June 15, 2026 at 12:00 am
Comments posted to this topic are about the item The Slow Growing Problems
June 15, 2026 at 7:49 am
I'm watching a serverless cloud data warehouse issue at the moment, where the data is growing naturally, and the costs are slowly creeping up because the time to process is creeping up.
We could choose to use a bigger warehouse size. The cost calculation isn't straightforward because the warehouse bills by the minute. A larger instance will cost more per minute, but the number of minutes consumed will drop. It isn't easy to predict what will actually happen to costs. Businesses like to know exactly what will happen when it comes to money.
We have several options we could pursue instead; however, business priorities are elsewhere because the system works at present. It is hard to get preventative work prioritised.
June 16, 2026 at 1:46 am
It is. Usually everyone wants to wait until its an emergency.
Which is never good timing.
June 17, 2026 at 8:52 am
Hear, hear as to emergencies being inconvenient times to fix stuff.
If it ain't broke, there's less motivation to fix it. But a stitch in time saves nine.
Maybe a bit of judicious extrapolation, with currency signs attached, would focus people's attention.
And chuck a few suitable phrases in there -- "proactive", "forward-looking", etc. I don't know -- people are perverse.
Good success. May we all come out smiling.
MarkD
June 18, 2026 at 5:09 pm
Hear, hear as to emergencies being inconvenient times to fix stuff. If it ain't broke, there's less motivation to fix it. But a stitch in time saves nine. Maybe a bit of judicious extrapolation, with currency signs attached, would focus people's attention. And chuck a few suitable phrases in there -- "proactive", "forward-looking", etc. I don't know -- people are perverse. Good success. May we all come out smiling. MarkD
Preach!
June 21, 2026 at 2:05 pm
We have a situation like this at work. We have a third-party product that is used by several people throughout the state to help them know what to do each day, how to perform treatment of patients, etc. But there are a lot of moving parts. Patient data from about the state gets sent to a business who collates and forwards it onto us daily. Then a Rhapsody job runs on our servers to take the data sent to us and move it to a location where a SQL Server SQL Job runs daily to process and load it into a SQL Server database, where the application can use it. For the most part this all works fine, but if anything along the pipeline gets mucked up, I'll hear about it by 8 AM. The users get pissed, fast. And then tend to blame the application they use as the culprit, but I see it as multiple software systems loosely connected where anything can go wrong. Occasionally the users will begin to think that replacing the third-party application will solve all their problems, but I don't see it that way. One thing that bothers me about this whole process is I don't think anyone has a deep understanding of the whole process from beginning to end. The most recent call to scrap the third-party app involved our rewriting the application. But I believe my management have no idea what we'd be getting ourselves into. The Federal government will sometimes change the requirements and regulations dictation what they will accept. And since a significant amount of funding comes through Federal grants to report to the Feds, that this software helps us generate those reports. My users and managers take the naive point of view that if we rewrote the application then it won't need to be changed for 10+ years. That simply isn't true. The Feds always make some change every few years. The vendor does a good job of keeping on top of the Fed's change and adjust their application accordingly. Fortunately, I think this demand is no longer being discussed.
I really like your idea, Steve, of an AI keeping track of all the moving parts, everywhere in the system! It makes a lot of sense. The only thing is that AI would have to have the permissions of all entities involved in order to do that kind of monitoring. I know that all I have access to is the vendor's application, not Rhapsody which is run in house, and certainly not whatever system(s) all the other parts of the system use to get the data to us.
Kindest Regards, Rod Connect with me on LinkedIn.
June 22, 2026 at 4:22 pm
Be thoughtful, be careful, be deliberate.
Analysis should include all of these, ending with a decision you've made (to do, or not to do something), not deferring things.
I find a lot of people want to make an amazing decision for the future, but they haven't really done the first two above. They just want problems to go away.
Viewing 7 posts - 1 through 7 (of 7 total)
You must be logged in to reply to this topic. Login to reply