Integration with CRIC

The Computing Resource Information System (CRIC) is a framework providing a centralized and flexible way to describe resources and their usage.

It is possible to integrate PanDA and CRIC so that administrators register various resources in CRIC by using WebUI and PanDA fetches information from CRIC, avoiding manual registration in the database. One of the scripts launched by PanDA daemon, configurator, periodically retrieves information from CRIC and populates database tables. To enable PanDA daemon and configurator, you need in panda_server.cfg

[daemon]
# whether to run daemons for PanDA
enable = True

config = {
    ...
    "configurator": {
        "enable": true, "module": "configurator", "period": 200, "sync": true},

and set the following parameters:

Name	Description	Example
CRIC_URL_SCHEDCONFIG	URL to get schedconfig json	https://datalake-cric.cern.ch/api/atlas/pandaqueue/query/?json
CRIC_URL_SITES	URL to get site json	https://datalake-cric.cern.ch/api/atlas/site/query/?json
CRIC_URL_DDMENDPOINTS	URL to get storage json	https://datalake-cric.cern.ch/api/atlas/ddmendpoint/query/?json
CRIC_URL_DDMBLACKLIST	URL to get blacklist json	https://datalake-cric.cern.ch/api/atlas/ddmendpointstatus/query/?json&activity=write_wan&fstate=OFF
CRIC_URL_CM	URL to get site matrix json	https://atlas-cric.cern.ch/api/core/sitematrix/query/?json&json_pretty=0

If you don’t have a real CRIC instance or have a partially functional instance, you can use local json files to describe resources. In this case the above parameters take the form of: file://<path_to_file> instead of https://<url>. It is possible to use local json files only for some parameters while others are taken from the real CRIC instance.