$ ringserv panel deployments
Every application on its own row, with the buttons that matter beside it.
Which applications are up, whether they actually answer, what they last said, and whether their records are still whole. One command opens it, and it needs no setting up.
$ ringserv panel deployments
Every application on its own row, with the buttons that matter beside it.
| Start & Stop | run an application, or stop it — this one only, never everything on the machine by accident. |
| Reload | make your edited code live without stopping the server, and it tells you whether every part of the server took it. How that works. |
| Its log | the last thing it printed, updating as it runs — enough to see what just went wrong. |
| Try a service | call any action with any payload and read the answer, right there. No separate tool, and no browser security dance. |
This is the part we care most about, because a monitoring page that quietly makes things up is worse than no page at all.
Running does not mean working. A program can be alive and answering nothing — a start that failed halfway leaves exactly that. So the panel asks the application whether it is really serving, rather than assuming it from the fact that a process exists.
What it cannot know, it says it cannot know. If it cannot work out which address an application uses, it shows that as unknown instead of inventing a plausible number. An invented number is the kind of thing someone acts on at three in the morning.
It is a window, not an archive. The log you see is the recent tail, not everything since the beginning. Keeping history is a job for your system's own tools, and pretending otherwise would just lose it slowly.
If your application keeps the kind of record that must never change — sales, payments, decisions — you can ask it at any time whether that record is still intact:
$ ringserv journal verify 7 event(s), chain INTACTE
Each entry is sealed against the one before it, so if a single one were altered the seal would not match and you would be told which entry broke it. This is a check you can run every night, and the answer is a fact rather than a reassurance.
One honest word about backups: an untested backup is a belief, not a backup. When you rehearse restoring one, count the entries that came back — an empty record passes an "is it whole?" check perfectly well, because nothing empty is broken. The deployment guide has the routine we use.
$ ringserv ls myapp port 8210 running 2026-08-25 21:12 UTC comptoir port 8250 stopped 2026-08-25 19:40 UTC
The same truths, without a browser — useful over a remote connection or from a script that checks things while you sleep. Here too, running means the application was asked and answered.
The panel can start and stop programs, so it refuses to listen to anything except the computer it runs on. That is not a setting you can change, and it is deliberate: a page that controls your applications should not be one firewall mistake away from the internet.
To reach it from elsewhere, do it the way you would reach anything else — through a proxy you control, with your own login in front. The full guide.