IMD SolutionComputing LibraryShop main store

Comparison

How AVEVA Historian publishes collected data to a client

Edge computing and IIoT Published 7 min read

Short answer

Yes — AVEVA Historian publishes collected data, but through a client-matched output rather than one universal server. Seven paths cover the cases: SuiteLink, PCS iData, OMI Replay, SQL History or Live tables, MQTT, OData, and Gateway as an OPC DA client. Keep the OPC UA link as collection, then pick the output by the client's protocol and whether it needs live or stored values.

On this page
  1. Can I configure Historian as a server to publish data?
  2. Which output path fits the client I already have?
  3. Does the client need live values or stored history?
  4. How do I keep the OPC UA collection separate from publishing?
  5. How do I publish only part of the collected data?
  6. What do I check when the client cannot read a value?
  7. How do I prove the whole path from PLC to client?
  8. Questions asked

Yes. AVEVA Historian collects through OPC UA on one side and publishes through seven client-facing output paths on the other: SuiteLink, PCS iData, OMI Replay, the SQL Server History and Live tables, MQTT, OData, and the Gateway acting as an OPC DA client. There is no single "historian as server" switch that exposes every collected tag over every protocol; you configure the output that matches each downstream client. Keep the OPC UA link exactly as it is — it is the collection leg, not publishing. If the platform choice itself is still open, plant historian selection by scale and support needs comes first; everything below assumes Historian is installed and collecting.

Can I configure Historian as a server to publish data? #

Yes, but "server" means one interface at a time, not a universal one: AVEVA Historian exposes collected data through seven output paths, and you enable the one each client speaks rather than opening the whole store.

The OPC UA setup already running is ingestion. Historian IDAS with the Gateway configured as an OPC UA client is the documented route for capturing variables from any OPC UA server into the archive, and it says nothing about how anything downstream reads them back. Publishing is configured on the output side, one interface at a time.

When you search the vendor documentation, these output paths are indexed under terms including aahIOSvrSvc, iData, Replay, Live and MQTT; search them alongside your installed version, because the list has grown by release and PCS iData appears only from Historian 2023 R2.

"Publish data" also means four different jobs here: serving each tag's latest value, answering queries against stored history, replaying history into a specific client, and streaming live values. The next two sections split those before any configuration screen is touched.

Which output path fits the client I already have? #

Start from the protocol the client already speaks, because that alone eliminates most of the list: an AVEVA application gets SuiteLink, a SQL consumer gets the History or Live tables, an API consumer gets OData, and an OPC consumer gets the Gateway route.

Downstream client Output path What it delivers Scope to verify
AVEVA application such as InTouch Integrated SuiteLink server The latest historical value per tag AVEVA clients only; not a live stream
PCS client PCS iData provider A PCS data feed Added in Historian 2023 R2
AVEVA OMI view OMI Replay Replay of stored history Historical replay, not current values
SQL Server client History and Live tables Stored history; current values Point each query at the table it needs
MQTT consumer MQTT publishing A feed to the broker Support is basic; test against client needs
REST or API client OData APIs Access to historical data Historical reads, not live streaming
OPC DA client Gateway as an OPC DA client Live value streaming The most complex route of the seven

Commit per client. An InTouch or other AVEVA client that only needs the current reading of each tag takes the integrated SuiteLink server. A reporting or BI tool that queries SQL Server takes the SQL output, with each query pointed at History or Live according to what it asks. An analytics platform that calls an API takes OData. An OPC consumer that must see values change takes the Gateway configured as an OPC DA client, at the cost of the extra configuration that route needs.

A broker-based consumer takes the MQTT output. Its documented support is basic, so verify the exact publish and subscription behavior the client needs before committing; where the built-in publishing falls short, bridging OPC UA PLC data to MQTT with an edge PC is the established pattern.

Does the client need live values or stored history? #

This question eliminates more options than protocol does: the Gateway OPC DA route is the only true live stream, SQL splits into the History table for stored values and the Live table for current ones, and SuiteLink, OMI Replay and OData each serve a historical reading.

SuiteLink publishes the latest historical value per tag — the archive's newest stored value — which is what an InTouch screen wants and not a substitute for streaming. OMI Replay pushes stored history into AVEVA OMI. OData answers historical queries over an API. The Live table answers what the value is now; the History table answers what it was over a range; a client pointed at the wrong one fails with collection running perfectly.

When the downstream stack instead expects latest values over open protocols such as OPC UA, exposing historian latest values over OPC UA or MQTT describes that pattern for stacks built around it.

How do I keep the OPC UA collection separate from publishing? #

Treat them as two independent legs: the OPC UA client connection reads PLC variables into the historian, and each output path is configured and tested on its own; a working input proves nothing about any output.

Work in this order:

  1. Confirm the PLC variables update in the archive through the existing OPC UA path.
  2. List the subset of collected tags the higher-level client needs, and mark each value as current, historical or replayed.
  3. Identify the protocol that client supports among the seven paths.
  4. Configure the matching output, then test it with the real downstream client, separately from collection.

