-
Notifications
You must be signed in to change notification settings - Fork 14
Pilot Options
Paul Nilsson edited this page Apr 25, 2022
·
1 revision
The table below contains a list of the various pilot launch options and their descriptions.
Example of pilot launch with a minimal set of options:
python pilot.py -q ANALY_CERN-PTEST --pilot-user=ATLAS
Option | Default value | Description |
---|---|---|
-a <value> | Pilot work directory | |
-d | False | Enable debug mode for logging messages |
-i <value> | PR | Version tag (optional: RC) |
-j <value> | ptest | Job prod_source_label, used to select job type from server |
-l <value> | 324000 | Pilot lifetime in seconds. queuedata.maxtime will be used if set |
-p <value> | 25443 | PanDA server port |
-q <value> | Required: Queue name (e.g., AGLT2_TEST-condor) | |
-r <value> | OBSOLETE: resource name (e.g., AGLT2_TEST). Can be safely removed from pilot wrapper | |
-s <value> | OBSOLETE: site name (e.g., AGLT2_TEST). Can be safely removed from pilot wrapper | |
-t | True | Disable proxy verification |
-u | True | Disable payload proxy verification |
-v <value> | 2 | Number of getjob requests |
-w <value> | generic | Pilot workflow |
-z | True | Disable server updates |
--allow-other-country <value> | PanDA server port | |
--allow-same-user <value> | Multi-jobs will only come from same taskID (and thus same user) | |
--cacert <value> | CA certificate to use with HTTPS calls to server, commonly X509 proxy | |
--capath <value> | CA certificates path | |
--cleanup <value> | True | Cleanup work directory after pilot has finished |
--country-group <value> | Country group option for getjob request | |
--es-executor-type <value> | generic | Executor type for event service (generic or raythena) |
--harvester-datadir <value> | Harvester data directory | |
--harvester-eventstatusdump <value> | Harvester event status dump json file containing processing status | |
--harvester-submit-mode <value> | PULL | Harvester submit mode (PUSH or PULL) |
--harvester-workdir <value> | Harvester work directory | |
--harvester-workerattributes <value> | Harvester worker attributes json file containing job status | |
--hpc-mode <value> | manytoone | HPC mode (manytoone, jumbojobs) |
--hpc-resource <value> | Name of the HPC resource (e.g. Summit) | |
--input-dir <value> | Input directory (pilot expects to find input files here if this is specified) | |
--job-type <value> | Job type (managed, user) | |
--no-pilot-log | False | Do not write the pilot log to file |
--output-dir <value> | Output directory (pilot expects to find output files here if this is specified) | |
--pilot-user <value> | generic | Required: Pilot user, e.g. name of experiment |
--queuedata-url <value> | Queue data server URL (default value stored in default.cfg) | |
--resource-type <value> | Resource type; MCORE, SCORE, SCORE_HIMEM or MCORE_HIMEM | |
--url <value> | PanDA server URL (default value stored in default.cfg) | |
--use-https <value> | True | Use HTTPS protocol for communications with server |
--working-group <value> | Working group option for getjob request |
- Introduction
- Pilot Architecture
- Pilot Workflows
- Event service
- Metadata
- Direct Access
- Signal Handling
- Error Codes
- Containers
- Special Algorithms
- Pilot Configuration
- Timing Measurements
- Copy Tools
- Pilot release procedure