×
‹
›

Search

Connectors

Contact support

Helpdesk portal

Peliqan documentation
/
On-prem connectivity
/
Peliqan Satellite - Windows on-prem Data sync Agent

Peliqan Satellite - Windows on-prem Data sync Agent

What is Peliqan Satellite

On-prem agent

Peliqan Satellite is a Windows agent you install on an on-prem server. It reads data from an on-prem database (SQL Server, or anything reachable via ODBC) and copies it into your Peliqan data warehouse (Postgres or BigQuery). Once the data is there, you can use it for BI/reporting, MCP access to your business data, build Data Apps, publish APIs etc.

What are “pipelines” in Peliqan Satellite

Each pipeline is one source configuration → one target schema. You can run several pipelines side by side. One source configuation can capture data from one schema in one database or it can capture data from multiple databases. Use multiple pipelines if you have different source configurations or a different target configuration.

Preparing the source database(s)

Create a dedicated database user on the source DB server, with read-only access. Use e.g. “peliqan” as the username.

Preparing SQL Server

  1. Make sure TCP/IP is enabled in SQL Server.
  2. Use a SQL Server user for authentication, not Windows auth. The Peliqan Windows Service usually runs under a different account than the logged-in user, so Windows auth commonly fails for scheduled (background) syncs even though it works fine when testing from the UI.

Enabling access to multiple databases

If new databases can appear over time and you want the agent to pick them up automatically, grant access to all databases (including new databases that are added in the future):

GRANT CONNECT ANY DATABASE TO [peliqan];
GRANT SELECT ALL USER SECURABLES TO [peliqan];

CONNECT ANY DATABASE lets the agent connect to every database on the instance, present and future. SELECT ALL USER SECURABLES lets it read from every table/view it can connect to.

Installing the agent

  1. Download the installer: https://peliqan.io/winagent/latest/PeliqanSatellite-win-x64.msi
  2. Run it and follow the steps in the Windows installer.

After installing, you'll see 5 tabs: Source, Target, Settings, Logs, and Manage pipelines. Visit the tabs one by one to configure the agent.

Configuration for specific data sources

Use Peliqan Satellite to sync data from your on-prem software, for example accounting software installed on your own computer.

Expert/M

Add a SQL Server user for Peliqan, on the SQL Server used by your Expert/M accounting software. Grant access to all current and future databases (see above).

  • Select “SQL Server” as the source type and fill in all the details.
  • Enable the option “Sync all databases”
  • Enable “Merge all source tables based on name” since Expert/M will create a separate DB for each end-customer (dossier).
  • DB include keywords: leave blank (this will select all databases)
  • Schema include keywords: leave blank (this will select all schemas in each selected DB)
  • Table include keywords: leave blank (this will select all tables in each selected schema)

Adsolut

Add a SQL Server user for Peliqan, on the SQL Server used by your Adsolut accounting software. Grant access to all current and future databases (see above).

  • Select “SQL Server” as the source type and fill in all the details.
  • Enable the option “Sync all databases”
  • Enable “Merge all source tables based on name” since Adsolut will create a separate DB for each end-customer (dossier).
  • DB include keywords: leave blank (this will select all databases)
  • Schema include keywords: leave blank (this will select all schemas in each selected DB)
  • Table include keywords: leave blank (this will select all tables in each selected schema)

Plenion

Add a SQL Server user for Peliqan, on the SQL Server used by your Plenion ERP software.

Select “SQL Server” as the source type and fill in all the details.

Exact Venice

Add a DB user for Peliqan, on the Actian Zen DB used by Exact Venice.

Create a Computer/System DSN.

Select “ODBC” as the source type and fill in all the details of the Actian Zen database.

Sage Bob 50

Coming soon.

Winbooks

Coming soon.

Other

Add a DB user for Peliqan, on the DB used by your on-prem software.

Source tab — where the data comes from

SQL Server

Use this for e.g. Expert/M, Adsolut, Plenion etc.

Source tab configured for SQL Server, single-schema scope, Windows Authentication
Source tab configured for SQL Server, single-schema scope, Windows Authentication
Field
What to put
Database type
SQL Server
Server
Examples: localhost, localhost\SQLEXPRESS (named instance), or 10.0.0.5,1433 (IP + port)
Port
Ignored when you use a host\instance name — SQL Server resolves the port itself in that case
Authentication
SQL Server Authentication (recommended) or Windows Authentication
Database name / Schema name
The single database + schema to sync (only shown in the default "single schema" scope — see below)

Important on Windows Authentication: Test connection and Sync now run as your Windows user, so they'll work fine. Scheduled syncs run as the Windows Service account instead (LocalSystem by default), which usually has no SQL access at all. If you need scheduled syncs to work with Windows Auth, change the service's log-on account via services.msc → Peliqan Satellite → Properties → Log On. Otherwise, just use SQL Server Authentication for anything that needs to run unattended.

ODBC

Use this for e.g. Exact Venice (using an Actian Zen DB).

Set Database type to ODBC and enter a single DSN or connection string, plus optional Username/Password.

