IMD SolutionComputing LibraryShop main store

Troubleshooting

Why the Ignition Edge historian keeps disappearing

Edge computing and IIoT Published 6 min read

Short answer

The Edge historian disappears because its saved record, not its module, is disabled: the EdgeHistorian row in config.idb's TAGHISTORYPROVIDEREP table carries ENABLED = 0, so startup skips registering the provider while Tag Historian reports Running. Back up config.idb, set ENABLED to 1, and the provider holds across reboots; a module restart only restores it until the next start.

On this page
  1. Why does the Edge historian disappear while the Tag Historian module shows Running?
  2. What does the Store and Forward message about installing the module mean?
  3. Which checks isolate a vanished provider from other history faults?
  4. How do you read the EdgeHistorian record in config.idb?
  5. Which recovery action actually survives a Gateway restart?
  6. Why did a restore in disabled state start this?
  7. Is the history from the missing interval lost?
  8. Questions asked

The Edge historian disappears because its saved record is disabled, not because the Tag Historian module failed. In config.idb, the Gateway's internal SQLite configuration database, the TAGHISTORYPROVIDEREP row with TYPE EdgeHistorian is stored with ENABLED = 0, so every Gateway startup skips registering the provider while the module loads and reports Running. A module restart brings the provider back until the next start; setting ENABLED to 1, with a backup of config.idb taken first, was reported to hold across reboots and Gateway restarts. The sections below confirm the record, apply the repair, and separate this fault from the ones that imitate it.

Why does the Edge historian disappear while the Tag Historian module shows Running? #

Because the provider's saved configuration is at fault, not the module: the EdgeHistorian record in the TAGHISTORYPROVIDEREP table of config.idb is stored with ENABLED = 0, so each Gateway startup skips registering it while the module loads and reports Running.

On affected gateways the symptom cluster is consistent. The Edge historian drops out of the Historical Providers list under Status → Tags, every tag stops saving and loading historical values, Store and Forward reports that the module needs to be installed, and the module page shows Tag Historian installed, activated and Running. Restarting the module restores the provider and the trends, and the loss returns at the next Gateway or server restart.

Two states are in play. Module status is runtime: loaded, activated, Running. Provider enablement is persisted configuration: the record that decides whether the historical provider is created at startup. A Running module does not prove an enabled provider, and a module restart re-registers the provider in memory without rewriting the saved record, which is why the durable fix is never another restart.

A module restart occurs when the Gateway restarts, when the module's configuration changes, and when a new module version is installed, so any of those events can appear to cure the historian for a while.

What does the Store and Forward message about installing the module mean? #

It means the history service Store and Forward can currently reach is absent, not that the module package is missing: a loaded module can supply a service that failed to register, so confirm module status and provider presence before reinstalling anything.

The message describes reachability, not installation. Affected sites had already tried the two obvious moves, restoring the Gateway backup (.gwbk) with everything enabled and reinstalling the Tag Historian module, and the provider still disappeared after restarts, because neither action touches the persisted provider record. Read the module page and the provider list together: both facts, not one, decide whether a reinstall is even relevant.

Which checks isolate a vanished provider from other history faults? #

Check them in this order: the Historical Providers list under Status → Tags, the Tag Historian module status, the Store and Forward page, one direct history query on a known tag, then live tag quality, because history access and live controller reads are separate paths.

What you see Where the fault sits Decision
Edge provider absent, Tag Historian Running provider enablement or registration read ENABLED on the EdgeHistorian record
Store and Forward asks for the module, module installed history service not registered verify the provider list before reinstalling
every historical trend fails together shared history infrastructure query one history-enabled tag directly
one trend fails, others load binding, tag path or query range compare it with a working history query
live tag quality is bad too tag provider, driver or controller link diagnose communication, not history

Live values can stay healthy while history is gone, because the two use different paths. Do not change controller logic to recover a Gateway history provider.

How do you read the EdgeHistorian record in config.idb? #

Open config.idb, the Gateway's internal SQLite configuration database, and read the TAGHISTORYPROVIDEREP table: the row with TYPE EdgeHistorian is the Edge historian, and on affected gateways that row carries ENABLED = 0 where a healthy installation carries 1.

Work on a copy with any SQLite browser. The reported columns are TAGHISTORYPROVIDEREP_ID, NAME, ENABLED, TYPE and DESCRIPTION. Identify the row by TYPE, not by ID: the example below had TAGHISTORYPROVIDEREP_ID 1 and NAME historian, but those are that gateway's values, not identifiers you can assume.

TAGHISTORYPROVIDEREP_ID NAME TYPE ENABLED Reading
1 historian EdgeHistorian 0 saved disabled, the fault
1 historian EdgeHistorian 1 healthy record

Which recovery action actually survives a Gateway restart? #

Back up config.idb, stop the Gateway, set ENABLED from 0 to 1 on the EdgeHistorian row, start the Gateway, then prove it with one controlled restart: that repair was reported to hold across reboots and Gateway restarts, where a module restart held only until the next start.

