Repeat, in case you didn't receive this the first time:
I reloaded the HACMP v4.3 MIB into our NetView host without trouble. Instead of
running mib2trap, I generated the trap definitions manually using the NetView
GUI Options->Event Configuration->Trap Customization: SNMP and then adding the
enterprise "HACMP" with oid 1.3.6.1.4.1.2.3.1.2.1.5. Next I added the following
traps under that enterprise with the specific IDs from the HACMP MIB as
follows:
clusterState specific 10
clusterSubstate specific 11
nodeState specific 12
networkState specific 13
addressState specific 14
clusterPrimary specific 15
applState specific 16
adapterSwap specific 17
Each of these traps has different integer values with the following definitions
(also taken from the HACMP MIB):
clusterState up =2 down = 4 unknown = 8
clusterSubstate unstable = 16 stable = 32 unknown = 8 error =
64
clusterPrimary 1=primary defined or 2=primary undefined
nodeState up = 2 down = 4 joining = 32 leaving = 64
networkState up = 2 down = 4 joining = 32 leaving = 64
addressState up = 2 down = 4 unknown = 8
This takes about 10 minutes to do this and once done is good for all HACMP
clusters.
The trap for clusterPrimary is now obsolete and is carried over from HACMP 3.1
where a primary node had to be defined regardless of cluster configuration.
Note that in case of an HACMP failover, the NetView control desk will receive an
alarm avalanche of 26 traps, 13 for one HACMP host going down and 13 for the
alternate host taking over.
Regards,
ken
kgarst@giantofmaryland.com
|