A Tale of Two Worlds

I continue to support clients who use SSIS for enterprise data engineering. Some of the enterprises are small-ish by comparison. Others are huge. Some friends also continue to support these clients and clients like them. That’s World 1.

World 2 is social media. I have some friends on social media. Many I know IRL (in real life). Most, though? Most are acquaintances. Most of my interactions on social media are with people I don’t really know or don’t know that well. Conversely, many of them – of you – don’t know me that well, either. I touched on this in a recent newsletter / post titled 2025: A Number Containing Two 2’s, One 5, and a 0.

Some Dissonance

• Although SSIS is older than many data engineering platforms, making it mature, it has held up well
• SSIS remains a viable enterprise data engineering platform, even now in 2026
• SSIS has strengths other, newer platforms lack; such as extensibility and a thriving third-party controls market

Many of the World 2 individuals with whom I interact have experience developing, supporting, and maintaining SSIS-base enterprise data engineering. Some of them have more experience, some less. I rarely know what experiences led them to perform enterprise data engineering or what led them to form the beliefs they hold about SSIS. I’ve stated many times over the years: “People believe what they believe for a reason.” Please hear me. I believe many have good reasons for disagreeing with me and my World 1 peers. You may be surprised to learn that, whenever World 2 people share their reason for disagreeing with me (and my World 1 peers), I most often agree with them, at least in principle.

Sometimes, I’m able to share something of which they’re unaware – some best practice or design pattern or, occasionally, an anti-pattern.

An Early Exchange

Advice: try to avoid writing more than one book at the same time.

If memory also serves, one of this gifted software developer’s complaints was about source control. In that first book, I authored two chapters. And durned if one of them wasn’t about using source control with SSIS. It’s prudent to pause here long enough to disclaim that SSIS never played well with many concepts included in Application Lifecycle Management (ALM). And that not-playing-well largely holds to this day.

I can hear some of you thinking, “In what ways does SSIS not play well with ALM, Andy?” That’s a great question. I’m glad you asked! You’re not the first person to ask, either. In fact, I was inspired to write this newsletter at this time because a dear friend and brother from another mother reached out to me to discuss a couple of SSIS’ quirks that interfere with practicing ALM in an enterprise that employs SSIS for enterprise data engineering.

I went back and forth with the gifted software gentleman, arguing that I agreed with a few of his points. I believe he took offense when I shared that I taught SSIS courses and that every. single. one. of my students knew how to manage the majority of the items he hated about SSIS. I went on to point out that I was working my way up to becoming a n00b in the software platform in which he specialized. I don’t think that part bothered him. Learning more from a master in that language was, in fact, why I followed him on social media to start with. I think what bugged him was when I stated that I would never write a list of things I hated about his platform of expertise unless and until I felt I knew said platform well enough. I may have also gone on to imply that not knowing 8 of 10 things all my students knew may qualify him as not knowing enough to hate them… memory fails me at this juncture.

I remember thinking that, though, and I know the Bible teaches a man is as he thinks in his heart. So, busted.

A More Recent Exchange

During the back and forth, he made several statements:

• SSIS is too old and outdated
• He didn’t understand why SSIS wasn’t deprecated
• He’d never actually deployed an SSIS package to Production. He was basing his opinion on what a friend who had deployed SSIS to Production told him

If memory serves, I covered why I disagreed with his opinion of SSIS, and with his stated reasons for his beliefs. My arguments (in brief) are:

• COBOL is older and touches many (if not most) of the financial transactions on Earth
• Years ago, I had a conversation with THE Buck Woody in which I asked Mr. Woody, “If shrinking a SQL Server database is so horrible, why does Microsoft still support the functionality in SQL Server?” I’ll never forget Buck’s response: “Because sometimes shrinking the database is the only viable option.”
• (I didn’t respond to this point. This is where I decided to stop investing time.)

The overwhelming majority of the World 2 people I interact with aren’t this… animated. Most have at least some idea how data engineering – and data engineering with SSIS – works.

So, It’s 2026…

• “We have thousands of SSIS packages and no reliable way to automate migration to a different platform with confidence all our use cases will be covered”
• “SSIS Just Works and remains supported by Microsoft; why should we change?”
• “We do not trust the cloud”
• “Migrating to the cloud will not save us money”

There are more use cases (don’t get me started on ‘just migrate the SSISDB database to the cloud’… even though that’s sometimes viable), but this newsletter is long enough already.

If you read those four use cases and balked, please leave a comment and share why. I know there are solutions out there for some of my clients who would like to migrate away from SSIS. I’m being serious here. Please share your thoughts about solutions for them (and others).

For some of my clients interested in migrating away from SSIS, rewriting their enterprise data engineering manually is not an attractive or inexpensive proposition. And, in my opinion, those previous bullets make the most sense for at least some of them.

What I’m Doing About It

I built Data Integration Lifecycle Management Suite. DILM Suite was built to support the SSIS lifecycle. Did I solve all the issues? Did I address all the SSIS quirks? No and no. I addressed many of them, though, and Kent Bradshaw and I continue to develop tools and utilities to address more.

In 2026…

“When will they be done, Andy?”
“They’ll be done when they pass all my tests.”

I’m also going to write more about SSIS.

And there’s stuff I’m not prepared to share at the moment…

A Request