Before you start
You need the relevant system-reading capability. Starting, stopping, restarting or creating a service additionally requires the corresponding administration permission and authority for its dependencies.
What to do
- Open Processes & services and inspect the service list. Search for the exact service name or description, then check its reported state.
- Use Running processes to inspect current process resource use. A busy process is not by itself evidence that it is faulty or safe to stop.
- Use Terminate only for a process whose identity, owner, dependencies and intended shutdown you have verified. Review the confirmation and enter its exact process identifier; a matching name alone is insufficient.
- For an authorised service action, check the exact service and affected workloads. Use Start, or review the Stop or Restart confirmation and enter the service name as requested.
- When Create service is available and a new workload is intended, review its service name, existing executable/start command, working directory and Linux user. Do not assume this creates the application's files or dependencies.
- Read the final operation output, refresh the service state and inspect the affected application through its ordinary interface before considering the change successful.
What happens next
You can distinguish recorded service state, live process information and an administration action's result. A completed service action does not by itself prove that every dependent application works.
If something goes wrong
Check the exact unit, command, directory, executable, permissions and relevant logs when a service fails. Do not stop another workload because of its age, name or resource use alone. An uncertain action requires inspecting its operation and current state before retrying. Contact the infrastructure administrator with the service and time, keeping credentials and private logs out of public messages.
