Influxdb is restarting constantly since my last reboot

It looks like /usr/lib/influxdb/scripts/influxd-systemd-start.sh is trying to do a health check:

 while [ "$result" != "200" ]; do
   sleep 1
   result=$(curl -s -o /dev/null http://$HOST:$PORT/health -w %{http_code})
 done

this is failing. From the file date, the start wrapper was only created on 21 July, so it looks like the start check is new.

If I manually try I get:

root@monitor$ curl https://127.0.0.1:8088/health
curl: (35) OpenSSL SSL_connect: Connection reset by peer in connection to 127.0.0.1:8088 

It fails for several reasons.

1. Because I have configured TLS it needs to be https
2. Because I have not explicitly defined the Bind Port, because I am using the default, the script gets the wrong port.
3. because TLS is enabled, it needs the FQDN, not localhost or the cert validation check fails.
4. the perms were also wrong on the default startup script

To resolve it I edited the /lib/systemd/system/influxdb.service file and

1. change Type=forking to Type=simple
2. change ExecStart to : ExecStart=/usr/bin/influxd -config /etc/influxdb/influxdb.conf --pidfile /var/lib/influxdb/influxd.pid $INFLUXD_OPTS

This is a bug introduced in Influxdb v1.8.7. Github Issue.

There's a variety of ways of fixing this, your solution being one of the ways. In our case Influx took a bit longer to startup than the 10 second window the startup script allows, so I simply changed the line sleep 1 in the file /usr/lib/influxdb/scripts/influxd-systemd-start.sh to sleep 2 to give Influx more time to startup.