Compare

The honest comparison

There are four realistic ways to work with data in InterSystems IRIS or Caché. Here is what each is good at — including the cases where we are not the right answer.

InterSystems Management Portal

The administrative interface that ships with IRIS. Unbeatable for security, configuration and operations. Not built for daily SQL work or for exploring data — the SQL page is a text area, not an editor.

Use it for: administering the instance.

DBeaver & DataGrip

Very good generic SQL clients that reach IRIS over JDBC. They see the SQL surface and nothing beneath it: no globals, no namespace overview, no cached queries, no IRIS-specific diagnostics. Cross-platform, which matters if you are not on Windows.

Use them for: many database systems at once.

Feature by feature

“Generic client” below means DBeaver or DataGrip connected to IRIS over JDBC. Rows marked with a dot are the ones that usually decide the question.

  Management Portal Generic client SQL DATA LENS
Browse globals — subscripts, $LISTBUILD values, mixed types A basic viewer Not at all Visual browser with $LISTBUILD resolution
SQL editor with IRIS-aware completion Minimal editor Generic SQL completion IRIS SQL grammar and completion
One statement across namespaces and servers No No Query Cloud, joined client-side
Cached query analysis and cleanup Manual, one screen at a time No Listed by cost, deletable
Redundant index and unindexed FK scanner No No Scans a namespace, groups findings
Execution plans Yes Generic, where the driver exposes it In the editor, IRIS-native
Tune Table statistics Yes No Yes, including cached-query recompile
Processes, locks, tasks, logs, CPF Yes — this is what it is for No Read-focused views next to your SQL
Security, user administration, roles, mirroring, gateways Yes — the right tool No Yes — users, roles and gateways are shown
Several IRIS/Caché driver versions at once Not applicable Manual driver setup per version 8 bundled, isolated per connection
Copy tables between namespaces and servers No Limited, via generic import/export Data Pumper
Class and routine source, including INT code Partly No Yes, and copy between servers
ER diagram from an existing schema No Yes Yes, stored in the namespace
Other database systems No Very broad — dozens of systems Solid JDBC set, plus DuckDB and CSV files
Query CSV, Parquet and JSON files directly No Varies Drag and drop onto DuckDB
Platforms Browser — any OS Windows, macOS, Linux Windows only today
Installed on the database server Part of IRIS Nothing Nothing
Price Included with IRIS Free to roughly $270 per year Free Edition, or $299 per user/year

When we would tell you to use something else

A comparison page that finds its own product best at everything is worth nothing. Three cases where the answer is genuinely not us:

  • You need a macOS or Linux desktop client today. Use DBeaver. A macOS build is on our roadmap but does not exist yet, and pretending otherwise would waste your afternoon.
  • You work across thirty database systems and IRIS is a footnote. DBeaver or DataGrip will serve you better. Our generic JDBC support is solid, not encyclopaedic — see Beyond IRIS for exactly what it covers.
  • Your task is administration — users, roles, mirroring, gateways, backups. The Management Portal, every time. We deliberately do not duplicate it.

And one case where the answer is clearly us: your work happens inside IRIS or Caché, and the questions you keep hitting are what is actually stored in this global, why did this query get slow, and how do I join data that lives in three namespaces. That is the whole design brief.

Questions people actually ask

Can I not just use DBeaver with the IRIS JDBC driver?

You can, and plenty of people do — it works for ordinary SELECT and INSERT work. What you do not get is anything IRIS-specific: no globals, no namespace tree, no cached queries, no CPF view, no Tune Table, no index scanner, and driver versions become your problem to manage. If your work is 95 % plain SQL against one namespace, DBeaver is fine. If your work involves asking why something is slow or what is really stored in a global, it stops being enough.

Does this replace the Management Portal?

No, and it is not trying to. Security, mirroring, backups and the deeper admin workflows remain the Management Portal's job. SQL DATA LENS does show users, roles and gateways, but it is mainly for daily data work, diagnosis and SQL analysis. Most users keep both open.

Does it replace VS Code with the InterSystems extensions?

No. VS Code is where you write and debug ObjectScript. SQL DATA LENS is where you look at data, run SQL and diagnose performance. They overlap barely at all, which is why the two sit side by side on most of our users' machines.

We already pay for DataGrip. Why add another licence?

Because the two answer different questions. DataGrip is an excellent general SQL IDE; it has no concept of a global, a namespace or a cached query. The honest test is to install the free trial, open your slowest namespace, and see whether the index scanner and cached-query list tell you something DataGrip cannot. If they do not, keep DataGrip and save the money.

Settle it against your own database

A full 30-day Pro trial, no registration, nothing installed on the server. Open your slowest namespace and see what the scanners find.

Windows 10, 11 and Windows Server (64-bit) · ~140 MB · version 3.24 · full 30-day Pro trial included