Valentin: we do get reqs from people willing to figure out how much content is used

for certain reqs you don't need a count, just whether it has been used by anybody

valentin proposes a grep in the srv log (!) 

fab mentions the opensearch which is supposed to be there

also hugo agrees, opensearch would solve this problem without instrumenting the clients, which would be a technically questionable option

Valenting: sft interested in the performance of publishing. one idea is s3 buckets to be indexless. Val discussed with Enrico. Maybe testable in the medium term.

fab proposes to start having a look together at the code of the gateway. Val: let's sit together virtually tomorrow afternoon