  • Enter either a plain DSN name (e.g. MyDatabase) or a full connection string (e.g. Driver={Pervasive ODBC Interface};dbq=MyDatabase;).
  • The DSN must be a Computer/System DSN, not a User DSN. Note: a User DSN isn't visible to the Windows Service, so scheduled syncs would fail to find it even though manual tests from the UI work fine.
  • Username/Password are optional. If the DSN itself already stores credentials (common for e.g. Exact Venice/Actian Zen setups), leave them blank.
Source tab configured for ODBC
Source tab configured for ODBC

Sync scope: one schema vs. everything

Source tab in "Sync multiple databases, schemas and tables" mode, with include-keyword fields
Source tab in "Sync multiple databases, schemas and tables" mode, with include-keyword fields

Sync scope “one database”

By default the agent syncs one schema in one database — just fill in Database name + Schema name and you're done.

Sync scope “multiple databases”

Switch to "Sync multiple databases, schemas and tables" when you need the agent to discover multiple databases/schemas/tables on its own. This is the mode various accounting-software integration uses (Expert/M, Adsolut), since those products create a separate database per end-customer.

In this mode you get three include keyword fields (all optional, all comma-separated, all case-insensitive substring matches):

  • DB include keywords — leave blank to scan every database on the server; or list keywords to only include databases whose name contains one of them.
  • Schema include keywords — same idea, scoped to schemas within each matched database.
  • Table include keywords — same idea, scoped to tables within each matched schema.
⚠️

These are substring matches, not exact-match and not "starts with". A keyword like invoice will also match a table called invoice_archive or old_invoices. If you only want a specific, exact set of tables, list them all explicitly and double-check none of your other (unwanted) table names happens to contain one of them as a substring.

How this works, step by step:

  1. The agent connects to the server and lists every database it can see.
  2. It keeps only the databases matching DB include keywords (or all of them, if that field is blank).
  3. For each matching database, it lists all schemas and keeps only the ones matching Schema include keywords.
  4. For each matching schema, it lists all tables (and views too, if "Include views" is on) and keeps only the ones matching Table include keywords.
  5. It combines the matched tables across every database/schema and applies the Merge/Separate setting below to decide how same-named tables land in the target.
  6. Newly discovered tables are added to this pipeline's table mappings automatically — nothing needs to be added by hand.

Also in this mode:

  • Merge all source tables based on name — same-named tables across different databases get merged into one target table, with the source DB/schema added as extra columns. This is what Expert/M, Adsolut etc. use, since they create one DB per customer/dossier but the table structure is identical across dossiers.
  • Create separate tables for each source DB — keeps every source database's tables completely separate in the target, with the DB/schema prepended to the table name instead.

Target tab — where the data goes

Get the connection details from your Peliqan account, under Connections → your data warehouse's connection card. For additional security, ask Peliqan support for a dedicated DB user scoped to write access on just one target schema, rather than reusing a shared admin user.

image

Postgres datawarehouse

Target tab configured for Postgres
Target tab configured for Postgres

Get the connection details from your Peliqan account, under Connections → your data warehouse's connection card.

Field
What to put
Hostname / Port / Username / Password
From your Peliqan connection card
SSL mode
Usually Prefer
Database name
The Postgres database (from your connection card)
Schema name
The schema this pipeline writes to, must be unique per pipeline (see the warning below)
Table name prefix
Optional — prepended to every target table name

BigQuery datawarehouse

Get the connection details from your Peliqan account, under Connections → your data warehouse's connection card.

Target tab configured for BigQuery
Target tab configured for BigQuery
Field
What to put
Service-account key (JSON)
Paste the full JSON key. It's hidden by default once saved — click into the box or "Show" to reveal/replace it.
Region
The BigQuery dataset's region, e.g. EU
Project
The GCP project ID
Dataset ID
The BigQuery dataset, must be unique per pipeline (see warning below)

Settings tab — schedule and options

Settings tab
Settings tab
Setting
What it does
Sync frequency
Hourly, every N hours, daily, or weekly
Sync start time (UTC)
Anchor time for daily/weekly schedules, in UTC, not converted from your local time
Scheduled sync enabled
Turns the scheduled runs on/off. "Sync now" always works regardless of this setting.
Push config.json to target
Writes a redacted copy of this pipeline's config to a __config table on the target, tagged with the pipeline's id and name, so you can audit what's configured for a given pipeline over time. Secrets (passwords, keys, connection strings) are always replaced with *** before this is written.
Push logs to target
Writes per-run log lines to a __logs table on the target

Sync now / Stop trigger or cancel a run for this pipeline immediately, independent of the schedule.

Logs tab — see what happened

Logs tab showing a completed run
Logs tab showing a completed run

Every run adds log lines as it progresses. Filter by free-text search or toggle Errors only. Open log folder jumps straight to the raw log files on disk if you need to send them to support.

Manage pipelines tab — running more than one pipeline

Manage pipelines tab with two active pipelines
Manage pipelines tab with two active pipelines

Each pipeline is one source configuration → to one target schema, with its own schedule and settings. Click + Add pipeline to create a new one, or use the ⋮ menu on a row to rename, activate/deactivate, or delete it.

