4 ms·

It has been known for DECADES that Oracle is outrageously expensive compared to the competition. But the unwillingness of larger companies to go with open-sourc

It has been known for DECADES that Oracle is outrageously expensive compared to the competition. But the unwillingness of larger companies to go with open-source databases which have no multi-billion dollar company behind them has kept Oracle in the saddle all these years.

qeternity 6y ago
I think that mission critical, legacy systems from a time before there was big OSS acceptance at the CTO level is the main reason. I’m not sure how many greenfield projects are using Oracle, but I’d expect that is rapidly decreasing over time. However, if you’ve had an Oracle db running for 20 years, and the cost of screwing up a migration starts at 9 figures, and runs into the 10 figures, then Oracle have you by the cost-benefit balls.
- swarnie_ 6y ago
  Oracle consultant here -> The Oracle world is shrinking admittedly and its shrinking from the smaller end up. I almost never get called by a company with less then 500 users or $100k a year of licence obligations. Oracle is a tried and tested safe bet when implementing corp wide, multi year projects. The HN crowd won't like it (and for good reason) but CTOs will always bet on name recognition when millions or billions are on the line.
  - miahi 6y ago
    As a fellow Oracle consultant, I see the same. Most of the small clients only involve me in migration projects to other DBs now (postgres or mysql), but the ones with >$1M licenses continue to pay their support and migrate to 19c. But that's also because it's way harder to migrate a huge project/organization to a completely different DB - it's just cheaper and less risky to continue with the current one.
skrebbel 6y ago
That, and a huge ERP thing.
snarfy 6y ago
Also, government lobbying. How many large government databases do you think are oracle? I suspect quite a lot.
AdrianB1 6y ago
I am a platform architect in a large company that has a few hundreds TB database servers that are not on open-source software; we do use Postgres in about the same number of instances, but not for the core databases. There are several reasons for that, from tooling (SSMS has no competition) or simplicity to manufacturer support; for example, using MS SQL setting up high availability or replication is a matter of a few clicks (or a page of code) with no extra add-ons, packages etc. and these features are perfectly integrated in the database engine, supported by the same company, come in the same installer. Even for smaller databases we prefer the free edition of MS SQL, for development, testing or validation we use the free developer edition, so the cost is not that bad, while the product is much more simple, coherent and better supported and this matters when you do the math. The day Postgress will have an integrated tool as good as SSMS, high availability and replication seamlessly integrated and the ease of use for the operations team to keep it in good shape without highly skilled DBA's, I am the first in line to evaluate and deploy it. Until then, it is not good enough, even for free.

qeternity 6y ago

I think that mission critical, legacy systems from a time before there was big OSS acceptance at the CTO level is the main reason. I’m not sure how many greenfield projects are using Oracle, but I’d expect that is rapidly decreasing over time. However, if you’ve had an Oracle db running for 20 years, and the cost of screwing up a migration starts at 9 figures, and runs into the 10 figures, then Oracle have you by the cost-benefit balls.

swarnie_ 6y ago
Oracle consultant here -> The Oracle world is shrinking admittedly and its shrinking from the smaller end up. I almost never get called by a company with less then 500 users or $100k a year of licence obligations. Oracle is a tried and tested safe bet when implementing corp wide, multi year projects. The HN crowd won't like it (and for good reason) but CTOs will always bet on name recognition when millions or billions are on the line.
- miahi 6y ago
  As a fellow Oracle consultant, I see the same. Most of the small clients only involve me in migration projects to other DBs now (postgres or mysql), but the ones with >$1M licenses continue to pay their support and migrate to 19c. But that's also because it's way harder to migrate a huge project/organization to a completely different DB - it's just cheaper and less risky to continue with the current one.

skrebbel 6y ago

That, and a huge ERP thing.

snarfy 6y ago

Also, government lobbying. How many large government databases do you think are oracle? I suspect quite a lot.

AdrianB1 6y ago

I am a platform architect in a large company that has a few hundreds TB database servers that are not on open-source software; we do use Postgres in about the same number of instances, but not for the core databases. There are several reasons for that, from tooling (SSMS has no competition) or simplicity to manufacturer support; for example, using MS SQL setting up high availability or replication is a matter of a few clicks (or a page of code) with no extra add-ons, packages etc. and these features are perfectly integrated in the database engine, supported by the same company, come in the same installer. Even for smaller databases we prefer the free edition of MS SQL, for development, testing or validation we use the free developer edition, so the cost is not that bad, while the product is much more simple, coherent and better supported and this matters when you do the math. The day Postgress will have an integrated tool as good as SSMS, high availability and replication seamlessly integrated and the ease of use for the operations team to keep it in good shape without highly skilled DBA's, I am the first in line to evaluate and deploy it. Until then, it is not good enough, even for free.