2 replies, from David Evans and Alan Reichert, who suggested putting debug statements in the script. I had already done this and the script itself was fine. After further discussion, David and I came to the conclusion that:
scripts in rc1.d do not run when the system boots normally
I have since put my script early in the sequence in rc2.d - and it is OK.
Why did all my other systems work?? Because this particular script, which lives in /etc/init.d , is linked to a script in BOTH rc1.d and rc2.d. I must have broken the link somehow, and even when I ended up with the file linked to a script in rc1.d, the script in rc2.d was different, and it was the one that was actually being run.
Sparc 10 clone was recently upgraded from 2.5 to 2.6 with no apparent problems. All rec.patches and Y2k patches installed, Sunscan 2.0 is happy.
Problem is, a script in rc1.d isn't running any more. A cron entry creates a flag file, then the system reboots. On the way up, the rc1.d script should identify the flag file, and the backups are run. Or at least it used to.
I have upgraded a dozen systems here, all use the same mechanism for backups, and no others have this problem. I have tried linking the file from /etc/init.d , and I have tried it on its own. I have verified that the script itself is okay by running it manually while in single user mode. The script must run in single user mode before the databases start up.
/etc/inittab was modified as part of the upgrade, but rc1 entries seem to be the same.
Has anyone else had this problem, or any ideas please?
Qld Health, Brisbane, Australia
This archive was generated by hypermail 2.1.2 : Fri Sep 28 2001 - 23:13:20 CDT