Hello team -

I was seeing this issue with RC2, and had hoped it would go away with the release code, but it didn't.

After building a CPS45FP1 farm (including pub apps of course) the servers don't immediately take incoming connections. With no intervention, apps don't launch from PN or from vAC. If I stop IMAService recreate the LHC, and restart the services, apps launch as expected and the server with the fresh LHC begins taking on users.

Presentation Servers won't take on any users until after the LHC has been recreated.

Any ideas what's happening? I could create a post-install package that recreates LHC, but that would seem superfluous.

Thank you!
RickD

Hi Rick,

are you sure that LHC is really corrupt?

I have sporadically a similar problem that sounds like yours, but in my case the LHC is okay, only the Load Evaluator reports untruly full load instead of 0. You can see this problem when you run “QFarm /load”. Workaround: Restart the IMA service or reboot the server or assign another Load Evaluator. In my case it was not necessary to recreate the LHC.

I have made many test installations to find the problem, but I had no luck.

I think this is not a general problem, it has an unknown factor. I have this problem in my own environment with CPS 4.5+R01,but not with CPS 4.0/3.0 and I know one customer environment with CPS 4.0+R02 and this problem, but I know many other CPS 4.5/4.0 customer environments without this problem.

I think the problem is not visionapp related, I think this is a general CPS 4.5/4.0 Hotfix Rollup Pack related problem. I have made some test installations without all visionapp related things, and I had the same problem!

If you look into the Citrix support forum you find others with similar problems. Maybe it has to do with this, but not 100% sure:
http://support.citrix.com/forums/thread....1&tstart=8 “There's another issue that started with 4.0 R02. They changed the default behavior of load evaluators. It used to be if they couldn't talk to PerfMon, they would go to 0. Now they go to 10000. (This is to prevent a black effect.), Check this: http://support.citrix.com/article/CTX112520

Another question, why do you report this now, and not after you have found it in RC2? :-)

Regards
Andreas

Thanks for the response Andreas!

When I do the next round of builds, I'll take a look at it more closely. With the build I wrote about, I was creating/assigning a custom load evaluator through vSM. From the looks of the articles you referenced above, I would expect the issue to persist through reboots, and I wouldn't expect dsmaint -recreatelhc to make it go away...

In retrospect, I should have bubbled this up to you during my RC2 work. At the time I didn't have it isolated very well, and figured it could have easily been a problem I introduced in one of my packages.

If I post on the next one I come up with, can I get a ride on the 'Ring in a stupid fast German sedan? ;-)

RickD

Hi Rick,

a colleague figured out that the problem has to do with the Load Evaluator in combination with Windows Multi User Interface (MUI) and/or Windows regional setting.

Workaround: http://support.citrix.com/article/CTX112520

Regards
Andreas