We’ve been running 3.5.5P (AIX 4.3.3) on a SAN for a few years now.
The directory ‘/hci’ is just a link to the ‘disc’ that contains ALL of the Cloverleaf files.
eg
[355] qdxi01p> cd /
[355] qdxi01p> ls -l hci
lrwxrwxrwx 1 root system 14 Jul 19 2002 hci -> /apps/qdxi01p/
When a failover is initiated (assuming that the ‘current’ server is still working) a script is issued (we wrote this) that stops all Cloverleaf sites. When the disc has been remounted on the alternate server a script is issued to start all Cloverleaf sites.
Our ‘startSite.sh’ script always perfoms a ‘crash recovery’ before a site is started, just in case.
I hope this helps.
Note that we are IN an HACMP environment, but we do not run the HACMP Cloverleaf software.
