rsyslog pushing "old" supervisor logs to syslog

[prsunny]

Description
While stopping and starting snmp service, it was observed that the same error message is repeatedly pushed to syslog.

Initial analysis suggest the logic here may have issues.

/etc/rsyslog.d/supervisor.conf => InputFileStateFile

Steps to reproduce the issue:

1. systemctl stop snmp, systemctl start snmp multiple times
2. tail -F /var/log/syslog | grep "not expected"
3. The following logging happens everytime snmp service start. Notice the syslog time vs supervisord time

Describe the results you received:

Apr 26 22:43:48.757317 str-a7060cx-acs-1 INFO supervisord 2018-04-26 21:07:20,342 INFO exited: snmpd (terminated by SIGHUP; not expected)
Apr 26 22:45:58.609038 str-a7060cx-acs-1 INFO supervisord 2018-04-26 21:07:20,342 INFO exited: snmpd (terminated by SIGHUP; not expected)
Apr 26 23:04:25.800888 str-a7060cx-acs-1 INFO supervisord 2018-04-26 21:07:20,342 INFO exited: snmpd (terminated by SIGHUP; not expected)

Describe the results you expected:
After deleting the above line from supervisor/supervisord.log, and repeat the same steps of stopping and starting snmp, the issue was not observed

Additional information you deem important (e.g. issue happens only occasionally):

**Output of `show version`:**

```
admin@str-a7060cx-acs-1:/$ show version 
SONiC Software Version: SONiC.20180330.16
```

[lguohan]

we can potentially classify the logs based on the docker, then the problem is that we do not have merged events logs

[lguohan]

another solution is the let supervisord to send the syslog directly to outside the docker, this will also solve the dual timestamp issues.

[lguohan]

I found the supervisord activity log also goes to stdout/stderr, therefore you can use docker logs to see them.

https://docs.docker.com/config/containers/logging/configure/#configure-the-delivery-mode-of-log-messages-from-container-to-log-driver