Edge VM memory utilization is increasing after upgrade to 3.0.3/3.1.X

Products

• VMware NSX

Issue/Introduction

vROPs and other monitoring devices using the following MP API in custom scripts may show high memory utilisation on edges after upgrade to 3.0.x/3.1.x (until 3.1.2).

"GET api/v1/transport-nodes/< Edge-TN-UUID>/status"

Symptoms:

• Only monitoring devices (VROPs, etc) using the above MP API report high memory utilisation on edges.
• Edge logs and CLI output do NOT show any sign of high memory utilisation.
• No DP impact, but monitoring devices send alerts of high memory utilisation.
• Memory usage stats can be accessed var/log/vmware/top-mem.log.

Environment

VMware NSX Datacenter.

Cause

The mem_used, mem_total and mem_cache fields of the API were actually generated by Agg Service, and this part of the code was changed in 3.0.x and 3.1.x (until 3.1.2). A new code was added to collect Edge DP memory dump stats, which was overwriting the values of mem_used, mem_total and mem_cache fields correctly generated by AggService earlier. The new code calculated the parameters incorrectly.

Resolution

Fixed in NSX release 3.1.3 and later.