I am having an issue at a site that currently communicates to our other sites via an IPSEC tunnel that terminates in our datacenter. The issue was first discovered when a user was attempting to check out a CAD license from a failover server at another location (all local licenses were taken). Connectivity to the server was fine but when the license was attempted to be checked out, the user loses internet connectivity and is unable to ping anything beyond the vlan he is in. The strange part is that while the user is experiencing this issue, we can remotely connect to the user and are able to ping him from the site where he attempted to check out a license. This has randomly happened to more than just one user but only at this site. It is now beginning to show up on other applications that rely on a license server to check out a license.
Initially, we opened tickets with the license server vendor and they insist it is something environmental and not with the license server. After doing some troubleshooting, I find that I am able to replicate the issue from the switch. If I ping from the core switch at this site to the core switch of the site with this license server, it pings fine and even the traceroute takes the correct patch. Strangely, when I ping from the core switch at the affected site to any internal host at the site with the license server, it pings fine, but after running traceroute, it takes the correct path but times out at the default gateway of the remote site, then I am unable to ping successfully again. It resolves itself after about 10 minutes and I can then ping again.
I have tried clearing the arp cache from the switches and firewalls. I have also rebooted all the switches, firewall, and router at the affected site, but still no luck. The core switch is an HP Procurve 2920. Just wondering if anybody has some suggestions on what to look for.
No comments:
Post a Comment