Rebooting every 2 to 3 months

[Davel323]

Metaframe 1.8, compaq proliant 800 server, windows NT 4.0 terminal server, service patch 6. This server only runs Smart switcher for lotus notes, allows multiple user to access email accounts. Notes resides on 4.11 Novell server. Problem: After running good for 2 to 3 months server will reboot and not allow anyone to log in. I've tried last good men, to get the last good configuration doesnt work, tried rescue disk, no way. The only way to recover, is to reinstall windows terminal server, metaframe everything. Im tired of recreating this server. Any Ideas would help.

Thanks

 

[CitrixEngineer]

2 - 3 months is a very long time for an NT4 TSE/MetaFrame server to keep running without being rebooted. Generally I write reboot scripts to make sure my servers are rebooted at least once a week.

If the period is 90 days, then it sounds like your NT4 Terminal Services licensing is not configured correctly, or you may be using a time-bombed version.

Is there anything in the Event log which may give any clues?

 

[Nostferatu]

I have to agree here with Engineer, many of us take for granted the actual function of Terminal server/Citrix servers and as such expect them to stay up like any other server, but if you really assess what these things are doing rebooting weekly just gives it that little break the poor things need.

I also run a script to reboot the servers and would recomend that once a month would be the very minumum frequency for this, there is no iron cast "Lore" for this, but just timelines found through experimentation.

Can't think why you have to rebuild, but I certainly would look through your event viewer very carefully.

If it is regular enough (I know you dont want it to happen again...) could you not sit next to the damn thing just before it goes pear shaped ? and see exactly what happens ?

All the above just to re-inforce engineers great advice, just giving a second opinion...

Regards,
Steve Marshall
Group IT
Technical Support Supervisor
(aka General Meddler)
The Expro Group
Reading
UK

 

[Voyager1]

Even Citrix Systems emphasizes the importance of what they term a "cycle boot". A TSE/Metaframe server keeps a lot of "balls" in the air, and really should be rebooted on a regular basis if for no other reason than to clear RAM and reset all memory pointers. They do, however, recommend that to get a clean reboot, especially by script in the middle of the night, stop the spooler service first. Add a "net stop spooler" to your reboot script... - Bill

"You can get anything you want out of life, if you'll just help enough other people get what they want" - Zig Ziglar

 

[Pritch5000]

I think Citrix recommend daily!
Curiously did you activate? because if you did then it won't be a time-limited trial version and also such details should be in the Citrix License Manager.
pritch

 

[Davel323]

yes, i have initialized the software. Why would you have to reboot your server daily. And if you do, have any of you wrote your own scripts to reboot at a certain time? If so, can you share?

Thanks

 

[gozer]

I may as well chime in here since this is exactly what I do with my two MF1.8 servers. The only "script" I wrote was a simple batch file I called reboot.bat. It consists of "shutdown /reboot". Then I use the scheduler that comes with Internet Explorer 5/5.5 (which I installed for just this reason) and schedule a nightly running of the batch file at around 3 or 4 am depending on which computer. The scheduler program is found my dbl clicking on "My Computer".

I haven't had any spontaneous reboots ever since!

Gozer the Gozarian

 

[Pritch5000]

You have to reboot daily because Metaframe is such a marvellous piece of software! no, seriously, its preventative maintenance, it clears all the registers. A Metaframe server is unlike other servers. Curiously how often do you shut your PCs down?
pritch.

 

[Davel323]

We only use it as a Dialin server, for people at home or away on business to recevied their email right now. We dont use it for any other applications because of the problems we've been having with it. I now have it scheduled to reboot every night. See how that goes.
Thank you all for your responces.

 

[clocke2000]

Has you ICA Browsing service stopped. This would prevent people from logging in. If you can't get in through a login at the console setup a pc with with srvmgr so you can open up the services and see what service has stopped. If you are unable to login to the console I believe there is a Microsoft Knowledge base article on this.

 

[MMB]

To resolve a host of issues I was having, I noticed that rebooting the MF 1.8 server resolved the issues.

So I wrote a batch file to reboot the server daily at 5:00am and my most of my problems have cleared up.

 

[Davel323]

yes, I started to reboot the server, it seam to clear up most of my problems.
Thank you all for you input