Looking at better logging in debug mode for snmp

[i3149]

Closes #705

Changes how debug level logging goes.

• Removes 1 NR log which is confusing to the user.

• Adds a log on each poll which looks like:

2024-04-29T22:26:23.125 ktranslate/ [Debug] KTranslate>bart (192.168.0.200) polling 1.3.6.1.2.1.1.4.0 using github.com/gosnmp/gosnmp.(*GoSNMP).WalkAll-fm took 6.968ms. attempt 0 of 2.

• Adds a log after each metric poll (of all the oids) which looks like:

2024-04-29T22:28:33.472 ktranslate/ [Debug] KTranslate>bart Metrics Poll: In 453.71575ms polled 14 mibs, 0 missing.

Thoughts?

[thezackm]

Can we add a tag to the kentik.snmp.Uptime metric that shows the current logging level? Something like tags.container_logging?
I'm concerned about blowing out storage if debug is left enabled. These logs are contextually valuable, but if left running I can see a problem on the horizon.

[i3149]

How does this look:

"SysLogLevel": "Info",

There's a hack where things prefixed Sys only get added to Uptime. So this lets things happen with a min of fuss.

[thezackm]

looks great; ty