A fault on one leg does not migrate to the other. A client binding failure looks the same whether the archive is full or empty, which is why the legs are proven independently before they are proven together.

How do I publish only part of the collected data? #

Publishing is a selected scope, not the whole store: name the tags in the chosen output interface and leave the rest internal, because no output path mirrors every collected tag by default.

Each interface defines its own published set, so what the client can see is a per-output decision you make explicitly. Keep that set as small as the client's function requires: every tag you expose is readable by anyone who reaches that client.

Worked example, with illustrative numbers. A collector archives 4,200 tags from the PLC through the OPC UA path. The higher-level client — a reporting tool that queries SQL Server — needs 310 of them, so 4,200 - 310 = 3,890 tags stay inside the historian. Of the 310, the tool's overview screen shows current values for 45 tags: those reads go to the Live table, and the remaining 310 - 45 = 265 are history-only reads against the History table. Decision: configure the SQL output over the 310-tag subset and stop there — not SuiteLink, because the client is not an AVEVA application; not OData, because it queries SQL Server rather than an API; not the Gateway OPC DA route, because nothing downstream consumes OPC.

What do I check when the client cannot read a value? #

Trace the failure from the client screen back through three checkpoints — client binding, output interface, input tag — because an absent source value and a broken client binding fail differently and get different fixes.

Observed result Check next What it usually means
Client cannot connect or browse The protocol and provider the client is pointed at Output selection or client binding problem
A tag is missing from the output Whether the tag is collected and inside the published scope Collection problem or scope selection
The value is stale The same tag in the archive, then at the client Stale at the source points at OPC UA collection; current at the source points at the output path
A history query returns nothing Which table the query hits, History or Live, and the time range Wrong table, or a range in which nothing was logged

Use one known, currently updating tag as the test point before widening the published scope. For SQL clients in particular, an empty History result often means the client asked a range in which nothing was logged, or asked the Live table a history question — neither is a collection fault.

How do I prove the whole path from PLC to client? #

Prove each leg, then the pair: confirm the test tag's value updates in the historian, read that same tag through the chosen output with the real client, and only then expand to the full published subset.

Watch one test tag update in the archive first. Then force or wait for a PLC value change and see it arrive at the client through the route you configured — the Live table, the MQTT feed or the OPC DA route for current values; a History query or an OData call for stored ones.

Then publish the full subset, confirm the excluded tags do not appear in the client's scope, and record for each client the output method and the value type it expects, so the next fault separates a source problem from a binding problem in one step. Every selection here reduces to two questions answered in order: what protocol does the client speak, and does it need live or stored values?

Questions asked

Can AVEVA Historian act as an OPC UA server for downstream clients?

Not in this configuration. The OPC UA link is the input leg: the Gateway side acts as an OPC UA client reading the PLC's server into the archive. For a downstream OPC consumer, the documented route is the Gateway configured as an OPC DA client for live value streaming.

Does the SuiteLink server stream live values?

No. The integrated SuiteLink server publishes the latest historical value of each tag — the newest stored value — which is what an InTouch screen needs. True live streaming to an OPC consumer goes through the Gateway OPC DA route; current-value SQL reads go to the Live table.

Which SQL table should a reporting client query?

Both, by query type: the Live table for current values, the History table for stored values over a time range. Point overview screens at Live and trend or report queries at History. Getting this backwards returns empty or stale-looking results while collection runs normally.

Is the MQTT output enough for a full IoT feed?

Treat it as basic until proven otherwise on your installed version. The checks that decide it are topic structure, update triggers, and whether the client needs to subscribe or only to receive published values. If any of those fail, route the feed through an edge PC bridging OPC UA data to MQTT instead of the built-in publisher.

Keep reading

  • How-to

    Deploying Edge Data Store in Docker on a controller PC

    Run Edge Data Store as a Docker container on the controller-class edge PC and let its OPC UA adapter ingest the local server over localhost. Define egress targets, buffering and forwarding schedules in the JSON configuration, size the buffer from tag count, scan rate and outage target—50 tags at 1 event/s across an 8-hour outage is about 288 MB—and sign off only after a live pulled-cable test.

    Edge computing and IIoT

  • How-to

    Expose historian latest values over OPC UA or MQTT

    Yes — connect an integration gateway to the historian's classic I/O Server over SuiteLink (application name aahiosvrsvc, topic TagName) and publish the current values as OPC DA and MQTT. After a historian restart the I/O Server and the Live table return nothing until the source sends a new value, so a true last-known-from-disk read needs T-SQL or the historian SDK.

    Edge computing and IIoT

  • How-to

    Forward Canary historian data through a SaF proxy

    Forwarding site history to a corporate Canary historian is a Store and Forward routing job. Enable the Proxy feature in the sending SaF service, set RouteDestinations to localhost plus the corporate historian, and open TCP 55291 inbound on the destination. An API token with Write permission to the Historian View is required only when Tag Security is enabled on the destination Identity service.

    Edge computing and IIoT