When building a Production, should I create separate message classes for each integration flow, or is it acceptable to reuse generic request/response classes across different Business Operations? I'm trying to understand how to keep things organized as the number of integrations grows.

What is the recommended way to handle errors inside a Business Process in IRIS? Should I use Try/Catch within the BPL, return error responses to the caller, or rely on the built-in retry mechanism of the Production? Looking for guidance on what's considered good practice.

What is the recommended approach for handling upgrades in an InterSystems IRIS Kubernetes environment?

For example, if we deploy version 1.0.0 of our product and subsequently need to upgrade to 1.0.1, and this upgrade requires changes to SQL tables containing customer data.

The quickest solution that comes to mind is creating an 'upgrade method' that runs on startup to check if any data migration actions are required. However, I'm wondering if there are better solutions or established best practices for this.

Thanks in advance!

What's the most straight-forward way to install this on an offline server? I'm trying to set this up on an Azure DevOps server to support our CI/CD pipelines. I've tried using zpm installing the tgz from the local filesystem. I note zpm seems to need a repo configured to install but I can't work out how to setup a bare-bones Filesystem repo (please point me to some documentation on this). I have no idea what I'm doing...

Hello Community,  

What is the most efficient, memory-safe way to get the names of the corrupted indexes on very large tables for a rebuild. However, if an index has millions of corrupted rows, the .errors array in %ValidateIndices grows too large and throws a errorerror.

Hi!
We are deploying the iris image in a Kubernetes environment and the cluster state is "Hung" , looking the alerts endpoint we get 2 alerts:

[
  {
     "time":"2026-03-24T13:45:44.548Z",
     "severity":"2",
     "message":"System appears to have failed over from node a69a9f137593"
  },
  {
     "time":"2026-03-24T13:46:30.274Z",
     "severity":"2",
     "message":"Error: <PROTECT>KillAlive+1^%SYS.CDIRECT in SERVERS"
  }
]
Any idea / help where those are comming from and how to address them?

I recently started using Cursor/VSCode with an IRIS container for development rather than Studio/Terminal.  
I've noticed that whenever I use %G (so basically all the time), when I exit %G, the terminal window simply closes, rather than returning me to my usual namespace prompt.  
%G also does not retain the command stack like it does in old school terminal, so I'm forced to constantly retype every global reference.  
Anyone figured out a solution to this?  It's a relatively minor problem in the grand scheme of things, but a time consuming and irritating one.  
 

When developing Python applications with InterSystems IRIS, you can quickly end up with several execution contexts:

• Python launched directly by IRIS with Embedded Python;
• a regular python3 process that loads the Embedded Python libraries from a local IRIS installation;
• an external Python application that connects to IRIS through the official native driver.

These three cases are useful, but they do not behave exactly the same way for imports, system configuration, object APIs, and SQL access.

Redoc is an Open Source solution capable of rendering API specifications in OpenAPI 2.0 or 3.0+ as very beautiful and functional web portals. Currently, to have something similar, we need the ZPM SwaggerUI extension or we need to install IAM - InterSystems API Manager and then configure the IAM Developer Portal. Well, now the community has one more option, iris-redoc. This solution installs a web application on your IRIS instance that uses Redoc to present a beautiful web portal for your REST APIs:

For those who learned Caché ObjectScript from scratch: what kind of personal or practice projects did you build to get comfortable with the language? I come from a C# background and I'm looking for project ideas that are small enough to be feasible but meaningful enough to actually teach the core COS concepts.

Hello Community!

We’re excited to invite you to our upcoming Hebrew webinar, presented by @Keren Skubach, an InterSystems Senior Sales Engineer:

👉A Practical Guide to Leveraging AWS Adapters and APIs 👈

📅 Date & time: June 2nd, 3:00 PM IDT

Discover how to accelerate cloud-based integration with InterSystems' native AWS adapters for S3, SQS, SNS, and CloudWatch.

This session provides a practical look at building modern interoperability workflows — from secure file ingestion and asynchronous messaging to automated notifications and centralized monitoring.

