6- Testing for failure senario

Owned by Copy Page Tree
2018/07/18
1 min read


Table of Contents



Prove of concept

MCCS is a high availability solution that ensures the service keep online of the system when meet the downtime events whether caused by failures within the system, network, application and storage, or planned downtime such as hardware or software update.
Therefore, before online an application service after all the installation and configuration are completed, few steps shown below should be done to check if MCCS meets the objectives.

Types of Fault

Test List

Test Result

Power Off

Power Off

Failover to the standby server.

System OS Shutdown

System OS Shutdown

Failover to the standby server.

Network Card Failure
Network Cable Failure

Remove network cable

Re-check the connectivity to active server defined counts or failover to standby server(Possible to specify the counts).

Disk Failure
Physical Disk Failure
I/O Controller Failure

Remove Disk

Failover to standby server.

Database Service Failure

Stop Database

Re-start failed service on local(Possible to specify the counts) or failover to standby server.

Operation Process Failure

Stop Operating Process

Re-start failed process on local(Possible to specify the counts) or failover to standby server.

Test list above should be done in condition where resource group is created, and resource dependency is configured.
Please refer to "MCCS Manual" for more details about how to create group and resources.
In addition, please refer to"Chapter8. Administrating after failure" of MCCS Manual for more details about various faults that might occur and how to solve those problems.