OW login

[manleyt]

Has anyone seen this issue? OneWorld does not respond at time of login (on Citrix), the user just gets an hourglass. This only happens after the automatic nightly reboot of the 5 Citrix servers, but never to more than one or two of them at a time, and even then not every day. The user may eventually login after 20 minutes or more, but response is slow after it does. All five servers are identical IBM xSeries 230 Dual 1Ghz PIII, with identical configuration.
When this issue occurs, I am still able to connect to that particular Citrix server with a console session (I login and get an NT virtual desktop session). I can also run any application on the server through this session without any trouble (except for OneWorld). A restart of the server usually clears this up, but sometimes it takes a "hard" shutdown. I did find that a "NETLOGON" error (event 5719) showed up on the server at the time of nightly reboot any time this happened (we have an admin account that "auto. logs on" at boot up).
I have since resolved the "NETLOGON" error. IBM and JDE support have not been able to come up with anything, but I found a reference to the error on Microsoft's KB. It appears that this can occur when a server or workstation is plugged directly into a switch (as mine are). The "Spanning Tree Algorithm blocks the port for 10-15 seconds when a port becomes active, while it checks for a loop condition. If the server is trying to login during this time, it will not find the domain because there is no traffic allowed through the port. Our servers are set to autologin, so when this happens, they log in with a cached account. Apparently the timing is such that sometimes the servers make it, other times they don't. The workaround for this was to add a parameter to the registry that would tell the server to wait 30 seconds before login. This resolved that symptom, but not the OW login problem. I thought I was still on the right track, so I disabled "spanning tree protocol" on all ports that the servers are plugged into, but this morning two of the servers had the OW login problem again.
One other thing I did notice is that I cannot see our AS400 enterprise server in "Network Neighborhood", although I am able to ping the name or IP address from any of the servers and get a response.

Any suggestions would be greatly appreciated!

 

[alexander_chev]

Hello, Terry,
I am not sure if it will help, but we had some Citrix login issue like you
described. It was caused that all users used the same JDE.INI file on the
Citrix server. So it was locked pretty often. It was resolved by setting
that each user had his/her own JDE.INI.
I hope it helps.

Alex.

 

[mats_thunholm]

Hi=20

Are you using the 'W' environment ?, we had a similar problem some month=20
back and we were using
the 'W' environment, usually this happened in the afternoons when quite a=20
lot of users were logged on
to our Citrix Server (NT4 at that time), in the debug log I could see that =

OneWorld was waiting to run
a BSFN called 'Vertex validation .............', that BSFN were mapped to=20
run on the ES in the 'W' env,
I mapped that + some more BSFNs to run locally on the Citrix Server and I=20
could see a bit better=20
logon performance but not much.
So we tested the normal env. (like PY7333) and the users were logged on in =

5-8 seconds.=20

Actually what the problem really was the debug log never told me, the=20
signon process was just
hanging at the BSFN Vertex .........

Hope it helps a bit.=20

Best regards=20
Mats Thunholm

---------------------------------------------------------------------------=
------------------------------
Mats Thunholm
Amersham Biosciences
Uppsala, Sweden
+46 18 612 0646 (w), +46 70 666 5819 (m)
[email protected]
---------------------------------------------------------------------------=
------------------------------




manleyt <[email protected]>
Sent by: [email protected]
2003-02-25 20:24
Please respond to jdelist

=20
To: [email protected]
cc:=20
Subject: OW login


Has anyone seen this issue? OneWorld does not respond at time of login=20
(on Citrix), the user just gets an hourglass. This only happens after the =

automatic nightly reboot of the 5 Citrix servers, but never to more than=20
one or two of them at a time, and even then not every day. The user may=20
eventually login after 20 minutes or more, but response is slow after it=20
does. All five servers are identical IBM xSeries 230 Dual 1Ghz PIII, with =

identical configuration.
When this issue occurs, I am still able to connect to that particular=20
Citrix server with a console session (I login and get an NT virtual=20
desktop session). I can also run any application on the server through=20
this session without any trouble (except for OneWorld). A restart of the=20
server usually clears this up, but sometimes it takes a "hard" shutdown. I =

did find that a "NETLOGON" error (event 5719) showed up on the server at=20
the time of nightly reboot any time this happened (we have an admin=20
account that "auto. lo!
gs on" at boot up).=20
I have since resolved the "NETLOGON" error. IBM and JDE support have not=20
been able to come up with anything, but I found a reference to the error=20
on Microsoft's KB. It appears that this can occur when a server or=20
workstation is plugged directly into a switch (as mine are). The "Spanning =

