vanillasite.at :: comments-2003-12-24

Wednesday, December 24 2003

After years and years (and, uhm, years) of hard and diligent work we are proud to release the hitherto longest-awaited Vanilla release ever: 0.6!

For the full scoop, all the new features, every single fixed bug: go read the vanilla-0.6-release-notes.

This is the first release since 714 days, and it's definitely been worth the wait.

David Collantes, 7668 days ago:
Would it be possible to download the templates this site uses, with images and all, to use as a reference towards my own branding? I love the way this site looks...

earl, 7667 days ago:
We're happy that you like it :), but no, we're currently not providing a bundled way to download the complete look and feel of this site.

jasonic, 7664 days ago:
Thanks for the new version :-)
Not time to test it yet, or figure out how to update my old site [dynasnips, merge customized vanilla.r etc].

This new vanillasite.at seems much faster -- Is it ?
Speed is my #1 concern about Vanilla.. Any new thoughts on this?

chris, 7664 days ago:
Jason, much faster than what? ;-)

jasonic, 7661 days ago:
...much faster than older/previous existing vanilla sites.

langreiter.com does seem swiftest, so I guess has been using v0.6 for a while,

possible reasons..
- no content yet
- the new code is better
- improved server/bandwidth
- just my imagination
- slimmer front page

Q1: Do Vanilla sites get slower and slower as they grow?

Q2: How does Vanilla scale?

Every time one indexes something, calls a snip or vanilla searches for some metadata or snip isn't it going take and longer and longer? Rebol can be both surprisingly fast or slow depending.

"find" is fast, but index is slow.

Becuase Vanilla is *so* cool, I am tempted to take for granted the magic of embedding snips, dynasnips etc.

That means a vanilla page can and up with a lot of work to do in the background beore it renders html. Probabably I am expecting it to do too much, and I am not caching or using sensible code to optimize where I should be.

I know there are some basic speed obstacles which cannot be avoided.

One example I refer to as a really fast data heavy site is Phil Askey's photo site
http://dpreview.com
written in ASP. It loads really well

As I understand it, Apache lacks native rebol multi-threading module. But in reality that may not prove to be such a big deal. I think you and others have commented on this in the past.

For example, my own [vanilla 0.5] customized testlab site is http://tranzilla.net/.v

1550 snips now and it can now be really slow to respond. For example [win98se 500mHz Vaio]:

10+secs to connect first time, even on fast DSL line.
Further hits are much better, say 4-5 secs.

And using a new 3Ghx WinXP notebook, I get about 30% better speed than those.

The server [hosted by dreamhost.com] is pretty swift.

I am not complaining, I just really want to understand what the limits, tricks and gotchas are for using Vanilla to host a fast site with lots of snips.

chris, 7661 days ago:
The basic problems are:

- "space scans", when Vanilla has to load every single snip (find) - as it is a CGI program, we can't keep any index/cached structures in memory.

- REBOL start-up overhead (for every single request, REBOL has to load, decompress mezzanines &c.).

- slow/massively shared hardware - dreamhost and friends are fine for static sites IMO, for dynamic applications a dedicated server (which can be pretty cheap nowadays) is of course the best solution.

Some options to improve performance:

- dedicated REBOL HTTP server: lots of work to make it stable enough for production use, usually impossible to set up on shared hosts à la dreamhost I think.

- dedicated REBOL server behind Apache (AJP13 &c.): hard to set up, usually impossible on shared hosts.

- "live" index structure generation: we do that in quite a few places (backlinks, recent-stores). hairy, introduces lots of dependencies.

- background index structure generation: quite a bit of work, requires cron access, usually impossible on shared hosts.

So the least work-intensive option is simply to get a dedicated server. A modern machine can easily handle dozens and dozens of Vanilla sites.

jasonic, 7655 days ago:
Chris,
Thanks very much. It really answers the question.I will look into dedicated server for sure.

Can you say more about what a vanilla-freindly rebol http server project would have to do?

chris, 7655 days ago:
Not that much besides being stable, gracefully handling malformed requests and being easy to extend with dynamic behaviour ...

Comments for this entry are now closed.