come up on demand? –  David Schwartz Oct 21 '11 at 22:04
    
It is always up. I'm not using any dynamic routing that I'm aware of. Any idea why this would take place with only the new switch in place and not the old? –  user98651 Oct 21 '11 at 23:12
    
Is the PPTP tunnel bridged to your LAN or routed to it? (No, no idea yet.) –  David Schwartz Oct 21 '11 at 23:20
    
It is currently bridged to the LAN. I'm going to try and separate the networks and test. –  user98651 Oct 21 '11 at 23:25
add comment

1 Answer

We have had some HP issues with access ports sort of 'freezing up' but still reporting 'up' in monitoring software. Most of the time a needed switch firmware update solved the issue.

share|improve this answer
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.