Tek-Tips is the largest IT community on the Internet today!

Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!

  • Congratulations Mike Lewis on being selected by the Tek-Tips community for having the most helpful posts in the forums last week. Way to Go!

Published Applications 1

Status
Not open for further replies.

RoyBottse

MIS
Jan 5, 2001
5
0
0
NL
The site is as follow:
1x Windows NT Server 4.0 SP6 (PDC)
2x Metaframe 1.8 SP2 (on WTSe SP6) both Member Server
2x Samba Servers v2.0.7 (To access User home dirs on
HPUnix11)
Clients use Xterminals with the latest ICA CLients for Unix
We work with Published Applications.

The Problem:
When the users logon, they randomly get some applications, sometimes all applications and sometimes they don't get any applications at all. In some cases the application appears in the published appl, with the custom ica connection icon. In other cases the icon of the application doesn't appear at all.
Citrix never heard of this problem. Any advise is welcome.
 
Hmmmm, I have had a similiar problem at one point. I have not worked with Unix so I may be way off.

When I did have the problem, it was because my citirx box was losing connection to my PDC... make sure your samba is doing what it should.

TinRam; CCEA, MCSE, CCNA
 
TinRam, thanx for the reaction, but Samba is doing the best it can at this moment. We made some modifications the past week and now it should be working correctly. The problem I think has to be within Citrix. I also keep getting CLTMGR messages on the Citrix Servers. Nowhere I can get a good answer. Not Microsoft nor Citrix. I contniue to search on the net.
 
Question: Are all your MetaFrame servers (UNIX and NT) on the same subnet?

The thing that immediately occurs to me is browsing. You know that by default the Windows 2000 servers will try to become Master browser.

Because it's an ICA browser, this shouldn't affect the UNIX boxes - unless they're on a different subnet, and the Gateways aren't exchanging information correctly.

I admit, it's a bit of a stab in the dark, but I'm working with a MFU/Win configuration at the moment, and was unable to connect to some apps until I'd updated the router table.

I hope this information is of some use to you
 
CitrixEngineer,

Thanx for the tip, but all the servers, including the HPUx, are in the same subnet. Further more we are in the same building (LAN situation). I myself think that there are maybe 2 problems occuring at the same time. This is making it all more difficult to solve.
I also have a new problem. As soon as one user (working on a X-Terminal) double klicks on his PN Icon, he gets a Dr. Watson generated by pn.exe. De message is: Dr. Watson Non Fatal Error, cannot open Dr. Watson logfile. Press OK to browse to new directory. If you select a new directory you get a Windows NT Errorcode =80. If you press cancel, the Windows NT Errorcode = 0.

Anyone help
 
Here's something you could try:

Try making one of your MFU servers the master browser. To do this, you will need to set ALL the other servers to Never become Master Browser.

The more I experiment with this, the more I am convinced that this is the area to examine. If a Master browser cannot get information from a Metaframe server in a specified time period, it will remove all the information it currently holds.

You will definitely need to test this out of hours!

Under MFU, the command is ctxbrcfg -m always, on the server chosen to become master, and ctxbrcfg -m never on the other. On the W2k servers, just click the radio button in Citrix Server Administrator. Then force an election, with ctxqserver -election.

For your other issue, try running Filemon, from
I hope this helps - good luck!
 
Hey Roy,

How are your client systems configured? Are they set to automatically search for available application farms or do they have static entries of the Citrix servers configured? Do you have the specific CLTMGR error that your coming across?

Da Ruster
 
He guys,

Problem solved. Thanx for the suggestions. The cause of all the problems, had to do with roaming profiles. Every Server (NT or CTX) had a profile stored for each user. Since there are three servers (1 PDC and 2 Ctx Servers) it could happen that if there occurs a network problem, one of the profiles could get corrupted. If you make the mistake to answer the question that your locally stored profile which is newer should be used, you could copy the corrupted profile to the other servers. On the long run it's possible that you have 3 differntly corrupted profiles and depending on the CTX Server you log on to, get different applications published. This we solved, by deleting all the userprofiles on all the servers.
On each CTX Server we added the DWORD Value DeleteRoamingCache=1 in the next key:
HKEY_LOCAL_MACHINE\SOFTWARE\Microsoft\WindowsNT\CurrentVersion\Winlogon
As far as I can tell, this only occurs when you work with Load Balancing and don't use Mandatory Profiles.
This we had to figure out our selves, there was no support from Citrix or Microsoft.

By the way user Royb = user RoyBottse


 
Status
Not open for further replies.

Part and Inventory Search

Sponsor

Back
Top