

Configuring the Matching Service

Configuring the matching service

This document covers the process for administrators to update a deployed matching tool with a new matching configuration file. It does not include a guide for modifying the configuration file itself. For that, see Modifying the Matching Service.

Given a new matching configuration file, it can be applied by putting it in the path /matching-tool/matcher/matcher_config.yaml, rebuilding and restarting the containers. From that point on, a user can simply upload and merge any file to trigger the matching process again with the new configuration, even a file with only one record.

Example: Given a new matching configuration file called new_matching.yaml:

cp new_matching.yaml matching-tool/matcher/matcher_config.yaml
cd matching-tool
./scripts/run rebuild
./scripts/run start

# At this point the user can upload a one-line file in the web browser to kick off matching.

If you would like to verify that the new matcher configuration is there, the matcher worker logs its config when it begins matching. Each of these lines is prefixed with "Matcher config", so you can filter to these lines using grep. For instance, if you start following the log from the end of the file before matching starts, and grep for lines with the prefix, you should see those lines scroll by when the matcher loads up, and no other lines.

docker logs matcher_worker --tail=5 --follow | grep "Matcher config"