vSAN SSD Congestion on vSAN Cluster during hosts upgrading

Products

• VMware vSphere ESXi

Issue/Introduction

• There are multi SSD Congestion issues observed from LSOM:Cache disk on vSAN Cluster during hosts upgrading.
• And noticed that ssdCongestion is continuous high.
• At the same time, we see high Device IOPS and the start of the SSD Congestion.
• From DOM:Cluster Resync ,noticed that resync IOPS indicates the resync is initiated and the outstanding IOs 

 

Environment

VMware vSAN 8.x

Resolution

This is an expected behaviour when there is a high load on the cached disk 

The Performance will improve, once all the outstanding IOs are cleared. 

Best practice recommendation:

1.Avoid placing multiple host on maintenance mode.

2.Place one host in maintenance mode and confirm the vSAN data resync is completed before placing the next host on maintenance mode.

3.Avoid remove and creation of disk group during the high load time or during production time as this will lead to performance issue.