Hello gurus,
you all were right: Network!!
Because I dont use NIS, NFS, DNS, BIND ... renamed some files in /sbin/rc3.d
in order to "optimize" the system:
no_S18nis
no_S19nfs
no_S20nfsmount
no_S40sendmail
no_S65lpd
Now I found error messages in syslog.dated/xxx/daemon.log:
Feb 25 16:52:13 GOAW15 inetd[716]: Cannot register service: RPC: Timed out
Feb 25 16:52:56 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:53:18 GOAW15 inetd[716]: Cannot register service: RPC: Timed out
Feb 25 16:54:01 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:54:23 GOAW15 inetd[716]: Cannot register service: RPC: Timed out
Feb 25 16:55:06 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:55:28 GOAW15 inetd[716]: Cannot register service: RPC: Timed out
Feb 25 16:56:11 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:56:33 GOAW15 inetd[716]: Cannot register service: RPC: Timed out
Feb 25 16:57:16 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:58:21 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 16:59:26 GOAW15 /usr/dt/bin/ttsession[982]: Cannot register service:
RPC:
Timed out
Feb 25 17:09:11 GOAW15 last message repeated 9 times
I dont know why I get theses errors - I think something /etc/inetd.conf is
wrong - but after renaming /sbin/rc3.d/no_S19nfs to S19nfs (as it was!) the
error messages went away and CDE started quick as expected.
My solution: nfs must be started at boottime.
Any ideas why I have to start nfs? What changes to /etc/inetd.conf are
necessary to kick nfs away?
Thanks again to all who answered, this is a great list.
Werner Rost
-------------------------------------------------
Hello gurus,
many thanks to your answers, specially to
Dr. Thomas.Blinn_at_Compaq.com [tpb_at_doctor.zk3.dec.com]
Octave Orgeron [unixconsole_at_yahoo.com]
Harald Baumgartner [hmb_at_mpe.mpg.de]
Reiner Dassing [dassing_at_wettzell.ifag.de]
Marco Benton [bentonm_at_polaroid.com]
Peter.Stern_at_weizmann.ac.il
Paul Henderson [Paul.G.Henderson_at_compaq.com]
Alex Harkema [HarkemaA_at_vertis.nl]
You will find my original question below.
Conclusion: Sorry, but I give up. Yet I am not able to solve the problem.
That is not very bad because I do not use the console very often. In most
cases (98 %) I log on via telnet, look around if there are any problems, not
more. I do not really need the console. Therefore I have configured XDM
instead of CDE and all works well (though I am not really happy because of
this unsolved problem).
Nevertheless lets talk about your answers and my actions.
Most of you told me to check my network configuration. But I cannot believe
that netwok cause the CDE-problem. I have configured a "small" network: Only
TCP/IP, no NIS, no NFS, no DNS, no BIND, nothing else. All works fine (exept
CDE ;-) ) - the proxy server SQUID is the only application that runs on my
workstation. It grants access to the internet via a parent proxy and a
firewall - no problems, excellent performance! SQUID asks a WIN NT 4.0
server for user authentication - no problems too.
There exists following entry in /etc/hosts (Paul asked for):
127.0.0.1 localhost
Renaming the directory /.dt to /.dt.old and logging in again (Peter Stern,
Octave Orgeron) did not solve the problem.
According Marco Benton I shold check /etc/svc.conf and make sure
hosts=local,bind. Because I dont use bind I did not change the entry
"hosts=local".
Checking /.dt/errorlog I found following error message:
/usr/dt/bin/sdtdbcache: A multi-line field in the definition "Xwd"
in the file
GOAW17:/usr/dt/appconfig/types/de_DE.ISO8859-1/xclients.dt"
is incomplete. A \\" character may be missing
in the multi-line field.
I found the wrong line in this file, after correction the error message went
away, but that did not solve the problem.
The core-file is generated from "Xdec" which tell me the file command.
Another observation: After successfull (but LONG) CDE-login I waited about
24 hours. Then I saw many processes dtscreen, dtexec:
785 processes /usr/dt/bin/dtexec
1177 processes /usr/dt/bin/dtscreen
I found these processes because system refused calling vi ("no more swap
space available").
I observe the CDE-problem on 3 workstations Digital Personal WorkStation
600au / Digital Personal WorkStation 500au. For each workstation I did:
- fresh installation of TU 5.1A
- Patchkit 1
- install DIA (DECevent report utility)
- install driver for powerstorm
- install SQUID and utilities.
Maybe the installation of SQUID or the utilities causes the problem? Harald
Baumgartner uses CDE after an upgrade with great success - but probaly
without SQUID. I am sure I will investigate some time checking this
possibility - if my boss allows ;-). Then I will tell you again stories
about CDE.
My original question:
Environment:
Digital Personal Workstation 600au
Tru64 V5.1A (Rev.1885) Patchkit 1
Firmware 7.2-1
Graphics: Powerstprm 4D51T
ingr0: Attached 4D51T.2.0.80.0xa FB.0.0.71.13.100.4 TP.45.1_at_1Mx16
Problem:
When I try to login at the CDE-console log on takes about 10 minutes! After
that all works fine.
Before I installed Patchkit 1 log on lasts about 10 minutes too but even
created a core file. After installation of PK 1 no core file is written.
Why have I to wait so long during log on?
thanks again to this great list
Werner Rost, KIR
ZF Boge GmbH
Tel.: 0228/3825 420
Fax: 0228/3825 398
---------------------------------------------------------
This Mail has been checked for Viruses
Attention: Encrypted mails can NOT be checked!
**
Diese Mail wurde auf Viren geprueft
Hinweis: Verschluesselte mails koennen NICHT auf Viren geprueft werden!
---------------------------------------------------------
Received on Mon Feb 25 2002 - 16:34:33 NZDT