Interrupted cfgmgr

Interrupted cfgmgr

Post by Rachid Kouch » Fri, 04 Jul 2003 20:14:43



Hi,

When "cfgmgr -l scsi0" is interrupted by a CTRL-C, a subsequent "rmdev"
returns :

# rmdev -l hdisk0 -d
rmdev: 0514-516 Device configuration database lock service timed out.

I suppose that "cfgmgr" issued an "odm_lock()" but after the
interruption, it
didn't released the lock.

So, in this situation, what is possible to do to find and remove the
lock ?

Regards.

P.S. : I am running AIX 4.3.2

 
 
 

Interrupted cfgmgr

Post by Rachid Kouch » Fri, 04 Jul 2003 20:49:02


Sorry,

"cfgmgr" was called by a script and the CTRL-C did not interrupt "cfgmgr"
but only the script. That's why the lock was still active...


> Hi,

> When "cfgmgr -l scsi0" is interrupted by a CTRL-C, a subsequent "rmdev"
> returns :

> # rmdev -l hdisk0 -d
> rmdev: 0514-516 Device configuration database lock service timed out.

> I suppose that "cfgmgr" issued an "odm_lock()" but after the
> interruption, it
> didn't released the lock.

> So, in this situation, what is possible to do to find and remove the
> lock ?

> Regards.

> P.S. : I am running AIX 4.3.2


 
 
 

Interrupted cfgmgr

Post by Andreas Schul » Sat, 05 Jul 2003 20:29:40



> Sorry,

> "cfgmgr" was called by a script and the CTRL-C did not interrupt "cfgmgr"
> but only the script. That's why the lock was still active...


> > Hi,

> > When "cfgmgr -l scsi0" is interrupted by a CTRL-C, a subsequent "rmdev"
> > returns :

> > # rmdev -l hdisk0 -d
> > rmdev: 0514-516 Device configuration database lock service timed out.

> > I suppose that "cfgmgr" issued an "odm_lock()" but after the
> > interruption, it
> > didn't released the lock.

> > So, in this situation, what is possible to do to find and remove the
> > lock ?

> > Regards.

> > P.S. : I am running AIX 4.3.2

Hi Rachid,

from your second posting I assume that the problem was solved without
your input when the cfgmgr finished. However, should you happen to get
the error message after completion of an operation that changed the
ODM the ODM lock file should be here: /etc/objrepos/config_lock

Check the 4.3 documentation how the check is done:
#include <cf.h>

            if (odm_initialize() < 0)

            exit(E_ODMINIT); /* initialization failed */

            if (odm_lock("/etc/objrepos/config_lock",0) == -1) {

            odm_terminate();

            exit(E_ODMLOCK); /* database lock failed */

Quote:}

Regards,
Andreas
 
 
 

1. NEx000 and 1.2.9: Interrupted while interrupts are masked!

i'm using the NEx000 ethernet driver (i have an E1000) and kernel
1.2.9.  while kernels before 1.2.5 or so used to crash with this card,
this one has been perfectly stable.  however, i get the following
message in /var/adm/messages:

kernel: eth0: Interrupted while interrupts are masked! isr=0x1 imr=0x5.

i get many of these messages a day, perhaps 20, and sometimes grouped
together (middle of transfer?) just a few seconds apart.  however,
there are no problems -- just this odd message.  anyone have
any idea what's going on?
--
Stephen Farrell

                                http://128.135.108.36/

2. regard perl5.004.03.SPARC.Solaris.2.5.pkg.tgz

3. /var/log messages: eth0: interrupted while interrupts are masked!

4. Red Hat on floppies

5. Is '/proc/interrupts' auto update when a interrupts come ?.

6. Configure Crack question

7. Interrupted while interrupts are masked!

8. I've a problem with a remote printer on SCO

9. switching to interrupt contex when no interrupts

10. Interrupted while interrupts are masked!

11. Locating interrupts RE /proc/interrupts

12. Slow Interrupt vs. Fast Interrupt

13. Pls Help - CPU Lock-up on Interrupt with Shared Interrupt Vector