SUMMARY: Machine Hanging repeatedly

D. Ellen March (emarch@pinole1.com)
Tue, 22 Apr 1997 17:17:20 -0700 (PDT)

Original question:

>>We have a sparc 4, running solaris 2.5 that hangs repeatedly.

>>I am seeing the following message in the /var/adm/messages file:

>>Apr 21 08:17:47 foobar rpc.ttdbserverd[306]: NetISAM: read failed
>>Apr 21 08:17:47 foobar NetISAM[306]: Fatal error: read failed - UNIX errno 11

>>Can anyone point me towards what NetISAM is? And also perhaps
>>how to fix it?

Thanks to the following managers:

Shirley Chee Mei Kin
Rene Occelli
Chan Cao
Alfredo Sola
Mark Hargrave
Michael Sullivan

Shirley and Rene said it best (here is Shirley's reply):

SOLUTION SUMMARY:

This message indicates that at least one of the tooltalk databases
located on your machine is having a problem. Either files are missing
or the state of the database indicates that it is not able to service
your requests.

If you see this try the following:-

# kill -STOP <pid-#-of-rpc.ttdbserverd>

# df -kF ufs | awk '{if (NR>1) print $6 "/TT_DB"}'

This should print a list of directories in the root of
filesystems called TT_DB E.g.

//TT_DB
/usr/TT_DB
/var/TT_DB
/ftp/TT_DB
/opt/TT_DB
/usr/openwin/TT_DB

If this encounters no problems then go ahead with the following
command, remember though to ensure the system has adequate back
ups, or alternatively, remove the directories manually.

# rm -rf `df -kF ufs | awk '{if (NR>1) print $6 "/TT_DB"}'`

Once the directories have been removed kill rpc.ttdbserverd as
follows:-

# kill -KILL <pid-#-of-rpc.ttdbserverd>

The next request to rpc.ttdbserverd will restart the daemon and the
databases will be rebuilt.

Ellen
emarch@pinole1.com