11-05-2013 02:30 PM
Hi all,
Service provider maintains our server services in two geographically distributed data centres. Our WAN infrastructure has a high concentration of DMVPN users and we are experiencing some random issues whereby users can take up to 20-mins to logon whilst others are very quick. Users utilise thin-client Wyse devices.
Not had much help from the service provider and we ourselves cannot narrow down a common issue. So we are about to get Wireshark to monitor but wondered if there are any other similar experiences out there as we suspect that the issue is with the provider but we have to prove it.
I have seen articles about MTU/MSS which indicates fragmentation might be issue although our routers and switches are set to MTU 1400.
So any experience out there, advice for things to look for given that ICA is new to us and looks to be not a pretty protocol?
Any help gratefully received.
iJT
Sent from Cisco Technical Support iPad App
11-06-2013 06:52 AM
The ICA protocol is not in play here if you are having long login times. Assuming you are using roaming profiles or some sort of profile management service like Citrix UPM, you should take a look at the traffic from you profile server to your XenApp servers (assuming terminal servers). Are your users logging into servers in one datacenter but profiles are in the other? This will cause the profile to be copied across the WAN. Also take a look at the domain controllers and profile servers themselves to make sure resources are free.
11-11-2013 10:29 AM
Thanks Scott
Apologies for delay in responding to your kind reply but I have been investigating.
Your comments are certainly valid and we were looking down these avenues already - everything is where it should be in terms of user authentication, profiles and data being in the same DC - we do appear to have a mix of MTU settings across the estate and equally I have some concerns around the Wyse client devices.
Currently from WAN router to XENAPP is ~30-50ms , switch is not much different yet client to XENAPP is pushing ~150ms. Clients are 100/Full and it's all looking good.
Odd, very odd.
Sent from Cisco Technical Support iPad App
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide