thingsboard

27075 Commits

101 Branches

81 Tags

387 MiB

Tree: bbdfc416ca

Commit Graph

Author	SHA1	Message	Date
Sergey Matvienko	8aef432958	docs(monitoring): document RPC monitoring config and behaviour (#15541 ) - tb-monitoring.yml: add the optional rpc.{enabled,request_timeout_ms} sub-block to every transport target with comments explaining the per-transport mechanism, plus a commented-out secure-MQTT example - monitoring/src/main/resources/README.md: new file with the YAML key table, a per-transport behaviour summary, the failure-attribution semantics (RpcInfo => "<transport> RPC"), and troubleshooting notes including the rpc.request_timeout_ms < monitoring.rest.request_timeout_ms ordering constraint - 2026-04-29-rpc-monitoring-design.md: update §5 to list MQTT + HTTP + CoAP + LwM2M as shipped phases (replacing the previous "MQTT only + follow-ups" framing), reword §6 open decisions as resolved decisions with all four transports in scope, add §4.6 explaining the LwM2M Read-vs-echo asymmetry and the conservative 6000 ms default	1 month ago
Sergey Matvienko	4d4a6d0dcd	docs(monitoring): RPC monitoring design (rc baseline) Adds an opt-in two-way RPC companion check to each transport target so the monitoring loop covers the cloud-to-device direction (REST → rule engine → transport → device → response), not just the uplink. The shape mirrors how the calculated_fields flag layers an extra assertion onto the existing telemetry check today: same loop, same auto-provisioned device, same MQTT client (also subscribes the RPC topic when rpc.enabled), one extra round-trip per cycle. Failure attribution flows through MonitoringReporter unchanged, so RPC failures appear as their own row in the IncidentManager header ("MQTT RPC") alongside the transport row. Includes ASCII + mermaid diagrams of the current and proposed flows, a sequence diagram of the round-trip, an incident-attribution table, phased implementation plan, and open decisions.	1 month ago

Author

SHA1

Message

Date

Sergey Matvienko

8aef432958

docs(monitoring): document RPC monitoring config and behaviour (#15541 )

- tb-monitoring.yml: add the optional rpc.{enabled,request_timeout_ms}
  sub-block to every transport target with comments explaining the
  per-transport mechanism, plus a commented-out secure-MQTT example
- monitoring/src/main/resources/README.md: new file with the YAML key
  table, a per-transport behaviour summary, the failure-attribution
  semantics (RpcInfo => "<transport> RPC"), and troubleshooting notes
  including the rpc.request_timeout_ms < monitoring.rest.request_timeout_ms
  ordering constraint
- 2026-04-29-rpc-monitoring-design.md: update §5 to list MQTT + HTTP +
  CoAP + LwM2M as shipped phases (replacing the previous "MQTT only +
  follow-ups" framing), reword §6 open decisions as resolved decisions
  with all four transports in scope, add §4.6 explaining the LwM2M
  Read-vs-echo asymmetry and the conservative 6000 ms default

1 month ago

Sergey Matvienko

4d4a6d0dcd

docs(monitoring): RPC monitoring design (rc baseline)

Adds an opt-in two-way RPC companion check to each transport target so
the monitoring loop covers the cloud-to-device direction (REST → rule
engine → transport → device → response), not just the uplink.

The shape mirrors how the calculated_fields flag layers an extra
assertion onto the existing telemetry check today: same loop, same
auto-provisioned device, same MQTT client (also subscribes the RPC
topic when rpc.enabled), one extra round-trip per cycle. Failure
attribution flows through MonitoringReporter unchanged, so RPC
failures appear as their own row in the IncidentManager header
("MQTT RPC") alongside the transport row.

Includes ASCII + mermaid diagrams of the current and proposed flows,
a sequence diagram of the round-trip, an incident-attribution table,
phased implementation plan, and open decisions.

1 month ago

2 Commits (bbdfc416ca38588dec6b282c7ca6da9c1dd08e98)