In some cases, a failed readiness probe can only be remedied by a restart, but after healthy restart we want downstream processes to get started
Proposed Change:
Add process_healthy_restart_allowed as a dependency condition, then the process will stay in pending until the process it depends on is healthy (even if restarts are needed) or it will be skipped if the dependent process’s max_restarts are met
Who Benefits From The Change(s)?
Those using crash-only designs
Alternative Approaches
Put the process restart into the process script itself, rather than using process-compose’s orchestration
Feature Request
Use Case:
In some cases, a failed readiness probe can only be remedied by a restart, but after healthy restart we want downstream processes to get started
Proposed Change:
Add
process_healthy_restart_allowed
as a dependency condition, then the process will stay inpending
until the process it depends on is healthy (even if restarts are needed) or it will be skipped if the dependent process’smax_restarts
are metWho Benefits From The Change(s)?
Those using crash-only designs
Alternative Approaches
Put the process restart into the process script itself, rather than using process-compose’s orchestration