Test Case 13 -Simulate Hba Failure - Avaya VSP 9000 Technical Configuration Manual

Avaya virtual services platform with coraid etherdrive srx-series storage appliances
Table of Contents

Advertisement

5.13 Test Case 13 –Simulate HBA failure
The objective of this test case is to test HBA recovery by removing both links from one HBA for 20
seconds and then plugging them back in to see if the PC recovers.
Test Steps
1)
Run Vdbench from all the PCs
again.
Confirm that Vdbench is running
correctly.
2)
Remove both links of an HBA for
20 seconds, either from the HBA
itself or from the VSP.
3)
Plug both HBAs back in.
Verify if the PC is stable and traffic
is transmitting again.
4)
Wait for Vdbench to finish reading
and writing to verify the results.
5)
Multiple failover times for this
scenario might need to be tested
to see what the threshold is for a
complete failover and the system
can recover.
Dec 2010
Expected Results
Vdbench should be reading and
writing on all PCs.
Vdbench should hang
Coraid states a failover of 15 seconds,
so if the links are out for 20 seconds,
we should lose traffic.
Because the RAID arrays are
mounted to the PCs, if the RAID is
down for too long, the PC acts like it
has lost a disk drive. This can cause
the PC to hang. Verify that within 20
seconds a PC can lose the connection
but recover.
Vdbench should finish with no errors
in the log file and all data should be
validated.
Try multiple dual link failure times to
verify if a certain time can be
established as the threshold where the
system won't recover.
VSP 9000 with Coraid EtherDrive SRX-Series
Storage Appliances TCG
avaya.com
Actual Results
Pass
Pass
Pass
Pass
Pass
29

Advertisement

Table of Contents
loading

Table of Contents