AW: AW: [Linux-HA] Apache failover / renaming the binary

[Dominik Klein]

Ehlers, Kolja wrote:
> it does not need to be psychic to try to rerun apache again and see if the
> problem was fixed. But you are right if there is a problem which has to be
> manually fixed you can manually tell it that its fine to start it on that
> node again. 

You effectively _removed_ httpd from the machine. So the cluster is 
right in never trying to start it there again until you tell it to. How 
should it know you re-installed it.

> But is that certain that this error code will only come up if
> something needs to be manually fixed? 

Depends on what the RA does and the RA developer thought/knew when 
writing it. If you want to be sure: read and understand the RA. If you 
need to fix anything - please provide the patches.

> To do more reading on the Error Codes,
> also to look ahead to tomcat, do I need to debug those ocf start scripts or
> is there a manual somehwere?

To learn about OCF scripts, look at http://www.linux-ha.org/OCFResourceAgent

Regards
Dominik