Loading…
Loading…
Resources · Interactive Guide
Networks drop. Servers reboot. Your data should not care. Every Wiman gateway has an inbuilt data logger that catches readings the moment the cloud is unreachable — and delivers them, in order, when the link returns. Here is the whole concept.
GOOD NEWS — every Wiman gateway ships with the data logger inbuilt. No missed readings, no lost data.
While the connection is healthy, readings stream to the server live. The moment the network fails or the server stops answering, the gateway keeps polling exactly as before — but every reading is written to the inbuilt logger instead of being lost. When the link returns, the logged records are forwarded with their original timestamps, and the server’s history closes up without a gap. Three logger modes control the behaviour: FIFO, Live and Disable.
Every Wiman gateway — MAX, PRO or LITE — comes with the data logger inbuilt. Whatever happens to the network, there is no chance of missing data.
Follow one full cycle — the status line narrates every phase. Online: a reading leaves the gateway every interval (minimum 5 s recommended), hops through the network, and lands on the server. Network failure: the tower goes red — readings drop into the logger and the LOG count climbs. Server down: the network recovers but the server does not answer — watch a packet reach it, get no acknowledgement, and bounce back red; logging continues. Recovery: live readings resume, the backlog drains in between the live packets — first-in, first-out — and the amber bars back-fill their original time slots until the LOG count reads zero.
Green = live reading. Amber 1 2 3 4 = logged readings, first-in first-out. Both failure types are shown: tower red = network failure; server LEDs red = server down — the packet that reaches it bounces back with no acknowledgement. On recovery the backlog drains in between live packets and the amber bars land in their original time slots — the LOG count walks 0→4→0.
Why the interval matters — network latency
4G / Cellular — the hop takes longer
LAN / Wi-Fi — same journey, quicker
Every packet must travel to the server and bring back an acknowledgement before it counts as delivered. The cellular hop alone can add hundreds of milliseconds to that round trip; a wired LAN or Wi-Fi hop adds almost nothing. That is why we recommend a minimum 5 s interval between packets — every reading completes its round trip, even on a slow network, with room for the backlog to drain in between.
All three modes poll the same way — the difference is what happens at the moment of reconnection. Watch the same three buffered readings (1, 2, 3) and the fresh reading (NOW) leave in a different order per mode:
FIFO — oldest first, then live
1 → 2 → 3 → NOW — the history arrives in the order it was measured
Live — newest first
NOW jumps the queue — dashboards recover instantly, the backlog follows
Disable — nothing was kept
only NOW goes out — the outage stays as a gap in the history
The logger drains oldest-to-newest: record 1 leaves before record 2, always. Live readings keep flowing at their normal interval, and the backlog slots into the gaps between them until the count reaches zero.
Pick it when the record matters: energy and billing, audits, reports, anything where every interval must exist.
The current reading takes priority: the moment the link returns, the freshest value goes out so dashboards and alarms recover instantly, and the logged backlog follows behind it.
Pick it when operators watch the screen: control rooms, live dashboards, alarm-driven operations.
The logger is off. While the link is down, readings are not kept — history simply has a gap for the outage window.
Pick it only when history has no value — pure live-view use, or when another system already stores the data.
The logger does not just watch the signal bars — it watches the whole delivery. Two different failures trigger the same protection:
That second case matters: plenty of systems survive a dead SIM but silently lose data during a server reboot or a cloud outage. The acknowledgement is the contract — a reading leaves the logger only once the server has actually taken it. The delivery mechanics (status codes, timeouts, retries) are the subject of the How Gateway Data Moves guide.
Every logged record carries the time it was measured, not the time it was finally sent. So when the backlog arrives — minutes or hours late — the server files each reading into its original slot. Trends, totals and reports look exactly as if the outage never happened; only the arrival was delayed, never the data.
delayed delivery, correct position — the timestamp travels with the reading
The data logger is standard across the range — MAX, PRO and LITE — not an add-on. You choose the mode (FIFO / Live / Disable) in the configuration utility, and the gateway handles the rest on its own, every time the connection misbehaves.