the metrics are very good for prometheus type supervision, but they are missing certain important metrics that we have on the admin page, namely: WOPI Host Views PID Document Memory Consumed Elapsed Time Idle Time Modified
Do you know how I could recover these values in order to extract them via prometheus
I need to log this information
THANKS
If you read the Prometheus output - the format is not well designed for exporting structured data; so we mis-use it a bit like this - if you look at the end you will see for each document:
doc_info{host=“localhost”,key=“https%3A%2F%2Flocalhost%3A9980%2Fwopi%2Ffiles%2Fopt%2Flibreoffice%2Fonline%2Ftest%2Fdata%2Fhello-world.odt”,filename=“hello-world.odt”,pid=“341946”} 1
doc_views{pid=“341946”} 1
doc_views_active{pid=“341946”} 1
doc_is_modified{pid=“341946”} 0
doc_memory_used_bytes{pid=“341946”} 14416
doc_cpu_used_seconds{pid=“341946”} 0.33
doc_open_time_seconds{pid=“341946”} 2
doc_idle_time_seconds{pid=“341946”} 2
doc_download_time_seconds{pid=“341946”} 0.02
doc_upload_time_seconds{pid=“341946”} 0
If you want to write a more powerful tool everything in the admin console comes from a websocket connection which you can make too - the protocol is not guarenteed to be stable, but pragmatically has mostly been so for quite a while; so you can connect, authenticate and scrape what you want.
If you want more fields there - its easy to add in the code, but each one adds weight to the prometheus data do:
$ git grep -10 doc_upload_time_seconds
Patches most welcome =)
OK thanks for your answer