Monitor the Availability of the STANDBY Cluster
The STANDBY cluster responds to Diameter Device-Watchdog-Request (DWR) messages to let the network application know it is running so it can be ready to take over processing if a failover occurs. You can monitor the sending and receiving of DW messages by monitoring SNMP notifications and by running the print_blade_stats.py script.
About this task
This procedure describes how to verify the receiving and sending of DWR messages, signaling the standby cluster is available.
Note: If the redundant link between the HA cluster pair
is not available and communication between the two MATRIXX Engine sites is broken, for example, when the standby engine is stopped because it
cannot keep up with transaction replay operations from the active cluster (the
configured maximum number of retry attempts has been reached), an SNMP
sysClusterPeerDisconnected trap is triggered. Network operators should monitor the
system for this notification. When the connection is restored, a
sysClusterPeerConnected trap is triggered.
Procedure
Open a terminal and type
the following command to display the DW statistics on the standby cluster.
Replace
eID with the ID of the secondary engine.
print_blade_stats.py -e
eID -D
For example:
print_blade_stats.py -e 2 -D
Results
0 or are not equal to each other, problems might exist
on the standby cluster.
topology: Parsing config file: /opt/mtx/conf/mtx_config.xml
Diameter PDU Stats
------------------
Response Time in micros | Total Total
Current | Total | Packets Packets
AppId Cmd Description Avg Max | Avg Max | Read Sent
=======================================================================================================
0 257 common:CE 0 0 0 581 35 22560
0 280 common:DW 0 0 0 0 301028 301028
0 282 common:DP 0 0 0 0 0 0
1 265 nasreq:AA 0 0 0 0 0 0
3 271 accounting:AC 0 0 0 0 0 0
4 258 credit-control:RA 0 0 0 0 13334 0
4 272 credit-control:CC 0 0 6961 126512 21313 21313
368618 430 private:mdc 0 0 0 0 0 0