Another important component of operating a vSphere infrastructure is configuration management. When talking about configuration management, the concern is with ensuring configurations are not unknowingly changed or drift from their intended configurations. You want configurations to match their intended configuration and be consistent across the environment. For example, you want your hosts to be running on a certain build of vSphere and to be consistent with the other hosts within the same cluster.
vCenter Operations Enterprise versions include vCenter Configuration Manager, which provides the ability to monitor configuration virtual infrastructure configurations. vCenter Operations are discussed in further detail in Chapter 4, “Managing the Environment.”
Even with a product like vCenter Operations, you still need to implement the most important part of a solid configuration management strategy. Policies and procedures for documenting configuration are the foundation to maintaining an environment with consistent and desired configurations.
When thinking about the configurations, the goal is to maintain many items that might not seem obvious initially. The following is a list of some of the pieces in your virtual infrastructure that might have configurations—in terms of software or firmware—to track and ensure are desired and consistent. Keep in mind this list is brief and we could easily drill even deeper.
- vCenter Server configuration
- Cluster configuration
- High Availability configuration
- DRS configuration
- Update Manager configuration
- vSphere host configuration
- vSphere drivers and operating software
- HBA drivers and firmware
- NIC drivers and firmware
- IP network configuration
- Storage network configuration
- Storage firmware and software versions
- Virtual guest configuration
If a product like Operations Manager is not a fit, you may also use Host Profiles. This feature is included with the Enterprise Plus level of vCenter licensing and is of great assistance with managing the delivery of consistent configurations. Additionally, you may use Scheduled Tasks in vCenter to define a scheduled compliance check that will notify you daily of any configuration drift.
After you’ve configured your first host to the desired gold state, you can simply create a profile using this host as a reference host. Then you can take your baseline profile and apply it to other hosts or clusters. You will be prompted to enter dynamic information, such as network information during the application, but other configuration settings will be applied consistently to your hosts.
At any time, you can check the host’s compliance against the profile or receive notification via email when a drift in configuration occurs. When the time comes to make a change to your standard configuration, the process is just as easy. Simply update your reference host’s configuration and then update the profile and reapply the configuration to your other hosts.
Even if you do not have Enterprise Plus licensing, you should consider using Host Profiles during your setup as part of your 60-day evaluation licensing.
So far, this chapter has discussed ways to operationally maintain the environment through updates and alerts. Another important operational step is to perform regular health checks of your environment. This may consist of a physical inspection as well as checking configurations. You may also be ensuring your configured alarms are configured as expected and manually checking for issues just in case. You may also be looking for drifts in configuration based on your organization’s standardized configuration.
These are all important things to do and there are many community resources that can assist in these efforts. One such resource is a daily health check script developed by Alan Renouf called vCheck, detailed further in Appendix A.
This script creates a daily report that gives a great report of the environment, including items such as snapshots and new virtual machines that have been created. The setup process has been made easy with an install script, and a great demo video is included on the site for guidance in setting the script up.
Continuing the discussion of performing health checks, another reason to do a health check might be to get a new perspective on the current state of the environment. You might think to yourself, “Well, nothing has changed in this environment in the last three months.” Considering that perhaps nothing has changed in the environment, you also need to consider what has changed externally to your environment. This doesn’t strictly refer to the storage or networking attached to your vSphere hosts, although checking on these is equally important. Technology is often updated or at times has vulnerabilities due to security flaws in the product.
Bugs, workarounds, patches, and best practices are regularly released and updated. Many individuals barely have the time to perform their regular day-to-day duties, and this information can be difficult to find at times. This is where the aid of someone focused on vSphere technologies is of great advantage.
VMware’s Health Check Delivery
VMware offers a Health Check service that can greatly aid in this need. Any of the information that is used during this process is available to anyone and you could use scripts like the ones mentioned to verify much of the same information. The time to do so could be substantial and unless you have significant experience across many environments, there may be the risk that you are missing something. The health check delivery has many big advantages, such as the following:
- Consultants will add in their experiences recently as well as perform additional checks.
- Consultants will have at minimum a VCP.
- Quick collection of data for analysis will be performed by an expert.
The result of the engagement is a report and in-depth analysis of the environment with suggestions and remediation. The suggestions are based on best-practice configuration and known issues across a wide range of industries and environments.