TM1: Query Cube (Agent)
Description
Section titled “Description”The on-premises twin of TM1: Query Cube: the same read — a saved view or an MDX query against a TM1 cube, with an optional point-of-view (POV) slice, landed as a project table — dispatched through an installed PlaidLink agent on your network instead of called cloud-direct. Use this step, instead of TM1: Query Cube, when the TM1 server is only reachable from inside your own network.
This step appears in the step menu under IBM TM1 / Planning Analytics.
What’s Different From TM1: Query Cube
Section titled “What’s Different From TM1: Query Cube”| TM1: Query Cube | TM1: Query Cube (Agent) | |
|---|---|---|
| Transport | Cloud-direct over REST | Dispatched to a PlaidLink agent on your network |
| Configuration fields | — | Identical, plus a required Agent |
| Cube / View discovery | Live, from the connected database | Not available — type the names by hand |
| Preview before saving | Yes | Not available |
Everything else — the connection, the database, the view/MDX choice, the point-of-view slice, Suppress Zero/Empty Cells, Row Limit, and the incremental-extract pattern by period — works exactly as described on the TM1: Query Cube page. Only the differences below are specific to this step.
Configuration
Section titled “Configuration”Source and Target
Section titled “Source and Target”| Field | Required | Notes |
|---|---|---|
| Agent | Yes | The PlaidLink agent to dispatch this read to. Its build must include the TM1 handler (see Requirements below). |
| TM1 Connection | Yes | The TM1 connection to read through. The connection picker is not filtered by the selected agent — share the connection with the agent separately (see below). |
| Environment | Yes | The connection’s environment. |
| Database | Yes | The TM1 database (instance) to query. |
| Cube | Yes | The TM1 cube to query. Typed by hand — there is no live discovery on this step. |
| Target Table | Yes | The project table the returned cellset lands in. |
Slice and Options
Section titled “Slice and Options”Same fields as TM1: Query Cube — Source Mode (By View / By MDX), View, MDX, Point of View, Suppress Zero/Empty Cells, and Row Limit — all typed entry rather than discovered, and all validated when the step runs rather than in a preview.
No Agent-Mediated Discovery
Section titled “No Agent-Mediated Discovery”TM1: Query Cube’s Cube, View, and Dimension pickers query the connected server live while you’re editing the step. That live discovery does not exist for this step: an agent is only reachable during a run, not while a form is open in your browser. Cube, view, and MDX are typed entry, and an unrecognized name fails the run with a clear error rather than failing silently. For the same reason there is no Preview button — a preview would have to reach TM1, which only the agent can do.
Requirements
Section titled “Requirements”Reading an on-premises TM1 server needs:
- An installed PlaidLink agent on a machine with network access to the TM1 server. See Install PlaidLink.
- A build of that agent that includes the TM1 handler. PlaidCloud checks the agent’s advertised capabilities before dispatching; an agent whose build predates this capability cannot serve the request. PlaidLink does not auto-update — an existing agent must be manually rebuilt/reinstalled to pick up TM1 support. See Upgrade PlaidLink.
- The TM1 connection shared with that agent, in addition to whatever member/group access it already has — grant it from the connection’s
Actions > Agentslist. See Security Model.
Related
Section titled “Related”- TM1: Query Cube — the cloud-direct twin this step mirrors.
- TM1: Read Dimension — load a TM1 hierarchy into a PlaidCloud dimension, cloud-direct.
- Connect to TM1 (guide) — create the connection.
- TM1 Connector — connection field reference, including the on-prem agent path.
- PlaidLink — installing and upgrading agents.
- IBM TM1 / Planning Analytics Steps