BigJob (>0.4.65) provides a command-line interface (pilot-cli), which is installed with the BigJob Package.
To start a pilot:
$ pilot-cli --coordination redis://localhost:6379 --submit_pilot fork://localhost
Started Pilot: redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost
List all pilots:
$ pilot-cli --list_pilots
Pilot Compute State
-----------------------------------------------------------------------------------------------------
redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost Running
redis://localhost:6379/bigjob:bj-a174e4a2-b18c-11e1-bc81-109addae22a3:localhost Running
Submit a simple command as compute unit to a pilot:
pilot-cli --submit_cu redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost /bin/date
submit CU to redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost
Started ComputeUnit: redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost:jobs:sj-dbb62be0-b190-11e1-b6e0-109addae22a3
Check the state of all compute units of a pilot:
$ pilot-cli --list_cus redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost
Pilot Compute: redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost
State: Running
# Compute Unit State
-------------------------------------------------------------------
1 sj-dbb62be0-b190-11e1-b6e0-109addae22a3 Done
Terminate Pilot:
pilot-cli --cancel_pilot redis://localhost:6379/bigjob:bj-2ccb20f8-b18c-11e1-b5e4-109addae22a3:localhost
Help:
$ pilot-cli --help
usage: pilot-cli [-h] [--coordination COORDINATION] [--clean] [--version]
[--number_cores NUMBER_CORES]
[--cores_per_node CORES_PER_NODE]
[--submit_pilot RESOURCE_URL] [--cancel_pilot PILOT_URL]
[--list_pilots] [--wait_cus] [--list_cus PILOT_URL]
[--submit_cu PILOT_URL [COMMAND ARGS ...]]
[--run_cu PILOT_URL [COMMAND ARGS ...]]
[--get_cu_state CU_URL] [--cancel_cu CU_URL]
BigJob Command Line Utility
optional arguments:
-h, --help show this help message and exit
--coordination COORDINATION, -c COORDINATION
--clean
--version
Manage pilots:
--number_cores NUMBER_CORES
--cores_per_node CORES_PER_NODE
--submit_pilot RESOURCE_URL
submit a pilot to specified resource, e.g.
fork://localhost
--cancel_pilot PILOT_URL
Cancel pilot
--list_pilots list all pilots
--wait_cus wait for termination of all CUs
--list_cus PILOT_URL
Manage compute units:
--submit_cu PILOT_URL [COMMAND ARGS ...]
submit CU to pilot
--run_cu PILOT_URL [COMMAND ARGS ...]
submit CU to pilot and wait for completion
--get_cu_state CU_URL
--cancel_cu CU_URL