Hi
all,
I just come back
from an event where we had strange network problems. The
setting:
- a server program
(VW with OpenTalk StSt with wrapper UI on WindowsXP SP2)
- some client
programs on different machines in a LAN (VW, OpenTalk StSt, Pollock UI,
WinXP)
We were getting
lots of timeouts or server-not-responding errors on the clients. First we tried
WLAN and switched to LAN later. LAN was much better but still quite bad. Then I
remembered that a long time ago, I read something about Windows no reacting to
socket events when there are no window events on the machine. So, I instructed
the groups to always shake the mouse when sending requests to the server. At the
same time I was moving the mouse on the server. This helped! (I
think).
Did anybody see
something like this? And if yes, how can this be fixed? I am thinking
about generating window events from the applications and add a heard-beat to the
clients, so that stale ones can be detected and removed from the
server.
But this sound
quite awkward. Any other ideas?
Another issue is
that I have no clear idea how to test this. During development with server and
clients on the same machine everything was completely stable and fast. Quick
tests on my LAN were also reliable (although, I didnt test this seriously since
I was not expecting any problems with the network).
Any hints are very
welcome.
Cheers,
Christian
|
Thanks. This is interesting - I was using the VW
7.4d VM for both programs.
P.S. I hope you dont mind
that I reply to the list, since this might be interesting for others
too.
|
In reply to this post by Christian Haider-2
Christian,
A potential fix: <a href="http://worsethanfailure.com/Articles/Fisher_Price_Technology_Integration.aspx" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)"> Fisher Price Technology Integration:)Sean
|
In reply to this post by Christian Haider-2
Hi,
Christian Haider wrote: > Hi all, > > I just come back from an event where we had strange network problems. > The setting: > - a server program (VW with OpenTalk StSt with wrapper UI on WindowsXP > SP2) > - some client programs on different machines in a LAN (VW, OpenTalk > StSt, Pollock UI, WinXP) > > We were getting lots of timeouts or server-not-responding errors on the > clients. First we tried WLAN and switched to LAN later. LAN was much > better but still quite bad. Then I remembered that a long time ago, I > read something about Windows no reacting to socket events when there are > no window events on the machine. So, I instructed the groups to always > shake the mouse when sending requests to the server. At the same time I > was moving the mouse on the server. This helped! (I think). > > Did anybody see something like this? And if yes, how can this be fixed? > I am thinking about generating window events from the applications and > add a heard-beat to the clients, so that stale ones can be detected and > removed from the server. > But this sound quite awkward. Any other ideas? Does keeping the VM busy help, i.e. doing something like [[10+1] repeat] forkAt: Processor systemRockBottomPriority. > > Another issue is that I have no clear idea how to test this. During > development with server and clients on the same machine everything was > completely stable and fast. Quick tests on my LAN were also reliable > (although, I didnt test this seriously since I was not expecting any > problems with the network). > > Any hints are very welcome. > > Cheers, > Christian > > > Smalltalked Visuals GmbH > Christian Haider > Wiesbaden, Germany > http://www.smallCharts.de > > -- Ralf Propach, [hidden email] Tel: +49 231 975 99 38 Fax: +49 231 975 99 20 Georg Heeg eK (Dortmund) Handelsregister: Amtsgericht Dortmund A 12812 |
In reply to this post by Christian Haider-2
Hi Ralf,
> Does keeping the VM busy help, i.e. doing something like > [[10+1] repeat] forkAt: Processor systemRockBottomPriority. Maybe... I have problems with a test case. On my home LAN it is very snappy and quite reliable. Only in one of about 10 or 15 trials I can manage to get it stuck, but it does happen. Then I let your little process run and I couldnt make it freeze - I tried about 30 or 40 times. Then I terminated the process and after maybe 10 trials, it got stuck again. So maybe it works - have to find a way to reliably reproduce the problem... Thanks for your help, Christian |
Free forum by Nabble | Edit this page |