Calculation of vSAN capacity utilization alarm 'What if the most consumed host fails'
Article ID: 376614
Updated On:
Products
Issue/Introduction
Alarm of "Capacity Utilization Health – Disk Space check in the vSAN Health Service"
Environment
VMware vSAN 7.0.x
VMware vSAN 8.0.x
Cause
This is an informational alarm only for administrators to consider capacity planning of the cluster.
If the most consumed ESXi host in the cluster fails, there are two further outcomes.
Firstly, the resources on that ESXi host such as the disk group (cache and capacity disks) are no longer available.
Second, vSAN attempts to re-protect (rebuild) the components of their respective objects that were located on the failed host which would be running with reduced redundancy due to the failure.
This health check simulates both actions described above, therefore allowing for capacity planning.
Configuration of vSAN cluster
The vSAN cluster formed by 4 nodes, which contains 2 x 32 GB capacity disks. The capacity of the vsanDatastore is 256 GB.
From the UI, the capacity of the vSAN is 255.46 GB. There are 177.87 GB used. The live usage is 177.97/255.46 = 69.67%.
Resolution
When the most consumed host failed, esx-03 in this case.
'What if the most consumed host fails' predicts how the vSAN cluster will behave from the storage and component utilization perspective after most consumed host fails.
Click "TROUBLESHOOT" to inspect the detailed information in regards to the alarm,
vSAN datastore will lose esx-03's capacity which is approximately 64 GB.
The remaining capacity of the vSAN datastore is 191.6 GB
vSAN will rebuild the components which resided on esx-03, so the Used capacity still remains at approximately 177.93 GB.
The usage of vSAN datastore is 177.93/191.6 = 92.86%, which triggered the alarm "What if the most consumed host fails".
Note:
If this Skyline health check reports that the disk space utilization would reach 100% if the most consumed host in the cluster fails, then should this host fail re-protection would fail for some objects because there are not enough resources available in the cluster.
Additional Information
Feedback
