Hello,

We have bought new PC's and installed them with windows 2000.
When we run Baan on these PC's the sessions seem to have
a sort of time out. When you enter a field in a session and wait to long to go to the next field, baan needs 4-5 seconds to actually go to the next field. It is like the PC is extremeley occupied and needs a few seconds to return to the baan session.

This symtom is only for our new PC's. We have tried the following with no results:
1 inserted a new different PCI network adapter
2 took the PC to another network also running Baan (a completly different environment)
3 disabled the virusscanner and energy policies
4 started the PC in saftey mode with betwork
5 installed BW client 061 (an old one)
6 installed the latest bw client
7 changed the virtual memory settings

We have also called in help from HP but they don't know what the problem is.
The problem ONLY occurs when running Baan. All other applications work just fine.

Has anyone had any experience with this.

Any help would be much apprieciated.

best regards
thieuf

Not sure what help I can offer. I am tempted to say that the footprint of bw should be minimal, since all 'heavy' transactions are handled by the VM on the server.
Basically, all your bw.exe does is handle transactions between the user interface and the VM.

When you look at the processes in your taskmanager, does bw.exe consume an excessive amount of CPU or memory resources?
If not, what does?

Does the whole PC become sluggish, or just your BW?

I would re-examine your virus-protection software and other background processes. From what you decribe, something is checking your input. Maybe also check for some 'prankware'?

Good luck!

Hi

We seem to experience the same issue using HP Bl20p G2 blade servers running windows 2000 (up to date) and Baan B4c4.77 connecting to a HPUX clustered Baan Environment.

Could this issue be HP driver related - It also happens when using Windows 2003 as terminal server OS. But on our Dell Workstations the issue does not take place.

My 2 cents: Is the system (CPU) or disk busy during that time you use BW? If not, I would say it is a network issue. Do you have static or dynamic IP adresses for PC's? In case of static: are you sure no one else is using the same IP address?

I know a lot of stupid questions, but you never know....

Kind regards,
Dick

Maybe it has to do with multi-threading on new CPUs. Have a look at SOL 204348 for more information.

Günther

We have similiar problems using Vc on IBM blades. Up to now we suspect the Gigabits network cards but haven't been able to solve it.

For one user that experiences this a lot (very fast typing) we changed the client but that didn't help.

Using combo mode didn't help.

This same user doesn't experience this problem connecting to a full host mode installation running on a small IBM server.

Han

For one user that experiences this a lot (very fast typing) we changed the client but that didn't help.

Have you checked if the network has been set consistent and to a fixed value? Autodetect is one of the worst invention they ever did in the network area.

Hope this helps,
Dick

building a standard terminalserver box using a Dell Poweredge 1750 did not bring anything. (Latest Bios, Drivers. Broadcom Gigabit adapter @100FD and 1 CPU.)

It could be the network card - as it seems HP has rebranded Broadcom chips to HP and renamed the Broadcom driver B57W2K.Sys to Q57W2k.sys. Our Dell server uses the same broadcom chip. Our older Dell PC's dont't (they have an Intel Nic)

The newer Dell PC's we're buying have the Broadcom chip as well.....Unfortunately I have to wait a couple of weeks to get a new machine for testing.

I've installed an older HP server with a NON-Broadcom nic and I'll let you know if this works.

*Kick*

My 2cents (for now)
Our (WinXP) Dell Optiplex GL270 don't have this problem. Specs: Intel NIC & Integrated Graphics by Intel

Our (Win2k, Citrix MFXP) HP BL20P G1 and G2 blades(Broadcom Nic & ATI Rage XL gpu) have this issue.
Dell PowerEdge 1750 (Ati Rage & Broadcom) with Win2k/Citrix

I'm curious if it may have to do with either te NIC or the GPU. I mean - BW is sort of an XServer right? - it may be possible that somehow this combination is having an issue rendering the screen.

...it seems that BW.exe can't handle multiple processors (and hyperthreading - haven't tried dualcore machines, but i guess that's not any different) My guess is that as soon as the exe comes back from the cache to another core/cpu than where it originated it can't deal with this new situation correctly.
We were able to reproduce this behaviour by starting bw, then starting and running a bunch of other programs for a while, and then clicking on a tab in a form, or on a node in the baanexplorer (excuse the terminology - i'm not a baanguru) and then it would sometimes happen, and sometimes not.



If you use the taskmanager to look at the bw.exe process and the select Process Affinity, it shows processor 0 and 1 ticked. Untick one and the problem is gone, for this bw session.
This only works for this instance of the process so to get rid of this for ever you either have to make windows think it's a uniprocessor machine by disabling Hyperthreading/dualcpu in the BIOS, or if disabling is not an option (you don't always want to castrate your PC...) you can use Microsoft's imagecfg.exe which is in the NT4 cd in the support directory or in the windows2000 resource kit sup1. The command to execute is:

imagecfg.exe /u <path to executable> i.e. c:\imagecfg.exe /u c:\program files\baan\bin\bw.exe

this forces the bw.exe to run in uniprocessor mode (one processor at the time), resolving the issue (for us)
bw.exe cannot be running when you do this, and you need local admin rights.

What a wonderful, wellprogrammed, welldesigned piece of software this bw.exe is. And so up to date. Hopefully SSAGlobal or whoever owns the Baan brand this month will set this flag in a new bw.exe release so we don't have to do this on every pc.

We have the problem on uniprocessors as well.

Han

We have the problem on uniprocessors as well.

Han


You had the issue on IBM blades? And these have 1 cpu, not HT enabled?
You use citrix/terminal server, or how do I have to picture this?

bw is running on client (running xp), a lot of these clients are still uniprocessor, no HT.

The bshell is running on the server which is an IBM blade (dual core, HT turned on).

Han

bw is running on client (running xp), a lot of these clients are still uniprocessor, no HT.

The bshell is running on the server which is an IBM blade (dual core, HT turned on).

Han


Strange. We run the stuff on a HPUX server. Clients on Citrix Blades (Dual Proc) or XP (Uni). No problems on the PC as long as HT is turned off.

Can't help you there, i'm afraid. You may want to try to switch off HT on the blade. It wouldn't surprise me if that fixes the issue for you. But, i'm not a baanperson, so what the implications are performancewise I don't know.

Hello Everyone,

Thank you all for your Input. I redirected the input to our
system network administrator and finally we struck GOLD as they say...

The hyperthreading was the problem for the BW client. He disabled it on the PC's running Windows 2000 and they are running fine now.

I hope this information is enough for you to solve the problem.
If not, I'll post a cscreendump ...

Thanks again to you all.
Season's greetings and best wishes for 2007

Thieuf