Oracle has a horrible reputation among devs. But I think they bypass devs for purchase decisions and straight up wine and dine and bamboozle the low information execs.
When I worked at a megacorp as a dev, I had near zero say in such purchase decisions. I had to work with what I was given. Thankfully I work for a much smaller shop now. Better pay and much better decision autonomy.
>Oracle has a horrible reputation among devs. But I think they bypass devs for purchase decisions and straight up wine and dine and bamboozle the low information execs.
That is only partially true. Oracle has a wide portfolio of a bunch of products and the "wine & dine the execs" is the sales cycle for software like ERP Oracle E-Business Suite and Oracle Health (Cerner). E.g. it's the hospital CFO & CIO that are the true "customers" of Oracle Health and not the frontline doctors and nurses that use it.
However, for the Oracle RDBMS database ... it was often the developers that chose it over competitors such as IBM DB2, Sybase, Informix, MS SQL Server.
In the late 1980s and early 1990s, a lot of us devs did a "bake off" with various databases and Oracle often won on technical merit. E.g. Oracle had true row-level locking but MS SQL Server before v6.5 was page-level locking. And the open source alternatives of MySQL and PostgreSQL at that early timeframe were not mature enough to compete with advanced Oracle features such as Parallel Query Execution and Recovery from a Standby database.
So young devs today who aren't aware of history will wonder why Amazon ever got locked into Oracle in the first place?!? It was because in 1994, Oracle db was a very reasonable technical choice for devs to make.
Devs get Oracle's best side. They don't know how lucky they are. PL/SQL is a fine language and Oracle throws everything and the kitchen sink into the APIs available inside the database.
As a sysadmin/dba I get to handle the nasty side of Oracle: the bugs, the patches that fail, the redundant tools, the wordy documentation that always feels like a never-ending advert.
The problem is that putting code into your database has all the problems of data migration with no testability. It's really hard to write tests and debug code compared to a proper language.
When I was young and beautiful and naive I enjoyed playing with setting things up, even though my main focus was on backend dev. Like, setup postgres, tweak settings to performance, that kind of thing. So I was the first one to get the "figure out on site installation steps" task to share with client admin.
Our small company was building enterprises things for large clients. And one of them wanted Oracle on Windows server. They also wanted a failover setup. How hard could that be?
Now I hate Oracle. I hate Oracle consultants. I despise the ignorance.
My favourite bit was Oracle uninstall procedure. I had like 4 pages printed just for this case.
Oracle is in the small family of companies that have a business model of "you give us money and we solve _all_ your problems."
So it's likely that, for a specific task, the oracle solution is crap, but oracle has crap for everything so the oracle sales droid can sell a "one throat to choke, one check to write" policy to a company that likely has technology problems but produces a "not technology" product.
Amazon's moving into this space and their crap is even stickier than oracle's...
Is this the same Oracle/Cerner system ("Millennium", I believe) that, despite protests from the medical staff, was deployed "big-bang style" in the Swedish region of Västra Götaland, with much the same results? (And where during a press conference, where the management was explaining how nearly everything was going to plan, a doctor, who had somehow sneaked in, got up, shouted something to the effect of "you’re lying, it’s a bloody disaster!" and stormed out.)
(That was not an outage, though — as far as I understand the system was working, it just didn’t actually work...)
Swedish privacy laws for EHRs are quite different compared to other countries.
So I'm not surprised any software not purpose built for for that market would fail. And we know Oracle is not investing in R&D for Millennium since all efforts are out into their forthcoming AI based EHR - whatever that actually means....
Could be, but oracle only bought cerner a few years ago and millennium has been around much longer. Other EMR systems are also dreadful - my spouse has some stories about EPIC for example.
I just watched a YC video about vibe coding. He says he used Claude Sonnet 3.7 to configure his DNS servers & Heroku hosting. What could possibly go wrong?!
I managed to break production before vibe coding was cool.
20 years old me had rm -fr root access to way too many systems.
I don't think it's much different today.
If anything, I think the youngsters will learn faster from their mistakes because they already have good mentor for the easy stuff and will get grinded on the hard stuff sooner.
