From Dapper To Lucid, Four Years Of Ubuntu Benchmarks

[Roman]

Launchpad bug report: https://bugs.launchpad.net/bugs/567302

[darkbasic]

It tests performance under ideal conditions, assuming that Apache Bench is not running on the same system that is being tested.

I don't think so...

[kingN0thing]

How am I supposed to choose Ubuntu for servers with such big regressions in server apps(Apache and PostgreSQL)?

Well, maybe just start by installing the Ubuntu Server Edition instead of the Desktop Edition? What the benchmark has told me so far is, that the desktop crew did a extraordinary job (I do not run a production mode HTTP server on my desktop, just my development environment.. and with that I do not need thousands of request per second).

The Server Edition has a specially optimized kernel for server usage. The desktop edition ain't. You know, that's why it's called desktop edition.

[kbios]

I'd say there are better reasons than that not to use Ubuntu for any kind of serious server; it's fine for a home NAS and the like, but IMHO it's far too 'bleeding edge' for a server which has to be reliable and secure 24/7.

Well, I use Ubuntu for the server I administer since version 7.10 and it has always been very reliable (well, except when I had to use experimental git code and -rc kernels to get what I wanted, but that's not Ubuntu's fault). Given these results, I think I'll keep ext3 (over lvm over raid1) until btrfs is stable.

[Shining Arcanine]

I'd say there are better reasons than that not to use Ubuntu for any kind of serious server; it's fine for a home NAS and the like, but IMHO it's far too 'bleeding edge' for a server which has to be reliable and secure 24/7.

Bleeding edge can be a good thing when new versions of software fix more bugs than they introduce.

This is part of the reason why I would use Gentoo Linux if I ran my own server.

[srg_13]

I'd like to see how Cherokee fares compared to Apache in these tests. It generally tends to be faster than lighthttpd (which in turn is much faster than Apache).

[movieman]

How am I supposed to choose Ubuntu for servers with such big regressions in server apps(Apache and PostgreSQL)?

I'd say there are better reasons than that not to use Ubuntu for any kind of serious server; it's fine for a home NAS and the like, but IMHO it's far too 'bleeding edge' for a server which has to be reliable and secure 24/7.

[Shining Arcanine]

Phoronix Apache benchmark doesn't benchmark real world Apache performance, because visitors aren't connecting from the server where the site they're visiting is hosted Phoronix Apache benchmark tests something else, but I don't know what ;>

It tests performance under ideal conditions, assuming that Apache Bench is not running on the same system that is being tested.

[karl]

Could we get a Apache/Pg test in which the ext4 is set to the "unsafe" mode similar to etx3 level?

That way we could see if it's about etx4 or something else also.

[kraftman]

BTW such a regression in apache is a failure: any sane website admin will never accept an order of magnitude regression, mostly because this is a LTS release.

Phoronix Apache benchmark doesn't benchmark real world Apache performance, because visitors aren't connecting from the server where the site they're visiting is hosted Phoronix Apache benchmark tests something else, but I don't know what ;>