What your code can use during a run
Run details#
Every run gives your code a fixed set of values. You can read them from
os.environ in notebooks and, where listed, in job code.
| Name | Meaning |
|---|---|
USERNAME |
User who started the run |
USERGROUPS |
The user's groups, comma-separated |
SITE_NAME |
Site (tenant) name |
REGION |
Cloud region |
JOB_IDENTIFIER |
Id of this run |
SESSION_ID |
Id of the session |
WORKLOAD |
ETL_JOB or NOTEBOOK |
PAGE_NUMBER, PAGE_SIZE |
Paging of a dataset run |
Do not rely on any other values or try to override these.
Job placeholders#
Text in a job can contain ${NAME} placeholders that the runtime fills in when
the job runs.
| Placeholder | Meaning |
|---|---|
${STRTTM} |
Run start time |
${ENDTM} |
Run end time |
${CLNTNM}, ${SITENM} |
Client and site name |
${PRJNM}, ${JOBNM} |
Project and job name |
${STS}, ${ERR} |
Run status and error text |
${RTE}, ${RGN} |
Runtime name and region |
${USR}, ${USRGRPS} |
User and groups |
The Standard runtime uses a smaller set; see the Standard chapter.
Imported models#
A run can be given registered ML models. Groundzero finds each model's latest stored version and makes it available to your code.