Unix/Linux Go Back    


SCO Santa Cruz Operation (SCO) was a software company based in Santa Cruz, California which was best known for selling three UNIX variants for Intel x86.

Strange behaviour on Openserver 5.0.2 after 09/2015

SCO


Tags
solved by upgrading to 5.0.6

Closed    
 
Thread Tools Search this Thread Display Modes
    #1  
Old Unix and Linux 09-08-2015
ChipperEs ChipperEs is offline
Registered User
 
Join Date: Sep 2015
Last Activity: 15 April 2016, 6:36 PM EDT
Posts: 24
Thanks: 8
Thanked 5 Times in 4 Posts
Strange behaviour on Openserver 5.0.2 after 09/2015

Dear friends,
i am having some Problems with a Sco Openserver in a Box (normally on VMWare, importing it on a Virtualbox does the same).

When i boot up the system on xx/09/2015 i got fork failed - too many processes during startup.
The programs which should run on the server, do not run.
I can login, but i got no shell, instead i got: cannot fork to run /usr/bin/mapkey: Resource temporarily unavailable (error 11).

When i start the single user mode, then change the date to 08/2015, for example 31/08/2015, and then restart the server; it comes up like a charm, without any error.

If i change then the date of the system to the actual date the Server is working.

I have already tried a lot to make the server start the programs normally at boot time with the right date, but no luck:
disabling init scripts <P80
adding second swap space
adjusting Kernel Parameters (just doubling the values) on scoadmin -> kernel manager -> tune parameters -> users and group configuration

Does anybody know this kind of behaviour and has a hint for me?

thanks a lot
The Following 2 Users Say Thank You to ChipperEs For This Useful Post:
mig_2929 (10-05-2015), Retz (09-14-2015)
Sponsored Links
    #2  
Old Unix and Linux 09-08-2015
jgt's Unix or Linux Image
jgt jgt is offline Forum Advisor  
Registered User
 
Join Date: Apr 2007
Last Activity: 25 September 2017, 7:01 PM EDT
Location: 44.21.48N 80.50.15W
Posts: 1,977
Thanks: 2
Thanked 265 Times in 248 Posts
You cannot start the machine with a date earlier than the install date. The install date is entered MMDDYYYY, but your post uses DDMMYYYY. Did you possibly set up the system with the wrong date?
Sponsored Links
    #3  
Old Unix and Linux 09-08-2015
ChipperEs ChipperEs is offline
Registered User
 
Join Date: Sep 2015
Last Activity: 15 April 2016, 6:36 PM EDT
Posts: 24
Thanks: 8
Thanked 5 Times in 4 Posts
in Germany we write the date like this: DD.MM.YYYY
When i start the machine with a day in Aug 2015, it works
when i start it with the actual date 8th Sep 2015 i got the described errors.

the installation date is somewhere in 1998. that should not be the problem.
i also checked the license of the sco server. it should run through Jan 2038
    #4  
Old Unix and Linux 09-08-2015
jgt's Unix or Linux Image
jgt jgt is offline Forum Advisor  
Registered User
 
Join Date: Apr 2007
Last Activity: 25 September 2017, 7:01 PM EDT
Location: 44.21.48N 80.50.15W
Posts: 1,977
Thanks: 2
Thanked 265 Times in 248 Posts
Try starting the machine with an August date, then as root change the date.

Code:
#date -t 20150908hhmm.ss; setclk -r

Sponsored Links
    #5  
Old Unix and Linux 09-08-2015
ChipperEs ChipperEs is offline
Registered User
 
Join Date: Sep 2015
Last Activity: 15 April 2016, 6:36 PM EDT
Posts: 24
Thanks: 8
Thanked 5 Times in 4 Posts
yes, that works. That is what i already did to keep the services on the machine working.

i have to set the date in the past before i shutdown and then regain it with setclk -r after boot. on the last init script for example.

this is a good workaround.

but it would be nice if there is a way to make the sco machine start properly with the right actual date
Sponsored Links
    #6  
Old Unix and Linux 09-08-2015
jgt's Unix or Linux Image
jgt jgt is offline Forum Advisor  
Registered User
 
Join Date: Apr 2007
Last Activity: 25 September 2017, 7:01 PM EDT
Location: 44.21.48N 80.50.15W
Posts: 1,977
Thanks: 2
Thanked 265 Times in 248 Posts
Should I infer from your first post that you have, or can create, a test installation.
If so, is any evidence of where the failure occurs left in either /usr/adm/messages, or in the /etc/rc2.d/messages directory?
Is there a possibility that some non SCO application has an end date?
Sponsored Links
    #7  
Old Unix and Linux 09-08-2015
ChipperEs ChipperEs is offline
Registered User
 
Join Date: Sep 2015
Last Activity: 15 April 2016, 6:36 PM EDT
Posts: 24
Thanks: 8
Thanked 5 Times in 4 Posts
yes, i have a test system. every week, the system is been shut down and backup-ed automatically. that is why the clean startup with the right date is necessary.

/usr/adm/messages contains only the spalsh screen from the startup, and the .log files in /etc/rc2.d/messages contain only the messages with fork failed - too many processes.

the non-sco software running on the server is loaded by S99, long after the first fork failed message apperars

---------- Post updated at 08:17 AM ---------- Previous update was at 07:44 AM ----------

by the way, on normal startup the first message appears after "Starting TCP services: snmpd routed"
Sponsored Links
Closed

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Linux More UNIX and Linux Forum Topics You Might Find Helpful
Thread Thread Starter Forum Replies Last Post
Crontab strange behaviour enux Red Hat 3 04-17-2011 08:54 AM
strange behaviour from sed??? alirezan Shell Programming and Scripting 2 10-04-2008 06:19 PM
Strange sed behaviour vino UNIX for Advanced & Expert Users 8 02-12-2008 05:51 AM
A Strange Behaviour!!! navojit dutta Shell Programming and Scripting 5 12-21-2007 03:35 AM
/etc/passwd strange behaviour! penguin-friend Linux 0 06-06-2005 12:00 PM



All times are GMT -4. The time now is 05:09 AM.