  • Rename and Delete are both blocked while that specific pipeline is actively syncing — stop it first.
  • Active / Inactive reuses the same "Scheduled sync enabled" flag from the Settings tab — it's the same switch, shown in two places for convenience.
  • Clicking a row switches you to that pipeline everywhere else in the app (Source/Target/Settings/Logs all follow whichever pipeline is currently selected here).

Don't reuse the same schema/dataset across pipelines

⚠️ Do not configure two or more pipelines to write to the same target server + database + schema (or BigQuery project + dataset), because they will silently overwrite each other's metadata and data. The Target tab warns you about this automatically: if another pipeline already uses the same target, you'll see:

Warning: This target schema is already used in another pipeline. Set a unique schema for each pipeline!

This is a warning only — it won't block you from saving — but treat it as something to fix immediately, not something to ignore. Give each pipeline its own schema/dataset.

Advanced: setting up many pipelines at once by editing config.json directly

If you need to set up a large number of near-identical pipelines (e.g. one per end-customer dossier, each just pointing at a different DSN with a different target schema), clicking through the UI once per pipeline doesn't scale well. There's a faster path:

The agent's entire configuration is one plain JSON file:

C:\ProgramData\PeliqanSatellite\config.json

Its shape is a list of pipelines, each with source, target, settings, and table_mappings:

A few things that make this genuinely practical, not just theoretically possible:

  • Passwords can be plain text. The agent auto-encrypts every secret field the next time anything saves that pipeline (via the UI, or via its own first sync run) — you don't need to pre-encrypt anything yourself.
  • No new pipelines need table_mappings filled in. Leave it as [] — the agent discovers the schema automatically on first sync, exactly like a pipeline created through the UI.
  • The target schema/dataset gets created automatically on first sync if it doesn't already exist — no manual DB setup needed on the target side either.
  • Malformed JSON fails safely. If the file doesn't parse, the agent backs up the bad file (as config.json.corrupt-<timestamp>) and shows a clear error instead of silently wiping your configuration.
  • What it won't catch: valid JSON with a typo'd value (e.g. a wrong hostname) loads without any error — it only surfaces when that specific pipeline's first sync actually runs and tries to connect. Spot-check a few pipelines after a bulk import rather than assuming everything is correct just because the agent started cleanly.

To apply: stop the Peliqan Satellite service, replace config.json, start it again (or just reopen the desktop app — either picks up the file fresh). Always keep a backup of the previous file until you've confirmed the new one works.

FAQ

What are the alternatives to this for on-prem connectivity?

A VPN connection (e.g. Site to Site VPN) is the main alternative.

Can Peliqan Satellite write to the on-prem database?

No — it only reads data (SELECT only) and sends it to your Peliqan cloud data warehouse.

Do I need to open my firewall?

Only outbound: the agent needs to reach the Peliqan cloud data warehouse (Postgres or BigQuery). No inbound connections from the public internet are required.

Can it conflict with other tools using the same database (e.g. backup software)?

No — it only ever runs SELECT queries and never takes locks on tables.

Can I have two pipelines write to the same target at once?

No — each pipeline has exactly one source and one target; there's no such thing as two simultaneous targets for one pipeline. If you need the same data in two places, or two different sources feeding the same kind of target, that's two separate pipelines — just make sure they don't share the same target schema/dataset (see the warning above).

Current limitations

  • No incremental sync — every run does a full truncate + reload.
  • Supported sources: SQL Server and ODBC (covers on-prem accounting/ERP software such as Expert/M, Adsolut, Plenion, Exact Venice).
  • Supported targets: Postgres and BigQuery.

Releases

v1.5: various improvements

  • Warning on using same schema in 2 pipelines.
  • Improved logging of config (keep history of config).

v1.4: partner target configuration added

  • Added support to configure a target for specific partners using a code.

v1.3: Multiple pipelines and faster syncs at large scale

Performance improvements for large multi-database syncs:

  • Discovering and mapping tables now happens in one batch per source database instead of per table.
  • Copying data now streams each table over a single connection instead of reconnecting for every batch, which improves speed.

Reliability & visibility:

  • The sync log now streams to the target's log table incrementally during table discovery, not just at the end — so a long-running sync is visible in progress instead of appearing silent for hours.
  • Fixed a bug where a table that failed during discovery could make the whole run silently report "0 failed" even though something genuinely went wrong.

v1.2: Generic ODBC source support

  • Added support for connecting to source DBs using ODBC (using DSN or DSN-less connection string), used for e.g. Actian Zen DB (Exact Venice).
  • Fixed an issue affecting sync timing accuracy and a BigQuery key-field handling bug.
  • Improved memory handling (avoid out of memory error).

v1.1: Support for composite keys

  • Support added for composite keys
  • UI improvements

v1.0: Initial release

  • Support for SQL Server as source DB
  • Support for Postgres and Bigquery as target DWH

v0.x: Beta versions