vSAN 7.0 U1 Unveils Exciting New Features (Part 2)

My Journey from Infrastructure Admin to Cloud Architect: Embracing vSAN Capacity Reserve

As an infrastructure admin, I’ve always been focused on ensuring the smooth operation of our virtualized environment. However, with the recent upgrade to vSphere 7.0U1, I noticed some small but significant changes that have had a profound impact on my approach to managing our vSAN cluster. One of these changes is the introduction of Capacity Reserve, a feature that has helped me control vSAN capacity and prevent the datastore from becoming full. In this blog post, I’ll share my journey from infrastructure admin to cloud architect and how Capacity Reserve has helped me achieve this transition.

Background

———-

In our 4-node vSAN cluster, each host has approximately 17.5TB of capacity, and the total datastore capacity is around 69.86TB. When I upgraded to vSphere 7.0U1, I noticed that the Cluster -> Configure -> vSAN Services menu now includes an option for Enable Capacity Reserve. This feature was not available in previous versions of vSphere, and I was eager to explore its capabilities.

Pre-7.0 Era Challenges

————————

Before vSAN Capacity Reserve, our cluster’s datastore capacity was consistently reaching 70% utilization, with around 49.47TB of used capacity out of a total of 69.86TB. Although the datastore still appeared green and functional in the vCenter UI, I knew that we were approaching the limit of our available space. The only indication of potential issues was the vSAN Skyline Health check warning, “What if the most consumed host fails.” This alert suggested that if one of our hosts failed, we might not have enough space for rebuilds and cluster operations.

Managing Capacity with Fixed Slack Space

——————————————

To ensure that our cluster had sufficient capacity for rebuilds and other operations, we had to use a fixed 25-30% of slack (free) space regardless of the cluster size. This approach guaranteed enough space for rebuilds but often resulted in wasted capacity, especially in smaller clusters. In addition, not everyone was an expert in monitoring the cluster size, making it difficult to avoid creating automation tasks that provisioned hundreds of VMs over the weekend, potentially filling up the datastore.

Introducing Capacity Reserve

——————————

vSAN Capacity Reserve is a feature that helps control vSAN capacity and prevent the datastore from becoming full. This feature is specific to your cluster size and will protect the cluster from provisioning new VMs, ensuring that there is always enough space for rebuilds and other operations. When I enabled Capacity Reserve in our 4-node vSAN cluster, my VMs began using only 9.19TB of the datastore capacity, a significant reduction from the previous 49.47TB.

Capacity Reserve Calculations

——————————

The Capacity Reserve feature calculates the exact values for each host in the cluster based on the total capacity and the desired reserve percentage. In my case, the system reserved 16.54TB for host rebuilds (23.68% of the total capacity) and 1.78TB for operations (2.54%). When I created more VMs and vSAN tasks, my capacity utilization reached 45.45TB (almost the same as in the previous scenario). The Operations reserve increased to 7.08TB (10%), as more cluster resources were needed for operations.

Monitoring Capacity Utilization

——————————-

To monitor capacity utilization, I can view the Capacity Overview of the vSAN datastore in vCenter. This view shows me the total capacity, used capacity, and reserved capacity for each host in the cluster. I can also see that the system has reserved 16.54TB for host rebuilds (23.68%) and 1.78TB for operations (2.54%). The Operations reserve increased to 7.08TB (10%) as more cluster resources were needed for operations.

Alerts and Color-Coding

————————

When I create more VMs and vSAN tasks, my capacity utilization reaches 45.45TB, and the system triggers alerts to indicate that I am reaching my space limit. The Capacity Overview view now shows a red background to indicate that the cluster is nearing its capacity limit. Additionally, the system triggers alerts when the reserved capacity for host rebuilds or operations falls below a certain threshold.

Conclusion

———-

vSAN Capacity Reserve has been a game-changer for me as an infrastructure admin. This feature has allowed me to transition from managing capacity to designing and architecting our virtualized environment. With Capacity Reserve, I can now ensure that our cluster always has enough space for rebuilds and other operations, preventing the datastore from becoming full. By embracing this feature, I’ve been able to proactively manage our capacity utilization and avoid potential issues.

I hope that by sharing my journey from infrastructure admin to cloud architect, I can help others make a similar transition and take advantage of the many benefits that vSAN Capacity Reserve has to offer.