Fabric Manager fails to start. What do you read first?
Why A is the answer
Its service status and log
The log usually states the reason.
In detail
- The log usually states the reason.
- Read before you change anything.
Use case
An illustrative scenario, not a real incident.
Fabric Manager fails, and its log states the reason.
See it in different ways
flowchart TD A["NVSwitch job fails"] --> B["Is the service running?"] B --> C["Does its version match the driver?"] C --> D["Read the Fabric Manager log"]
sequenceDiagram participant O as Operator participant S as Service status participant V as Driver version participant L as Log O->>S: Is it running? O->>V: Does it match the driver? O->>L: Reads the reason
flowchart LR S0["Check status"] --> S1["Read log"] --> S2["Fix the cause"]
Key point: Read status and log first.
Why the other options are wrong
- B. The Slurm user list
- A user list is unrelated.
- C. The NGC catalog
- A catalog is unrelated.
- D. The Slurm controller log
- That log is for Slurm.