Steps to Reproduce:

1. Create many providers
2. Monitor results during 72-96 hours
3. Check memory usage of controller pod / inventory container with grafana

During the monitoring period, no actions were taken:

• Add provider
• Change provider Network
• Run migration plans

Test environment:

• Software versions:
  • CNV_2.6.3-16
  • MTV_2.0.0-21
  • latest u/s (May 26, 2021)

• cloud10:
  • 5 x VMware 6.5 (with ~4100VMs)
  • 2 x VMware 6.7 (with ~300VMs)
  • 1 x plan migration succeed

• cloud38 Providers:
  • 1 x VMware 6.5 (with ~4100VMs)
  • 1 x VMware 6.7 (with ~300VMs)
  • 1 x plan migration succeed

Results:

• cloud10 - Over 84 hours, the memory increased by 280 MiB, i.e. 3.3 MiB/hour
• cloud38 - Over 95 hours, the memory increased by 110 MiB, i.e. 1.15 MiB/hour