snmp_profiles: add SwOs and Edgeswitch

[FFHener]

This PR brings monitoring for Mikrotik SwitchOS and Ubiquitys Edgeswitches. In addition it fixes the monitoring of proprietary devices for our gateways.

• Temperatur for SwOs
• Uptime
• Port Status
• Add external SNMP monitoring to gateway devices (not rolled out yet as i didnt wanted to have unnecessary downtime at gateways)
• Fix Uptime Scale for Grafana Grafana: Add GUI for Switch Monitoring meta#29
• Better Display Port Status in Grafana Grafana: Add GUI for Switch Monitoring meta#29
• Better Display Port Speed in Gradana Grafana: Add GUI for Switch Monitoring meta#29

[FFHener]

As the grafanapart is independent from this part i moved this part into a new meta issue. We could discuss as well if we should add snmp-monitoring for all switches treewide.

[FFHener]

Some SwOS-Switches dont provide temperature data (in the same way?) resulting in this error in logread:
daemon.err collectd[9583]: snmp plugin: OID "iso.3.6.1.4.1.14988.1.1.3.11.0" is undefined (type ASN_NULL)
I personally can live with that right now. When we collected more experience with the monitoring of the different models of switches we could then move over to more individualized snmp-profiles, maybe in combination with #751 (defining switches as hosts and match the corresponding snmp-profile via the model).

[FFHener]

I found another error. At fesev-core collectd reported daemon.err collectd[2546]: uc_update: Value too old: name = fesev-switch/snmp/gauge; value time = 1706777397.464; last cache update = 1706777397.464;. The Switch has the Corerouter as the time server, so this needs to be further investigated.

EDIT: Changing the interval of collecting data for the edgeswitch at etc/collectd/conf.d/bbb-snmp.conf from 30 to 60 seconds solved this problem for now, but i would still like to understand the root cause of this problem.

[FFHener]

The Problem is solved.

TLDR: Names of the value descriptors wherent unique.

I was using the port names as descriptors for the individual names because i thought this would be more helpful in understanding the graphs in grafana (e.g. zwingli-sama helps more than 7). The problem is that port names sometimes get used multiple times e.g. dhcp, intern, ..., but collectd uses these names from my understanding as keys. This problem seems kind of commen as it is mentioned in there docs. After switching the names in TypeInstanceOID to Port Numbers like N or GigabitEthernetN the problem went away. So this PR is ready to be merged now.

[FFHener]

The Error regarding temperature values is fixed as i introduced a new snmp_profile for SwOS Lite.

Got Out of Band approval from @pktpls