jkellerer
jkellerer
Hi @allexzander, are you ok with the updates to your review or do you expect additional changes?
Hi @kesselb, I see your point. The reason for the config value is primarily to have a starting point for the auto-sensing that doesn't hurt performance too much (since both...
Hi @kesselb, @allexzander, Have updated the implementation as follows: * On client restart `_min` and `_max` chunk size applies (always), until the first failure. * On the first failure, the...
The latest commit addresses `_initial` chunk size to be persisted only when chunk size is dynamic (and supported). However I guess this value is better kept as runtime value in...
As far as I understand this has to be fixed in upstream (not to have conflicts in user IDs).
Not that I would be aware of and in fact the `node_exporter` pkg has a config option for the service user. However without an assigned user in [UIDs](https://github.com/opnsense/ports/blob/master/UIDs), this is...
Yes that's what I'm thinking, unless there is an other option to create a service user.
Fully agree, I also thought about creating a user dynamically but this would be the only plugin doing that and it wouldn't be very clean. For the moment I'd just...
Ok. I'll check what needs to be changed and prepare a PR.
Just found this entry: prometheus:*:478:478::0:0:Prometheus Daemon:/var/tmp/prometheus:/usr/sbin/nologin Looks like this case is already fixed partially with the move to FreeBSD. Seems to originate from the Prometheus daemon that collects input from...