Added changes needed for the Docker in Docker execution #32
Conversation
| project: dev | ||
| campaign: quick | ||
| campaign: "G3W06" |
There was a problem hiding this comment.
I'm not sure what G3W06 means. Guessing it would be a Gen3 run using weekly stack w_2021_06 (which I don't think is the case in the rest of the yaml). Since example yaml, it doesn't have to be real, but I would try to not make it confusing.
| submitPath: ${PWD}/submit/{outCollection} | ||
| container_obs_panda_edge_node_dir: "/opt/lsst/software/stack/stack/current/Linux64/ctrl_bps/21.0.0-6-g55737e1+e68d6e10c0/python/lsst/ctrl/bps/wms/panda/edgenode" | ||
| container_obs_panda_edge_node_dir: /opt/lsst/software/stack/stack/miniconda3-py38_4.9.2-0.6.0/Linux64/ctrl_bps/21.0.0-18-gf2cd492+6c749b2ca5/python/lsst/ctrl/bps/wms/panda/edgenode | ||
| container_CTRL_MPEXEC_DIR: /opt/lsst/software/stack/stack/miniconda3-py38_4.9.2-0.6.0/Linux64/ctrl_mpexec/21.0.0-30-g82f2559+c64cb64c6b/ |
There was a problem hiding this comment.
Note: When the lazy command line changes are made in bps, the need for this should go away.
| butlerConfig: s3://butler-us-central1-panda-dev/hsc/butler.yaml | ||
| inCollection: HSC/defaults | ||
| output: "RC2/w_2021_13/FULLRC2" | ||
| outCollection: "shared/pipecheck/{timestamp}" |
There was a problem hiding this comment.
Should the example be for a non-production run and let the production folks change it correctly for production runs? If so the outCollection needs to follow the pattern in https://dmtn-167.lsst.io/ (pattern: u//xyz)
There was a problem hiding this comment.
Also I'm not sure what pipetask does if the run collection (outCollection) isn't an "extension" of the chained output collection (output). In the main bps example yaml file (the run collection is just the output collection with a run timestamp):
output: "u/${USER}/pipelines_check"
outCollection: "{output}/{timestamp}"
| wmsServiceClass: lsst.ctrl.bps.wms.panda.panda_service.PanDAService | ||
| clusterAlgorithm: lsst.ctrl.bps.quantum_clustering_funcs.single_quantum_clustering | ||
| createQuantumGraph: '${CTRL_MPEXEC_DIR}/bin/pipetask qgraph -d "{dataQuery}" -b {butlerConfig} -i {inCollection} -p {pipelineYaml} -q {qgraphFile}' | ||
| runQuantumCommand: 'pipetask run -b {butlerConfig} --input {inCollection} --output-run {outCollection} --extend-run --skip-init-writes --clobber-partial-outputs --qgraph {bucket}/{payload_folder}/{uniqProcName}/${{IN/L}}' | ||
| createQuantumGraph: '${CTRL_MPEXEC_DIR}/bin/pipetask qgraph -d "{dataQuery}" -b {butlerConfig} -i {inCollection} -p {pipelineYaml} -q {qgraphFile} --qgraph-dot {qgraphFile}.dot --instrument lsst.obs.subaru.HyperSuprimeCam' |
There was a problem hiding this comment.
I am surprised to see instrument on the command line. That's normally defined in the pipeline yaml. Not that it is wrong, but if the bps yaml doesn't need it, then it's one less place to change it.
| outCollection: "shared/pipecheck/{timestamp}" | ||
| dataQuery: "skymap='discrete/ci_hsc' AND tract=0 AND patch=69" | ||
| #Very small (~150 jobs) workflow | ||
| dataQuery: "tract = 9615 and patch=30 and detector IN (10..20) and instrument='HSC' and skymap='hsc_rings_v1' and band in ('r', 'i')" |
There was a problem hiding this comment.
Is instrument needed here since the given inCollection limits it to that instrument?
…polish required by the review
No description provided.