Linux clients (e.g., RHEL) running on Element storage clusters (SolidFire or NetApp HCI)
While newer versions of ESX can run UNMAP automatically, linux clients are typically not set up to run TRIM automatically to reclaim block space.
This causes inaccurate capacity reporting and triggers alerts such as:
"blockClusterFull - Due to high capacity consumption Helix data protection will not recover if a node fails. Creating new Volumes or Snapshots is not permitted until additional capacity is available. Add additional capacity or free up capacity immediately.".
In this scenario, TRIM must be manually to reclaim the space and clear the alert.