How do we know that it was just a "human error" as the article/headline implies?
Answer: we do not know either, but this is the standard response so that companies (or governments or whoever this concerns) are absolved of any responsibility. In general, it is not possible to know for a specific case until a proper investigation is carries out (which is rarely the case). But most of the times, experience says that it is company policies/procedures that either create circumstances that make such errors more probable, or simply allow these errors to happen too easily due to lack of enough guardrails. And usually it is due to push for "shipping products/features fast" or similar with little concern to other regards.
It could be a different case here, but seeing it is about oracle and having in mind that oracle is really bad at taking accountability about anything going wrong, I doubt it (very recently they were denying a hack on their systems and the leak of data of companies that use them until the very last moment).
Title should be "Oracle execs caused five days software outage at US hospitals". If the systems helped save lots of life, you bet the engineers wouldn't be thanked for that and it probably wouldn't even be a news article
Oracle offers everything. Databases, web frameworks, programming languages, CRMs, cloud management tools, auto scaling clusters, identity management, BI, document search, email servers, colocation, mainframes, calendar sync. You name it, they've got it. Oracle has products ready to go to run an entire country if it needs to.
Sure, Postgres beats OracleDB, but Postgres doesn't integrate as well with Oracle Fusion and you need to migrate the code yourself. They're like SAP: they're big enough that you can make a career just out of configuring their software packages and make good money while doing so.
It's expensive and certainly not the best, but it's reasonably stable and has a huge company backing it. Oracle won't disappear any time soon and they're not as likely as Google or Microsoft to shut down their services within a few years notice.
In some countries, Oracle is also very good at doing what Google and Microsoft are doing to students. The Brazilian programmers I've spoken to specifically learned OracleDB when they were taught relational databases. They learned to program in Java, and I'm sure Oracle also sponsored other parts of the professional tooling they got to use for free. Microsoft, on the other hand, didn't seem to generous towards their educational facility (no free MS tooling for their schools like they offer over here). If all you know is Photoshop/Windows/Maya/OracleDB/iOS, you're going to look for jobs where you can use Photoshop/Windows/Maya/OracleDB/iOS, and employers looking for cheap juniors will need to offer Photoshop/Windows/Maya/OracleDB/iOS to make the best use of them.
Are we sure? I'm by no means a DBA, but DBA at our company (who is freaking smart btw) said if money wasn't an issue, he would actually go with OracleDB.
Because Oracle comes to any country/industry with trucks of money to corrupt officials and lobby their adoption.
And after a few years you find yourself in a situation when you already paid for Oracle so much, integrated it so deeply, so switching to any alternative is a massive pain and in most cases it’s safer and easier to keep paying Oracle.
I believe this is how they grew and how they remain big. While smaller companies aim for managers, Oracle targets CTOs and CEOs, takes them out for expensive dinners and promises them the world. Then, they legacy handcuffed them forever. And virtually no one chooses to invest crazy amounts of money on migrating or even starting a new codebase when your company lifeblood has been poured into the Faustian Oracle machinery.
It's legacy lock in. You can't just switch stuff. it's incredibly hard to move logic out of store procedures from on DB to another. Its the same reason why the US government runs on Cobol. People don't think about the strategic implication of the tech they choose and how it may come back to limit them in the future.
Most Oracle "shops" i know have used it for decades. When they started using it, there weren't many options, so Oracle was what was used.
COBOL is in the same category. When invented, it was the absolute easiest programming language to learn and use, so of course it gained popularity.
It then turned out to be rather good at what it did, along with running on hardware that "never fails", so most places didn't even think about replacing it until sometime in the 90s.
Also keep in mind that the reason companies are migrating away from COBOL is not due to the language as much as due to young people not taking an interest in COBOL and Mainframes, making it hard to recruit new talent.
Even then, a migration away from a typical mainframe is a huge task. In most cases you're talking 50-60 years of "legacy" (still running, still working, still updated) code, and replacing a system that has evolved for half a century is not an easy undertaking, at least not if you plan on getting it 100% right the first time.
