Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

Scalling an application is easy. You just stick a load balancer in front of more and more webheads. The hard part is scaling a database!

Twitter is still running a single master database system and - from what they've put out - only three database servers. They might also have some pretty, but wasteful queries that do a lot of joins. Joins are not good for high-hit applications.



I thought this also wasn't hard by using horizontal partitioning..? Which, I think, would have to map back into the application too.




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

Search: