Nginx crashes after disabling the source IP persistence feature in LB with ongoing traffic

Products

• VMware NSX

Issue/Introduction

• One Nginx process crashes, and the high CPU usage may be observed on the remaining Nginx processes of the same LB.
• The persistence feature is enabled in L4 virtual server of LB.
• This issue is specifically triggered by disabling the persistence feature on the L4 virtual server of LB. If the feature is enabled, and never disabled, the issue would not be triggered.

Environment

• VMware NSX-T Data Center
• VMware NSX

Cause

The issue happens due to a race condition during persistence profile reconfiguration.

Resolution

This issue is resolved in VMware NSX-T 3.2.4.3 and VMware NSX 4.2.2.2, 4.2.3.1, 4.2.4, 9.0.1.0 available at Broadcom downloads.

If you are having difficulty finding and downloading software, please review the Download Broadcom products and software KB.

 

Workaround

• Disable source IP persistence -  if the feature is disabled with ongoing traffic, please detach & re-attach the LB or enable/disable maintenance mode on LB, so the dirty data could be cleared.
• Failover the edge node manually

Additional Information

トラフィックが継続している状態でロードバランサーの送信元 IP パーシステンス機能を無効にした後、 Nginx がクラッシュする