it does not anyhow inform of this but the whole domain can be missing even thou the login goes well.

quote:

---

KiXtart 2001 and the console
KiXtart 2001 is provided in two ‘flavors’: the standard console-based version and a Windows version. The Windows version will only display a console if and when any output is sent to the screen. If desired, this behavior can be overridden using the /I (Invisible) commandline option.

Note
By default, the Windows version of KiXtart 2001 runs as an asynchronous process. This means that if you start WKIX32.EXE from a batchfile, the batchfile will not wait for KiXtart to exit and will continue processing. This behavior can cause problems if KiXtart is being used as part of the logon process, especially on Windows 9x clients. To prevent these problems, WKIX32.EXE should be started from a batchfile using the START command with the wait option, e.g.: “START /W WKIX32.EXE”. Optionally, on Windows NT or higher, you can also specify the /B option with the START command, to prevent the creation of an additional window.

The console version behaves exactly like KiXtart 95, and will automatically cause a console window to be created upon startup.

---

hope this helped...

Since you state the mappings are taking effect, no real problem exists here except for the time delay. To my end users I would say: Yes, it is slower because you are on a dial-up line, nothing will speed this process up except the end-user having a faster connection to the internet (or ISDN dial-up lines), you will have to learn to live with it, and dispose of this issue.

Bill

Perhaps I am mistaken on the underlying KiX technology, but essentially a net use function is taking place by KiX initiating a command of some time for Windows to use the shared resource (perhaps simply executing NET USE - notice the KiX syntax USE H:, USE H: /DELETE which closely mimics NET USE command structure ).

In another example, try a KiX COPY command, and compare the speed with a Windows COPY command. Would you expect one command to perform faster than the other? I would not.

Alex's suggestion may provide some performance improvement. I can not say as I have never tested. I recommend you attempt some timing tests to see if in fact there is significant performance improvement.

Additionally, you likely would need to look at where the login script resides when executing - on the client, or on the server. You certainly want the DLL's on the clients. As for the script(s) on the client, well, that is not a solution I would put in place yet.

Regardless, each network admin must decide the acceptable performance of the systems connecting to their network. I am comfortable telling my end users they will suffer a performance lag via dial-up. They in fact are quite prepared to accept that a dial-up 56/33/28kbps connection will never perform any network functions at the same speed as the 100mbps LAN. Should such timing differrences translate into a material impact on business operations, then of course one should consider revisiting it.

Bill

I am writing this from home right now, RAS'd in to my work at 26.4 Kbps, and there is no significant difference in speed testing either "NET USE" or KiX's "USE". Both were nearly instant.

Try RASing in without VPN and without running the logon script. From a DOS prompt issue a NET USE command and time it. Then write a one line KiX script with the USE command and time it. Then try it with VPN over RAS.

It may be useful to know what OS the clients are. Also and whether there are protocol, provider, or WINS issues. If you have unnecessary protocols, the binding order may not be optimal. If you have more than one provider, the order may be wrong.

Other possible issues may be with name resolution. Try populating your LMHOSTS and HOSTS with your servers' names and IPs to troubleshoot.