Is it possible to delay gcstartup in a multinode RAC environment, delay resource intensive scripts,

Looking for a way to either delay agent startup on some RAC nodes or delay the start of the grid control scripts that get kicked off either by initial agent startup or by the monitoring scripts.

We have experienced resource issues in the past where if one or two AIX RAC nodes running GPFS crash on a heavily used/busy RAC environment. Grid kicks off it's scripts to inquire about the environment such as CPU, memory, disk storage, etc. Each node has anywhere from 8 to 12 CPUS and we have around 6-8 terrabytes of storage depending on the RAC environment.

Due to the amount of storage it may take 20 minutes for grid to complete it's storage assessment. During this time we use a lot of CPU for this task. When we have two nodes crash we have both agents kicking off the same scripts. Sometimes GPFS is not all of the way up when these scripts are kicked off. This will sometimes create a RAC node to panic and reboot itself.