How to reduce volume of messages to Jitsi Videobridge log jvb.log

I have followed the quick-install instructions to install Jitsi Meet on a Debian cloud VM.

After about 6 days with very little activity on the server, the log file /var/log/jitsi/jvb.log has grown to 16 MB. Every 10 seconds it is generating messages like

2020-04-10 11:49:30.318 INFO: [17] Health.doRun#294: Performed a successful health check in 4ms. Sticky failure: false
2020-04-10 11:49:40.318 INFO: [17] Videobridge.createConference#326: create_conf, id=52c5f4df03ebb6a9 gid=null logging=false

Also, most log files in Debian (including jicofo.log) get routinely closed and renamed (e.g. with suffix .1, .2 … ), but that does not appear to happen with jvb.log .

Is there any way to reduce the volume of messages and possibly the size of the log files?

This is a minor concern - Overall I was really impressed with the quick-install, especially the built-in Lets Encrypt interface! Thanks to the developers.

There should be logrotate config file for videobridge:
/etc/logrotate.d/jitsi-videobridge
You can edit this file suitable for your expectations.

Thank you for the reply! There is such a file - but it does not appear to be having an effect. Here are the contents:

/var/log/jitsi/jvb.log {
  hourly
  size 100M
  missingok
  rotate 24
  compress
  delaycompress
  notifempty
  copytruncate
  su jvb jitsi
}

The manpage says that the “daily” parameter will cause the log files to rotate daily.

Also, is there a config file somewhere that would let me reduce the number of messages?

You can adjust the log levels in /etc/jitsi/videobridge/logging.properties

The logrotate file for jvb works!
Look excatly at the manpage on logrotate.conf and at the difference between the size and maxsize parameters!

Mine looks like (and rotates every day):

/var/log/jitsi/jvb.log {
  daily
  maxsize 100M
  missingok
  rotate 7
  compress
  delaycompress
  notifempty
  copytruncate
  su jvb jitsi
}

Thanks - that fixed it.

To reduce the amount of logging in that case (10seconds interval for succesful health check), you could also change the log level for that specific component.

I added this line to /etc/jitsi/jicofo/logging.properties

org.jitsi.jicofo.health.Health.level=WARNING

which results in not logging succesful health checks. it makes the jicofo.log readable again.

In my recent Debian installation I see HealthChecker events every 10 seconds. I tried removing them adding org.jitsi.jicofo.health.Health.level=WARNING to /etc/jitsi/jicofo/logging.properties and also to /etc/jitsi/videobrdige/logging.properties, but it has no effect in neither case. I must also say that those messages appear in jvb.log, not in jicofo.log

sed -i 's/^.level=.*/.level=WARNING/' /etc/jitsi/videobridge/logging.properties
systemctl restart jitsi-videobridge2.service

Can you paste the log message you’re seeing?

I do not want to disable all logging.

The logs are the usual ones for HealthChecker that are reported above. Logrotate works, but the presence of the HealthChecker messages makes it very difficult to spot other messages which is why I tried the org.jitsi.jicofo.health.Health.level=WARNING option, without success.

2020-09-23 19:33:38.108 INFO: [17] HealthChecker.run#169: Performed a successful health check in PT0.005804S. Sticky failure: false
2020-09-23 19:33:48.105 INFO: [17] Videobridge.createConference#258: create_conf, id=e13a77a94ae9bf77 gid=-1 logging=false
2020-09-23 19:33:48.110 INFO: [17] HealthChecker.run#169: Performed a successful health check in PT0.007622S. Sticky failure: false
2020-09-23 19:33:58.102 INFO: [17] Videobridge.createConference#258: create_conf, id=fa265d779c38abe4 gid=-1 logging=false
2020-09-23 19:33:58.107 INFO: [17] HealthChecker.run#169: Performed a successful health check in PT0.004909S. Sticky failure: false

Unfortunately that class is in another package and has its own logger, so it’s actually org.jitsi.health.HealthChecker.