Since my first submission to Harbour (and that was 9 years, 11 months ago!) I’ve missed email notifications for app reviews. Going app by app, one page at a time to just see if I got any new review is frustrating, both for the developers, who might miss some good or bad feedback, and for the users who might feel as though they’re being ignored.
I’ve raised my concerns at the community meeting as well, where Jolla acknowledged that the “critique” was valid, but said that there are other priorities at the moment. Which I fully understand.
So, long story short: I wrote a Harbour exporter for Prometheus, I built a Grafana dashboard and setup an alert to get notified about new reviews and got historical data in return (how many people install them and then uninstall them soon after? Do users increase with a release? etc…).
If you want to setup this, the stack is available at the main project repo. The default setup scrapes Harbour every hour; this should be ok for Jolla, I hope.
18 Likes
WOW! How amazing job, you followed every “best possible way to operate” that person can think of…. Hats off <3 You raised your concerns of the problem → took it to the community meeting → got confirmation Jolla can’t do anything currently with available resources → you made your own frigging solution! And shared it with everyone. Daam, didn’t know one could be fan of dev, but here I am 
6 Likes
Hi,
i have the example from github now running. However i have no data. Not sure if the credentials are not ok, of if there are other scraping problems.
How can i check the logs for scraping?
When does the initial scraping happen, or do i have to run the container for at least (one or three hours) depending on the scraping interval?
Will the data be persisted somewhere and what about restarting the PC? Will the scraped data survive a restart or is persisted volumne required, since i have no server running 24/7?
Cheers,
Andreas
docker logs harbour-exporter.
When does the initial scraping happen, or do i have to run the container for at least (one or three hours) depending on the scraping interval?
It depends on scrape_interval, defaults to 1h. On a fresh start, Prometheus, by design, will not scrape it before scrape_interval.
Will the data be persisted somewhere and what about restarting the PC? Will the scraped data survive a restart or is persisted volumne required, since i have no server running 24/7?
Yes, in the example config, data is kept the in the prometheus volume.
However, if you have “holes” in your data, you may want to modify the Grafana queries because the default ones look for a maximum of 2 hours back.
1 Like
