This will make stageing files optional, since we are migrating to a different storage architecture, which is faster than our dedicated raw data server.
Therefore we do not need to stage input files. To solve this, an analysis' infiles will be read from their storage location. Where as parameter passed stagefiles will be mounted to the analysis rundir. This is done so we can still do e.g. --tdb /path/to/dbs/* to include multiple files.
An env var is used to toggle this behaviour, simply deploy without the STAGESHARE env variable if you run without a stage disk.
This will make stageing files optional, since we are migrating to a different storage architecture, which is faster than our dedicated raw data server.
Therefore we do not need to stage input files. To solve this, an analysis'
infileswill be read from their storage location. Where as parameter passedstagefileswill be mounted to the analysisrundir. This is done so we can still do e.g.--tdb /path/to/dbs/*to include multiple files.An env var is used to toggle this behaviour, simply deploy without the
STAGESHAREenv variable if you run without a stage disk.