fix(snmp): walk only what the switch can safely answer

stacks/observability/snmp-exporter/generator.yaml

@@ -92,12 +92,42 @@ modules:
   # per-request timeout, so those values are pinned here rather than left to
   # the exporter's defaults (max_repetitions 25, retries 3, timeout 5s).
   #
-  # 5 * ~22 ifTable columns is a handful of round trips against a switch that
-  # cannot absorb 25. Worst case here is retries * timeout = 21s, comfortably
-  # inside the job's 45s scrape_timeout.
+  # Worst case here is retries * timeout = 21s, comfortably inside the job's 45s
+  # scrape_timeout.
+  #
+  # Only the five columns anything actually consumes are walked. This switch was
+  # bricked once by a full walk and has wedged since, so scrape volume is a
+  # safety property here, not an efficiency one. Five columns x 26 ports is ~130
+  # varbinds, about a quarter of what walking the whole ifTable costs.
+  #
+  # Column 22 (ifSpecific) must never be fetched. This switch returns it with a
+  # zero-length value, seen on the wire as
+  #
+  #   .1.3.6.1.2.1.2.2.1.22.26=
+  #
+  # An OBJECT IDENTIFIER cannot be zero-length, so snmp-exporter rejects the
+  # entire GETBULK response with
+  #
+  #   error in unmarshalResponse: error parsing OID Value: invalid OID length
+  #
+  # and the scrape yields nothing — one malformed varbind loses every metric.
+  #
+  # Note this is why column 21 is absent too, and why listing columns 1-21 did
+  # NOT fix it: GETBULK returns the next N varbinds and does not stop at a
+  # subtree boundary, so walking .21 overshoots into .22 and the bad varbind
+  # comes back anyway. The last column walked must be far enough from .22 that
+  # max_repetitions cannot reach it. `overrides: ifSpecific: ignore` does not
+  # help either — that only drops the metric, the reply must still be decoded.
+  #
+  # Adding a column is cheap; adding .21 or .22 is not. Anything past .16 needs
+  # to be checked against this.
   mokerlink:
     walk:
-      - 1.3.6.1.2.1.2.2  # IF-MIB::ifTable
+      - 1.3.6.1.2.1.2.2.1.1  # ifIndex
+      - 1.3.6.1.2.1.2.2.1.2  # ifDescr       (label lookup)
+      - 1.3.6.1.2.1.2.2.1.8  # ifOperStatus  (SwitchPortDown)
+      - 1.3.6.1.2.1.2.2.1.10 # ifInOctets
+      - 1.3.6.1.2.1.2.2.1.16 # ifOutOctets
     max_repetitions: 5
     retries: 2
     timeout: 7s