Recently, a question on the Community was asked by @Vermon Ferre about storing data from inherited classes in different globals. So, I decided to simulate the following behavior: I created a superclass called Article.MainClass and two subclasses, Article.Class1 and Article.Class2. By default, when each class extends %Persistent IRIS creates independent storage structures for them. This will work as intended if the first class in the list of superclasses is %Persistent. But it also means that if there are any parameters in the main class, they will be lost, because only parameters from the first class in the list get inherited.

IRIS 2026.1 provided Partitioned Tables as a new option for large data sets
It's a great improvement as it offers standardization of this feature.

Though:
It was possible also before, matching requirements and leaving room for creativity.
less elegant with a bit more coding and fewer automatisms.

Introduction

Today, coding assistants like Claude, GitHub Copilot and Cursor have transformed the way developers write code. However, these tools are limited by being isolated from the systems and data sources that developers work with daily. This limitation can be overcome through the Model Context Protocol (MCP), an open standard designed to connect AI assistants to external data sources and tools in a secure and standardized way.

In this review article, we'll explore the current state-of-the-art regarding the MCP within the InterSystems ecosystem.

1. VS Code version 1.60.0 and above

2. install below 3 extensions

   

3. Ctrl + Shift + P show all commands

4. InterSystems: Connect to Server

5. input connection details 

   IP：localhost

        Port: 52773

       choose global & http connection method

       Input username and password of Intersystems management portal

       screenshot for success connection:

This article presents a straightforward approach to automatically and efficiently tune hyperparameters for machine learning models using Optuna as the optimisation framework. We explore how to use both Optuna’s native storage options and InterSystems IRIS as a database backend to track the progress of hyperparameter searches. We also show how MLflow can be used to monitor experiments and manage models through its tracking and model registry UI.

This article is based on this Kaggle Notebook, which you can run and directly edit yourself.

When managing critical healthcare data through an integration engine you want to know the moment a queue starts backing up or a service drops. 

Unfortunately, this usually leads to an email inbox stuffed full of notifications that can sometimes seem impossible to maintain.

Well, I’ve not solved that problem. 

But I have created something that could help...

Key Responsibilities & Requirements

Core Infrastructure & Cloud Management

• Full-Stack Administration: Maintain and optimize both Linux (Ubuntu/RHEL) and Windows Server environments.
• Cloud Orchestration: Take full ownership of VM cloud infrastructure, including provisioning, health monitoring, backup and resource scaling.
• Policy Governance: Design and enforce comprehensive SysAdmin and Deployment policies (CI/CD pipelines, automated provisioning, and security hardening).

InterSystems Database Administration

• Platform Expertise: Expert-level administration of both InterSystems IRIS and Caché.

Data privacy regulations such as GDPR, LGPD, and HIPAA demand that organizations know exactly where Personally Identifiable Information (PII) lives inside their databases. Yet in practice, most teams rely on manual inventories, tribal knowledge, or external scanning tools that require data to leave the database engine — a process that itself creates privacy and security risks.

This article presents an MVP that takes a different approach: it runs PII detection inside InterSystems IRIS using Embedded Python, analyzing data where it lives and never exporting it to an external process.

InterSystems IRIS globals are one of the platform's core strengths: they store hierarchical data in a direct, ordered, and efficient structure. But when working from Python, manipulating globals can sometimes feel closer to a low-level API than to the natural habits of the language.

The iris-global-reference project provides a Python layer on top of IRIS globals. Its goal is simple: make access to globals more readable, more idiomatic, and easier to integrate into modern Python code, without hiding the underlying hierarchical model.

What is wrong with the *inc file code below:

#define ArrayToString(%array,%out,%del)      set %out="" for { ##continue
                                                                                   set key=$order(%array("")) quit:key="" ##continue
                                                                                   set %out=%out_%del_key ##continue
                                                                               } quit

Compilation at calling classmethod as : Set tQStr = $$$ArrayToString(pQArray,tQStr,"##") brings error as: