Setting Up an HP-UX CSE Serviceguard Cluster
- 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
Chapter Syllabus
-
25.1 The Cookbook for Setting Up a Serviceguard Package-less Cluster
-
25.2 The Basics of a Failure
-
25.3 The Basics of a Cluster
-
25.4 The "Split-Brain" Syndrome
-
25.5 Hardware and Software Considerations for Setting Up a Cluster
-
25.6 Testing Critical Hardware before Setting Up a Cluster
-
25.7 Setting Up a Serviceguard Package-less Cluster
-
25.8 Constant Monitoring
This is where we get down to the nitty-gritty of setting up a Serviceguard cluster on HP-UX. We mention concepts such as "mirroring your root disk." The mechanics of performing such tasks are not covered in this module because they are covered elsewhere. Specific tasks that are relevant for setting up a Serviceguard high availability cluster are covered and, where appropriate, screenshots and example commands are used to help illustrate key points.
25.1 The Cookbook for Setting Up a Serviceguard Package-less Cluster
Before we can start to use this cookbook, we need to understand what Serviceguard is and what it is trying to achieve. I suggest that you not jump straight to the cookbook (see Table 25-1) because having an understanding of the concepts and limitations of Serviceguard can influence your decisions on how to construct your cluster. Each bullet point in the cookbook should be studied, understood, and implemented carefully. So here it is.
Table 25-1. Cookbook for Setting Up a Serviceguard Package-less Cluster
Cookbook for Setting Up a Serviceguard Package-less Cluster: |
---|
1. Understand the hardware and software implications of setting up a cluster. |
2. Set up NTP between all cluster members. |
3. Ensure that any shared LVM volume groups are not activated at boot time. |
4. Install Serviceguard and any related Serviceguard patches. |
5. Install a Quorum Server (optional in a basic cluster). |
6. Enable remote access to all nodes in the cluster. |
7. Create a default ASCII cluster configuration file (cmquerycl). |
8. Update the ASCII cluster configuration file. |
9. Check the updated ASCII cluster configuration file (cmcheckconf). |
10. Compile and distribute the binary cluster configuration file (cmapplyconf). |
11. Back up LVM structures of any cluster lock volume groups (vgcfgbackup). |
12. Start cluster services (cmruncl). |
13. Test cluster functionality. |
Before we get started, we need to begin by talking about the unthinkablea failure. What constitutes a failure? Different types of failure will prompt different responses from Serviceguard. This is where we start our discussion.