Delayed iMonnit Alerts After an Outage: Separate Event Time from Upload Time
Updated: Nov 5
An iMonnit notification arrives just after a communications outage, but the dashboard now shows an ordinary reading. Has the sensor failed, has the alert arrived late, or did a real excursion occur while the connection was unavailable? Treat those as separate questions. First establish the current condition using the site's approved procedure, then reconcile the notification with the recorded history.
This guide covers delayed historical alerts in Monnit ALTA gateway-based monitoring. It is an investigation method, not a guarantee of uninterrupted records or immediate alarm delivery. For a safety, food-quality or other critical duty, the responsible person must assess the actual event and the suitability of the monitoring arrangements.
Why an old reading can generate a new notification
Monnit documents that an ALTA sensor which loses its radio link after previously reporting can store readings. After reconnection, it sends a current reading and uploads its backlog in batches. Historical readings can trigger configured Rules and notifications, even when the triggering condition occurred on an earlier day. See Understanding ALTA Sensor Data Logging.
That behaviour makes notification arrival time insufficient evidence of when the measured condition occurred. A current normal value also cannot establish that the asset remained within limits during the communications gap. Keep the question of present condition separate from the investigation of the earlier excursion.
Identify where the backlog was held
A lost sensor-to-gateway radio link differs from a gateway losing its server connection. Monnit explains that a powered ALTA gateway can retain messages from its connected sensors while server communication is unavailable, then upload them in batches. Logging is not supported through the Serial MODBUS Gateway or the Ethernet gateway's MODBUS TCP/SNMP interfaces. Check the actual operating path against ALTA Gateway Data Logging.
The pictured MONNIT Ethernet Wireless Sensor Gateway is the ProSense catalogue product associated with this example. Confirm the installed model and firmware from the actual device record; a similar-looking enclosure does not establish its logging behaviour.
Build a short event record before changing settings
Use the following checks as an investigation worksheet. Record observations and uncertainties rather than converting assumptions into timestamps.
Confirm the point: record the asset, sensor ID, gateway ID, rule name, measured units and threshold. Check that the alert and the displayed history concern the same sensor.
Capture the notification: retain its arrival time and any event time it actually contains. Record the time zone and UTC offset. Do not relabel message arrival as measurement time.
Inspect the history: save an authorised export or screenshots spanning the outage and reconnection. Record the timestamps and values shown, and note when older entries became visible if that was observed.
Check present conditions separately: use the approved on-site verification method and record its time, instrument and result. Escalate a current abnormal condition through the site's response procedure.
Compare independent evidence: authorised network outage records, gateway communications and operator notes can help establish the sequence. Mark missing evidence explicitly.
A hypothetical recovery timeline
Suppose a non-critical room-monitoring sensor loses its radio link at 09:00, records a high reading at 09:40, and reconnects at 11:00. Its current reading is normal, while a notification associated with the earlier high value arrives at 11:05. These times are illustrative, not predicted device performance.
The investigation record should distinguish the 09:40 recorded excursion, the 11:00 reconnection and the 11:05 notification arrival. It should also identify the time and result of any independent current-condition check. Do not close the event merely because the latest reading is normal; the responsible person still needs to assess what the earlier excursion meant for the monitored asset.
Check timestamp and power-loss limitations
Monnit warns that a gateway offline for days can develop a clock offset. Reconnection resynchronises its clock, but affected uploaded readings retain their offset timestamps. Preserve the original record and flag uncertain timing instead of silently shifting the data to fit an expected story.
Power-loss retention also depends on firmware. Monnit lists Ethernet Gateway 4 firmware 1.0.7.1 and later as supporting retention of logged messages through power loss. Verify the installed version and the manufacturer's current guidance before relying on that capability.
A gap in the recovered history needs its own assessment. Monnit notes that sensor logging stops when its message capacity is reached. Do not infer continuous coverage simply because some older readings reappear. Avoid promising a recovery duration or storage capacity without checking the exact sensor and configuration.
Common mistakes and the next useful action
Avoid repeatedly power-cycling equipment, changing thresholds to silence recovery notifications, or clearing evidence before the investigation is complete. Any authorised change should have a separate reason, approval and validation record. A delayed alert may reveal a real historical problem as well as a weakness in the communications or response plan.
Prepare a support pack containing device models and IDs, firmware versions, the outage window, the rule involved, the original notification, history extracts and the independent observations. Exclude passwords and security codes. Ask for help reconciling the record before changing the monitoring design; do not assume every delayed notification has the same cause.
Need help with your application?
For instruments and product selection, visit ProSense Instruments, email sales@prosenseinstruments.com.au or call 1800 560 854.
For a complete monitoring or automation project, visit ProSense Solutions or email sales@prosense.com.au.
For software enquiries, visit ProSight Software or email info@prosightsoftware.com. Confirm integration requirements for the exact devices and system before selecting an architecture.





Comments