Lifecycle of ACS Pod instances

This topic describes the special lifecycle states of ACS pod instances.

Terminal pod instances

When an ACS pod instance reaches a terminal state, ACS reclaims the underlying compute resources. By default, other resources created with the pod instance — such as elastic IP addresses (EIPs) — are also released.

ACS pod instances enter one of two terminal states: Failed or Succeeded.

Metadata retention limits

ACS retains metadata for terminal pod instances as follows:

Time since termination	Records retained
Within 30 minutes	All records
After 30 minutes	100 most recent records per region

References

For general pod lifecycle documentation, see Pod Lifecycle in the Kubernetes documentation.