Detect if NTRIP service is not sending data and restart it

[gunicsba]

Is your feature request related to a problem? Please describe.
Yesterday we had an issue with centipede, and today all PI-s have to be rebooted.

Describe the solution you'd like
It would be nice if we could detect the unavailability of NTRIP Service and auto restart it after 5-10 minutes.

[Stefal]

Hi gunicsba !
Yes, I'm thinking about this problem since yesterday. I see 3 differents solutions :

• Storing the caster ip address, and checking it with a deamon running a tool like nslookup, and restarting the ntrip service is the ip address change
• Connecting to the str2str log output and detect when it doesn't send any byte
• Fixing the problem directly inside str2str (Rtklib).

btw : You don't have to reboot your computer, you can just restart the ntrip service.

More background :

2023-10-01

• There was a major failure somewhere on the renater network or in the datacenter where the caster.centipede.fr server is located.
• A backup caster was started on another server, with a different ip address and the caster.centipede.fr dns entry was updated
• A big part of the running Gnss base didn't "updated" to the new ip address, and are unavailable.
• 2023-10-02 @ 07H10 the main server is back online, and the dns entry was updated to point to the old ip address.