Hacking and Other Thoughts

Tue, 10 Apr 2007

Solaris scalability...

I'm still fuming over Jeff Bonwick's entry into the anti-Linux FUD campaign. He should really take a look at this thread on linux-kernel, before spewing like that.

Now the context of that thread is that SGI is running into scalability issues on some "new big machine", and the problem they are running into is the traversal of the child task list of the init task.

What you have to understand is that they aren't specifying how large this new machine is, but since we know that they've been running Linux on 1024 cpu machines for quite some time at SGI you can be sure this machine is likely enormous.

And here's the punch line, Solaris has never even run on a 1024 cpu system let alone one as big this new SGI system, and Linux has handled it just fine for years. Yet Mr. Bonwick feels compelled to imply that Linux doesn't scale and Solaris does. To claim that Solaris is more ready to scale on large multi-core systems is pure FUD, and I'm saddened to see someone as technically gifted as Jeff stoop to this level.

Now, this all would be amusing if this were the early 90's and us Linux folk were "just a bunch of silly hobbyists." Yet these Solaris guys behave as if we're still in that era.

They must really be feeling the pressure in the Solaris group.

The Solaris FUD machine continues...

Peter Zaitcev makes reference to an amusing blog entry by Jeff Bonwick. To think that I used to respect this guy for his SLAB work.

Now he's stooped to the point of spewing a bunch of blatant lies about Linux. Here is what John states you need for an embedded storage appliance:

	You need something rock-solid, so it doesn't break or corrupt
The implication is that Linux is not rock-solid and that it does break and corrupt people's data. Whereas on the other hand Solaris, unlike the rest of the software in this world, is without any bugs and therefore won't ever break or corrupt your data.
	You need something that scales, so you can take advantage of
	all those cores the microprocessor folks will be giving to you.
Last time I checked, Linux scales the crap out of Solaris. Regretfully, Solaris cannot make use of the advanced SMP scaling techniques Linux has such as RCU locking, but that disappointment is no reason to spread the FUD around like this.

He goes on to implicate lies about a bunch of other things he thinks Linux doesn't have. But the most blatant lie is this one:

	And you need a community of equally serious developers who can
	help you out.
Give me a friggin' break. The one thing everyone knows for sure is that the only real consequence of Sun openning up Solaris is that now the entire world gets to watch how glacial Solaris development is compared to Linux and how miniscule the community behind Solaris is.
	The other operating systems one might consider are so far behind on
	so many of these metrics, it just seems like a no-brainer.
The only real no-brainer is that his blog entry is full of crap.

Linux has developer momentum, feature lists, performance, and scalability the likes of which Solaris can only dream about.

And let it not be missed how much Sun enjoys attacking it's competition via blogs, so they don't have to make any official statements in this area or stand behind what they say in any official capacity. And that, my friends, makes for one big coward of a company.

Peter's quotes from Larry McVoy in 1993 deserve duplication here, it's amazing how prophetic his srcos paper was:

One the other hand, if the politics kill such an idea [open-sourced Solaris], most software developers will simply give up on Sun and SPARC and focus instead on Linux and Intel. Sun has most to lost here.
Can you believe it, 13 years ago? Amazing.