In a configuration that uses Oracle ASM with Solaris Volume Manager mirrored logical volumes, If availability is lost for a service that depends on a SUNW.ScalDeviceGroup resource, the problem might be a failure of the SUNW.ScalDeviceGroup probe. This problem can be mitigated by increasing the value of the IOTimeout property of the SUNW.ScalDeviceGroup resource type.
# clresource set -p IOTimeout=timeout-value-in-seconds resource-name
The default of the IOTimeout property is 30 seconds. This value can be set much higher and should be determined for each system individually. Start by increasing the value to 240 seconds, then tune the property further as needed to achieve optimal results.