It's pretty common for business contracts to just end up stuck on mainframes or scattered across various systems from vendors like Oracle (the paper trails are often thrown away long time ago).
And let's be honest, a lot of folks in IT aren't exactly top performers and don't seem to care all that much. It's really the developers you find on forums like this who are genuinely passionate. You're not likely to bump into the people actually buying those big Oracle or IBM systems around here though :)
I know they get a lot of hate but I personally used Oracle’s cloud and it seemed like a decent piece of infrastructure with really solid engineering team behind it. They had their share of problems (and their first line of support is really really bad and not well suited to handle real issues) but not really any different from any other similar products.
I made a similar experience. The database just stopped working one day, triggered by a normal restart without any updates or patches. Fixed the configuration, tested it, restarted multiple times, everything works. After a week it was not working again, without me being able to fix it except doing a fresh reinstall on a clean VM.
Luckily more and more customers switched to Postgres and I no longer have to deal with it.
Used to work on CHS systems for Cerner. If I am not mistaken they were a "communityworks" client, which meant their databases were shared with a number of other clients - a "multitenent" envirnment, we called it. Completely bumblefucked design. Not surprised somthing like this happened.
Also - cerner software in general allowed hospitals to freely completely fuck up their own architecture. Sometimes ireperably.
If anyone has details about how this happened Id love to hear.
Readit News
When I worked at a megacorp as a dev, I had near zero say in such purchase decisions. I had to work with what I was given. Thankfully I work for a much smaller shop now. Better pay and much better decision autonomy.
That is only partially true. Oracle has a wide portfolio of a bunch of products and the "wine & dine the execs" is the sales cycle for software like ERP Oracle E-Business Suite and Oracle Health (Cerner). E.g. it's the hospital CFO & CIO that are the true "customers" of Oracle Health and not the frontline doctors and nurses that use it.
However, for the Oracle RDBMS database ... it was often the developers that chose it over competitors such as IBM DB2, Sybase, Informix, MS SQL Server.
In the late 1980s and early 1990s, a lot of us devs did a "bake off" with various databases and Oracle often won on technical merit. E.g. Oracle had true row-level locking but MS SQL Server before v6.5 was page-level locking. And the open source alternatives of MySQL and PostgreSQL at that early timeframe were not mature enough to compete with advanced Oracle features such as Parallel Query Execution and Recovery from a Standby database.
E.g. the C Language programmer Shel Kaphan at Amazon chose Oracle in 1994: https://www.linkedin.com/posts/jpcharles_in-1994-amazons-fir...
(that anecdote cited this deep link: https://www.youtube.com/watch?v=u3qIWN-ZIPk&t=1h11m56s)
It took Amazon 25 years to finally migrate off of all Oracle databases: https://www.google.com/search?q=oracle+shuts+off+last+oracle...
So young devs today who aren't aware of history will wonder why Amazon ever got locked into Oracle in the first place?!? It was because in 1994, Oracle db was a very reasonable technical choice for devs to make.
As a sysadmin/dba I get to handle the nasty side of Oracle: the bugs, the patches that fail, the redundant tools, the wordy documentation that always feels like a never-ending advert.
Our small company was building enterprises things for large clients. And one of them wanted Oracle on Windows server. They also wanted a failover setup. How hard could that be?
Now I hate Oracle. I hate Oracle consultants. I despise the ignorance.
My favourite bit was Oracle uninstall procedure. I had like 4 pages printed just for this case.
Oracle bad. Postgres good.
So it's likely that, for a specific task, the oracle solution is crap, but oracle has crap for everything so the oracle sales droid can sell a "one throat to choke, one check to write" policy to a company that likely has technology problems but produces a "not technology" product.
Amazon's moving into this space and their crap is even stickier than oracle's...
(That was not an outage, though — as far as I understand the system was working, it just didn’t actually work...)
So I'm not surprised any software not purpose built for for that market would fail. And we know Oracle is not investing in R&D for Millennium since all efforts are out into their forthcoming AI based EHR - whatever that actually means....
Deleted Comment
Must be engineers who write requirements and unrealistic deadlines that lead to such issues.