Tree Algorithm blocks the port for 10-15 seconds when a port becomes=20
active, while it checks for a loop condition. If the server is trying to=20
login during this time, it will not find the domain because there is no=20
traffic allowed through the port. Our servers are set to autologin, so=20
when this happens, they log in with a cached account. Apparently the=20
timing is such that sometimes the servers make it, other times they don't. =

The workaround for this was to add a parameter to the registry that would=20
tell the server to wait 30 seconds before login. This resolved that=20
symptom, but not the OW login problem. I thought I was still on the right=20
track, so I disabled "spanning tree!
protocol" on all ports that the servers are plugged into, but this=20
morning two of the servers had the OW login problem again.=20
One other thing I did notice is that I cannot see our AS400 enterprise=20
server in "Network Neighborhood", although I am able to ping the name or=20
IP address from any of the servers and get a response.=20
Any suggestions would be greatly appreciated!
Terry Manley
IT Specialist
Timken Super Precision
Xe, SP19.1,Update 6, AS400,
Citrix Metaframe 1.8 on NT4-TSE
--------------------------
To view this thread, go to:
http://www.jdelist.com/ubb/showthreaded.php?Cat=3D&Board=3DOW&Number=3D50466
+ - - - - - - - - - - - - - - - - - - - - - - - -+
This is the JDEList One World=AE / XE mailing list/forum.=20
Archives and information on how to SUBSCRIBE, and=20
UNSUBSCRIBE can be found on the JDEList Forum at=20
http://www.JDEList.com

JDEList is not affiliated with JDEdwards=AE

+ - - - - - - - - - - - - - - - - - - - - - - - -+





OW XE, SP18.1 + one off's, AS400 V4R5
Mats Thunholm
AmershamBiosciences
Uppsala, Sweden

 

[pianosynth]

We had something similar quite some time ago. The best I can remember, there was something wrong with one of the NT BDCs that was causing DNS resolution to intermittently fail. Once we identified the offending BDC and took it offline, everything worked properly. HTH...

 

[manleyt]

Re: RE: OW login

Alex,

I use roaming profiles served from a file server share, so everyone has their own jde.ini file. The only time they use the ini on the Citrix box is if I delete their roaming profile. The user then gets the server's 'Default" profile, which then gets copied off to their individual roaming profile share when they log out.

Thanks for your reply, it was good information anyway.

 

[manleyt]

Were you running multiple Citrix servers? Would you have some servers work fine, with one or more having issues at the same time?

I have 2 BDCs on site, so I may try shutting down one at a time. Thanks for the info. Any more detials you can think of would be apppreciated.

Terry

 

[mats_thunholm]

No, at that time we had only one Citrix server.

Cheers=20
Mats Thunholm

---------------------------------------------------------------------------=
------------------------------
Mats Thunholm
Amersham Biosciences
Uppsala, Sweden
+46 18 612 0646 (w), +46 70 666 5819 (m)
[email protected]
---------------------------------------------------------------------------=
------------------------------




manleyt <[email protected]>
Sent by: [email protected]
2003-02-25 22:50
Please respond to jdelist

=20
To: [email protected]
cc:=20
Subject: Re: OW login


Were you running multiple Citrix servers? Would you have some servers=20
work fine, with one or more having issues at the same time?
I have 2 BDCs on site, so I may try shutting down one at a time. Thanks=20
for the info. Any more detials you can think of would be apppreciated.
Terry
Terry Manley
IT Specialist
Timken Super Precision
Xe, SP19.1,Update 6, AS400,
Citrix Metaframe 1.8 on NT4-TSE
--------------------------
To view this thread, go to:
http://www.jdelist.com/ubb/showthreaded.php?Cat=3D&Board=3DOW&Number=3D50502
+ - - - - - - - - - - - - - - - - - - - - - - - -+
This is the JDEList One World=AE / XE mailing list/forum.=20
Archives and information on how to SUBSCRIBE, and=20
UNSUBSCRIBE can be found on the JDEList Forum at=20
http://www.JDEList.com

JDEList is not affiliated with JDEdwards=AE

+ - - - - - - - - - - - - - - - - - - - - - - - -+





OW XE, SP18.1 + one off's, AS400 V4R5
Mats Thunholm
AmershamBiosciences
Uppsala, Sweden

 

[pianosynth]

At the time, we had about 6 BDCs at our main site with another dozen or so around the world. Also had 3 Citrix servers in a farm. That was when we still had NT4 and Citrix 1.8, so it may apply to your situation. I'll talk to our NT guys and see if anyone clearly remembers all of the details. I was already full-time CNC by then so I wasn't closely involved in the solution.