- The Cookbook for Setting Up a Serviceguard Package-less Cluster
- The Basics of a Failure
- The Basics of a Cluster
- The "Split-Brain" Syndrome
- Hardware and Software Considerations for Setting Up a Cluster
- Testing Critical Hardware before Setting Up a Cluster
- Setting Up a Serviceguard Package-less Cluster
- Constant Monitoring
- Chapter Review
- Test Your Knowledge
- Answers to Test Your Knowledge
- Chapter Review Questions
- Answers to Chapter Review Questions
Test Your Knowledge
1: It is important that all nodes in the cluster have the same IO tree for shared devices. If not, we have to get involved with recreating the Instance numbers assigned to devices via reconfiguring the /etc/ioconfig and /stand/ioconfig files. True or False?
2: Every HEARTBEAT_INTERVAL node is transmitting heartbeat packets. After two subsequent HEARTBEAT_INTERVALS where the heartbeat packet does not reach the cluster coordinator, a cluster reformation commences. True or False?
3: Which of the following events will trigger a cluster reformation? Select all the correct answers.
-
A node leaves the cluster.
-
An application is moved to an adoptive node.
-
An application fails on its current node.
-
A node joins the cluster.
-
A Primary LAN interface fails on the Cluster Coordinator node.
-
The cluster starts up automatically.
-
The cluster starts up manually.
4: A network interface designated as a STATIONARY_IP interface will not have its IP configuration moved to a Standby LAN interface in the event of a LAN card failure. True or False?
5: To create the cluster binary file, we use the cmapplyconf command. We need 100 percent attendance when the cluster binary file is first created. The first time we start the cluster, we don't need 100 percent node attendance as long as all nodes have the newly created cluster binary file. True or False?