vSAN ESA Virtual Machine Read and Write Latency Spikes During UNMAP or File Overwrite Workloads

Products

• VMware vSAN

Issue/Introduction

When running latency-sensitive or high-frequency write/UNMAP workloads (such as trading session persistence, database logging, or rapid temporary file overwrites) on a vSAN Express Storage Architecture (ESA) cluster with Deduplication & Compression enabled:

• Virtual machine disk read and write latencies experience severe transient spikes ranging from 400 ms to over 2,000 ms (2+ seconds).
• The performance degradation correlates directly with High-frequency, small-block, latency-sensitive write/rewrite workloads paired with active TRIM/UNMAP space reclamation

Environment

vSAN ESA

Version: ESXi 9.1.0.0 (Build 25370933)

Resolution

Workaround:
To stabilize virtual machine write latency and prevent guest I/O from inheriting multi-second UNMAP throttle waits, reduce the UNMAP throttle timeout parameters down to 1 millisecond across all hosts in the vSAN ESA cluster.

Run the following commands via SSH:

esxcfg-advcfg -s 1 /VSAN/zDOMUnmapTimeoutMsec
esxcfg-advcfg -s 1 /VSAN/zDOMUnmapMaxTimeoutMsec

Additional Information

Please contact Broadcom support for further investigation and assistance.