Container (Docker) Gateway issue Cluster_info is not cleaned up and every container restart fills up the table

Products

• CA API Gateway

Issue/Introduction

We are trying to configuring container (docker) gateway version 10.0 using docker hub image of  api gateway.

After a few restarts the cluster_info table is filled with failed container id . 

Which are also showing up the policymanager dashboard.

How is this cleaned up ?

Environment

Release : 10.0

Component : API GATEWAY

Resolution

There is a process in the gateway which cleans up this table but the default is that the entry must be inactive for 90 days.

You can set the following property in the system.properties file to change this

com.l7tech.server.clusterStaleNodeCleanupTimeoutSeconds=86400s

which means  clean up the entry after 1 day inactivity

To set this in docker you have to add it to the yaml file 

https://techdocs.broadcom.com/us/en/ca-enterprise-software/layer7-api-management/api-gateway/congw-10-1/reference/gateway-system-properties.html

The other option is to clear the entries manual from the cluster_info table in the mysql database when needed.

Currently the clean task is run in background every hour and this interval is not able to changed.  So the com.l7tech.server.clusterStaleNodeCleanupTimeoutSeconds value less than 1 hour will see the node be removed after 1 hour has been reached.