Your independent source for application and desktop virtualization.
Sign in
|
Join
Topics
Blogs
Forums
Videos
Books
Events
About Us
Discussion Forums
»
Tech Support Forums
»
Citrix XenApp / Presentation Server
»
Protocol Driver Error - users getting dioconnected from applications after connecting successfully
Protocol Driver Error - users getting dioconnected from applications after connecting successfully, in the Citrix XenApp / Presentation Server forum on BrianMadden.com
rated by 0 users
This post has 0 verified answers | 3 Replies | 0 Followers
Suggest an Answer
Reply
Points 1,381
Paul Reed
posted on Mon, Jun 2 2008 8:14 AM
rated by 0 users
Our biggest problem, that I've tried to troubleshoot from several angles, is users getting disconnected from their Citrix session. They'll be working away for an hour or so and then get a Protocol Driver Error popup and the session disconnects. Its not a network issues as we have SNMP configured on all routers to alert us if there are any outages. We have 2 CSG's using NLB and if one user gets diosconnected, other users can continue to work. We currently have 6 production farm servers (which we'll be increasing to 9) which are PS4.0 and pacthed up to date. The disconnects are not linked to a particular server either, though they are all equally loaded. Clients are Web ICA 9.23 on Windows XP.
In short, it is extremely difficult to troubleshoot as it is so intermittent. Also, most of the results from internet searches (this site adn the official Citrix support site) mention Protocol Driver Error relating to problems launching an application (usually STA errors) rather than after an application has been successfully connected?
What can be causing these disconnects? Would farm server load cause it? CSG load?
Thanks
Macs and Citrix...what else could anyone possibly need?
| Post Points: 20
All Replies
Points 945
Reply
Jody Howard
replied on
Mon, Jun 2 2008 9:22 AM
rated by 0 users
We had a similar problem related to NetScaler deployment. User would log on and work fine then suddenly get the PDE. A ping at the start of the session to the Web Interface showed IP A and ping after PDE showed WI IP as B.
Turns out the load balancing was bouncing between the primary and secondary sites, XML request were taking longer than expected. Our issue was addressed by decreasing the sensitivity for fail over on NetScalers load balancing. Another option is to look at the session persistence via the client machine IP address.
Hope this puts you in the right direction.
JL
| Post Points: 20
Points 1,381
Reply
Paul Reed
replied on
Mon, Jun 2 2008 10:30 AM
rated by 0 users
Thanks Jody. Sadly I've gone down this route to some extent by turning off one of the servers in the NLB cluster. I have set Single Affinity in the NLB properties anyway, which is supposed to safe-guard against this anyway.
Macs and Citrix...what else could anyone possibly need?
| Post Points: 5
Points 1,381
Reply
Paul Reed
replied on
Wed, Jun 4 2008 5:32 AM
rated by 0 users
It looks like a problem with Session Reliability and Client 9.23. Even though Citrix say this was fixed in Hotfix SGE300W005 (we have Hotfix SGE300W008) installed, its still a problem.
I installed 10.2 on a test machine, pulled the cable, reconnected it and it worked ok.
Macs and Citrix...what else could anyone possibly need?
| Post Points: 5
Previous
|
Next
Page 1 of 1 (4 items) |
RSS
Copyright © 1997-2010 TechTarget |
Disclosures
|
Privacy Policy
|
Contact Info
BrianMadden.com
|
SearchVirtualDesktop.com
|
SearchEnterpriseDesktop.com
|
SearchServerVirtualization.com
|
SearchVMware.com