Hacker News new | past | comments | ask | show | jobs | submit login

I am a great fan of scaling vertically as far sa possible on DB servers. These days that is pretty damn high. It avoids a lot of prickly edge cases.

It is definitively not one solution for all. There are many cases where it just won't work.

I would like to see more IBM Z servers being used. $$$$$$$$ though




It doesn't solve for required multi-region data storage. Nor for data center failure resilience.

Scaling up is fine for a few things, but hopeless for many others.


For data-center failure, it does: the underlying storage can be resilient.

For multi-region, indeed, that will not be possible. Master-slave would be the way.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: