> Here's the problem with IBM writing your startup script for you -
> WebSphere changes over time. Servers are added. Servers are removed.
> Server names
> can change. All of this must be reflected in your startup script.

Easy. That's what /etc/sysconfig files are for. 

Define one init.d instance (eg srv1, srv2, etc) for each server instance
-- named the same as the server instance -- and configure the
/etc/sysconfig/$0 (eg, /etc/sysconfig/srv1, /etc/sysconfig/srv2, etc)
files with the same name as the instance. The same script can be used if
the variables are appropriately substituted in the script. 

The Oracle example I cited in an earlier note demonstrates this trick
(eg, you can use the same script to start multiple Oracle instances with
completely different parameters by changing the /etc/sysconfig files for
each instance). And, both YaST and the RH widget have methods for
editing /etc/sysconfig entries in nice GUIs for the command-line
impaired. 

The only thing that inittab does right now that /etc/init.d scripts
can't do is automatically respawn a failed application if it dies. We'd
just need to write a "spawner" application that initialized, forked a
specified command line as a child, and monitored it. If it exited
unexpectedly, then the spawner app should respawn it. Write that app
once, and then you can take all that stuff out of /etc/inittab and put
in proper startup scripts. 

-- db

----------------------------------------------------------------------
For LINUX-390 subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: INFO LINUX-390 or visit
http://www.marist.edu/htbin/wlvindex?LINUX-390

Reply via email to