Fred Gleason wrote on 6/13/21 6:01 AM:
> With the exception of (some) things related to JACK, it should never be
> necessary to restart the rivendell service after making configuration
> changes in rdadmin(1). If you find a situation where that is necessary,
> please report it as a bug!
>
> Cheers!
>
>
> |---------------------------------------------------------------------|
> | Frederick F. Gleason, Jr. | Chief Developer |
> | | Paravel Systems
I don't know if this qualifies for Fred's statement or not.
I had a standalone system (rd1) running the station. I built a second
computer (rd2) to use in the studio. I restored the rd2 database from a
backup copy of the primary machine (rd1) and ran it on the bench for a
while as a standalone. It was generating logs and chaining appropriately.
I took the new computer (rd2) to the station and installed it in the
studio. It did not run rdairplay or a log, it just basically sat there
on the network. I forgot to stop the log generation cron job, so that
script ran each night generating logs on this station (rd2).
I later returned to join the two computers to a single system. I mounted
/var/snd from rd1 and changed rd.conf to use the database on rd1. I
disabled mariadb on rd2. I went into rdadmin and set up rd2 in the
hosts section.
That night at midnight, the next day's log did not load, and we had dead
air until the next day when staff loaded and started the log.
In reviewing the operations log for rd1 for previous days, I discovered
that each night at midnight the CHAINTO cart ran on the log followed by
an RML command from *rd2* to load the next days log. When I connected
rd2 to the rd1 database the RML was no longer sent at midnight. I
verified all settings in rdadmin were correct. The next night, the same
thing. The CHAINTO cut played, but no log loaded. The operations logs
did not go back far enough to see when the RML started coming from rd2
instead of rd1. I assume it was when I installed it in the studio and
connected it to the network, however, it was still in standalone
configuration and should not have been sending anything to rd1.
Only after restarting the rivendell service, which locked up rdairplay
requiring a forced kill and restart, taking us off the air for a few
minutes, did things return to normal. At midnight, the CHAINTO cart
played and an RML from rd1 correctly loaded and started the next day's log.
So, this is the situation where I made configuration changes in rdadmin
and a restart of the rivendell service was necessary to get things back
on track.
Michael
_______________________________________________
Rivendell-dev mailing list
[email protected]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
_______________________________________________
Rivendell-dev mailing list
[email protected]
http://caspian.paravelsystems.com/mailman/listinfo/rivendell-dev