"How To Get The Most Out Of Vibe Coding | Startup School " => https://www.youtube.com/watch?v=BJjsfNO5JTo&t=494s
I like his advice on downloading relevant documentation and putting it in the code base. That makes sense to me for targeted use cases.
20 years old me had rm -fr root access to way too many systems.
I don't think it's much different today.
If anything, I think the youngsters will learn faster from their mistakes because they already have good mentor for the easy stuff and will get grinded on the hard stuff sooner.
Sometimes people make mistakes, sometimes people are incompetent, sometimes managers suck, sometimes it's a multi-layered issue.
Answer: we do not know either, but this is the standard response so that companies (or governments or whoever this concerns) are absolved of any responsibility. In general, it is not possible to know for a specific case until a proper investigation is carries out (which is rarely the case). But most of the times, experience says that it is company policies/procedures that either create circumstances that make such errors more probable, or simply allow these errors to happen too easily due to lack of enough guardrails. And usually it is due to push for "shipping products/features fast" or similar with little concern to other regards.
It could be a different case here, but seeing it is about oracle and having in mind that oracle is really bad at taking accountability about anything going wrong, I doubt it (very recently they were denying a hack on their systems and the leak of data of companies that use them until the very last moment).
I would regularly write massive update/insert statements on production DBs to fix issues.
So, yeah, id imagine this was the engineers fault.
Sure, Postgres beats OracleDB, but Postgres doesn't integrate as well with Oracle Fusion and you need to migrate the code yourself. They're like SAP: they're big enough that you can make a career just out of configuring their software packages and make good money while doing so.
It's expensive and certainly not the best, but it's reasonably stable and has a huge company backing it. Oracle won't disappear any time soon and they're not as likely as Google or Microsoft to shut down their services within a few years notice.
In some countries, Oracle is also very good at doing what Google and Microsoft are doing to students. The Brazilian programmers I've spoken to specifically learned OracleDB when they were taught relational databases. They learned to program in Java, and I'm sure Oracle also sponsored other parts of the professional tooling they got to use for free. Microsoft, on the other hand, didn't seem to generous towards their educational facility (no free MS tooling for their schools like they offer over here). If all you know is Photoshop/Windows/Maya/OracleDB/iOS, you're going to look for jobs where you can use Photoshop/Windows/Maya/OracleDB/iOS, and employers looking for cheap juniors will need to offer Photoshop/Windows/Maya/OracleDB/iOS to make the best use of them.
Are we sure? I'm by no means a DBA, but DBA at our company (who is freaking smart btw) said if money wasn't an issue, he would actually go with OracleDB.
Oracle have a relatively big presence here and there’s a comfortable “mates” system that runs the Australia (soft bribery).
And after a few years you find yourself in a situation when you already paid for Oracle so much, integrated it so deeply, so switching to any alternative is a massive pain and in most cases it’s safer and easier to keep paying Oracle.
COBOL is in the same category. When invented, it was the absolute easiest programming language to learn and use, so of course it gained popularity.
It then turned out to be rather good at what it did, along with running on hardware that "never fails", so most places didn't even think about replacing it until sometime in the 90s.
Also keep in mind that the reason companies are migrating away from COBOL is not due to the language as much as due to young people not taking an interest in COBOL and Mainframes, making it hard to recruit new talent.
Even then, a migration away from a typical mainframe is a huge task. In most cases you're talking 50-60 years of "legacy" (still running, still working, still updated) code, and replacing a system that has evolved for half a century is not an easy undertaking, at least not if you plan on getting it 100% right the first time.
And let's be honest, a lot of folks in IT aren't exactly top performers and don't seem to care all that much. It's really the developers you find on forums like this who are genuinely passionate. You're not likely to bump into the people actually buying those big Oracle or IBM systems around here though :)
Theory: A society collectivey buys Oracle when they no longer view armed revolution as acceptable.
Luckily more and more customers switched to Postgres and I no longer have to deal with it.
Also - cerner software in general allowed hospitals to freely completely fuck up their own architecture. Sometimes ireperably.
If anyone has details about how this happened Id love to hear.