  1. Back up config.idb.
  2. Stop the Gateway, so the database is closed and the change is read at startup.
  3. Set ENABLED from 0 to 1 on the EdgeHistorian row and change nothing else.
  4. Start the Gateway, confirm the provider is listed under Status → Tags → Historical Providers, and confirm Store and Forward no longer asks for the module.
  5. Prove persistence with one controlled Gateway restart: on affected sites a restart was the reliable trigger, so it is also the honest test.
Action Result on affected gateways Decision
restart the Tag Historian module provider and trends return, loss recurs at the next start immediate recovery only
restart Ignition Edge or the server re-triggered the disappearance every time persistence test after the repair, not a recovery
restore the .gwbk with everything enabled recurrence continued not sufficient; read ENABLED after any restore
reinstall the Tag Historian module recurrence continued skip it while the module is installed and Running
set ENABLED to 1 on the EdgeHistorian row reported to hold across reboots and Gateway restarts the durable repair

config.idb is internal application configuration: keep the backup, change the one value, and if the vendor documents a different supported repair path for your version, follow that instead of broader database edits.

Why did a restore in disabled state start this? #

A backup restored in disabled state persists the disabled flags of the objects it contains, and every startup reapplies them: one affected site traced its first disappearance to exactly that restore, and the later power failures only re-exposed what was still saved.

Restoring in disabled state disables broadly and takes real work to undo. Ignition 8.1.11 added the ability to re-enable the Edge tag provider from the Gateway webpage; before that release it was not editable in Edge, so a restore-disabled left it stuck. The Edge tag provider and the historical provider are different objects, so recovering live tags does not complete the history checks. Reports span Edge 8.1.14 on RHEL 8.5 x86, with recurrences on 8.1.32 and 8.1.42, which are reported environments, not a bounded affected range.

Is the history from the missing interval lost? #

No, the interval is not backfilled: while the provider is missing, tags stop saving history and nothing reconstructs those timestamps afterward, so read the trend from the first gap and treat every day without the repair as history you are losing.

Two follow-ups make the recovery honest. Confirm new writes are landing where the Edge tag historian writes its data, and read the central trend the way it will actually render, because why Edge historian gaps look like held values matters here: the outage can look shorter centrally than it was.

Questions asked

Is it safe to edit config.idb while the Gateway is running?

No. Stop the Gateway first. config.idb is an SQLite database with a single writer, and a live edit can race with the Gateway's own writes or be overwritten at shutdown. Copy the file, stop the Gateway, change the one value, start it back up, and keep the copy until the provider has survived a full restart.

How do I find when the historian actually stopped?

The wrapper log rarely names the historian directly. Bracket the stop instead: note the timestamp of the last good point on a trend, the first history query that returned nothing, and every Gateway restart in between, read from the wrapper log before it rotates. A nearby power failure is a clue, not a timestamp.

Does the same check apply to a full Ignition gateway, not only Edge?

The table and column are the same: TAGHISTORYPROVIDEREP with an ENABLED column exists on a full gateway too. The confirmed disappearing-provider reports are on Edge, spanning 8.1.14 on RHEL 8.5 x86 plus 8.1.32 and 8.1.42, so on a full gateway treat the check as diagnostic, not proven, until you read the row.

What does Tag History Data Sync have to do with this?

Nothing demonstrated. It was flagged as one more artifact to inspect after a backup restored in disabled state, alongside the Edge tag provider, because restore-disabled disables broadly. If you run Tag History Data Sync, check its enabled state and its diagnostics separately; no report ties it to this fault.

Why does the historian sometimes disappear with no restart at all?

That case is reported but not explained by the saved flag alone. ENABLED = 0 fully accounts for loss at every startup; a provider that vanishes while the Gateway stays up points instead at registration or a runtime failure in the module. If the row reads 1 after such an event, capture the startup and module logs and escalate.

Keep reading

  • Troubleshooting

    Historian stores zero after an MQTT edge node reconnects

    A 0 stored just before the true value at every MQTT edge reconnect is a reported pattern on Ignition 8.1.5 through 8.1.15 with the MQTT Transmission and Engine modules 4.0.6 through 4.0.10. Confirm it by querying stored records around each reconnect, exclude those zeros from process data, and remove them with a timestamped rule instead of a blanket zero filter.

    Edge computing and IIoT

  • Comparison

    MQTT Recorder or Tag Historian for plant data

    Tag Historian stores timestamped values for individual tags in a SQL database, built for trends and time-based queries. MQTT Recorder receives a block of data from an MQTT server and inserts it as one row in a database table, built for completed field-generated records such as tickets and batch results. Choose by the unit of information, then put buffering where the link fails.

    Edge computing and IIoT

  • Comparison

    Evaluating Ignition historian with TimescaleDB for your plant

    The Gold Technology Provider announcement is a reason to test TimescaleDB behind Ignition, not proof it replaces your historian. No footprint, compression or performance number is documented for this pairing. Adopt it only after a replay test on your exact Ignition 8.3 build measures growth, ingestion continuity, query response and backup/restore against your current store.

    Edge computing and IIoT