Job Metrics
Use job metrics to improve performance.
Metrics are automatically available in the oci_datascience_jobrun namespace for any Data Science job run. You don't need to enable monitoring on OCI resources to get these metrics.
Data Science job run metrics include these dimensions:
- resourceId
-
The OCID of the job run.
- resourceDisplayName
-
The user-supplied display name of the job run.
- shape
-
The Compute instance shape of the job run.
| Metric Name | Display Name | Unit | Description | Dimensions |
|---|---|---|---|---|
|
|
CPU Utilization |
Percent |
Activity level from the CPU when a CPU shape is used. Expressed as a percentage of the busy time compared with the total time (busy and idle). |
|
|
|
GPU Utilization |
Percent |
Activity level from the GPU when a GPU shape is used. Expressed as a percentage of the busy time compared with the total time (busy and idle). |
|
|
|
Disk Utilization |
Percent |
Disk space in use. Expressed as a percentage of the used disk space compared with total disk space. |
|
|
|
Memory Utilization |
Percent |
Memory in use. Measured by pages. The percentage of RAM that's used, which is measured by the number of memory pages used relative to the total number of memory pages. |
|
|
|
Network Bytes In |
Byte |
Network receipt throughput. Expressed as bytes per second. |
|
|
|
Network Bytes Out |
Byte |
Network transmission throughput. Expressed as bytes transmitted per second. |
|
Viewing Job Run Metrics
You can view the default metric charts using a job run details page in the Data Science service.
For more information about the emitted metrics, see job run metrics.
If you don't see the metrics data for a job run that you expect, see missing metrics data.