Re: Crontab - HA

Perhaps best to use DBMS_SCHEDULER if you are on 10g.

A very good solution specifically for your situation (VCS) is here:

http://www.vxideas.org/cron-edit

This shell script is designed to be a resource in your VCS group. I used it for several years in a production cluster -- both a primary and a logical standby database, each with its own set of associated cron jobs, and this tool handled them both, whether running together on one node or separately on different nodes -- the correct cron jobs always ran exactly where they were supposed to, nowhere else. No need to clutter up your crontab entries themselves with some ugly if-then logic.

Unfortunately there was one small bug in the script, which is fixed by commenting out one line in the script as shown here. I sent the fix to the author but never got a response. :-(

add_entry() {
# this next line was a bug, un-doing what remove_entry did # cp $before $after

-- 
Mark Bole
http://www.bincomputing.com

Madhu R. Vajrala wrote:


> Hi All,

> 

> We recently implemented V.C.S. The only problem we had so far is with

> the "CRONTAB". Whenever the database(service group) fail-over to the

> other server in cluster, I will have to manually comment the entries in

> the crontab file in the primary server. I would like this to be taken

> care of automatically w/o DBA intervention. probably you might have hit

> the similar situation in the past, please advise from your experiences

> on how to handle.




--
http://www.freelists.org/webpage/oracle-l