[Borges-users] performance/memory

[Eric Hodel]

??????? ???????? (slepnev_v at rambler.ru) wrote:

> Hi,
> 
> today I implemented my first small Borges app - wanted to show off 
> Borges to my employer on Monday. I tried a small stress test: just 
> clicking on links a lot of times, and watching the Task Manager 
> window. 
> 
> The first version (which stored a continuation for every click) 
> increased the memory about a megabyte per click, and started thrashing 
> wildly at something like 80 megs. No response in 5 minutes. 
> Unacceptable.
> 
> The second version did not store any continuations, but increased the 
> memory about 500K a click. I decided not to wait until it starts 
> thrashing =) So:
> 
> 1) Eric, you have a major bug somewhere. (Or is it Matz? =)

I'm quite certain its mine.  (I would be very, very suprised if it was a
Ruby flaw.)

There are two problems with Borges, one is a bug where memory grows
seemingly unbounded.  I need to track this down still.

The other is that Seaside uses *lots* of small objects, and
Smalltalk is very good at this.  I don't think Ruby is quite that good,
and certainly is not nearly as fast as Smalltalk at handling them.  As I
see fit, I'm rearchitecting parts of Borges to remove/reduce these tiny,
short-lived Objects in favor of other structures.  Last weekend I
finished up changes to the HTML renderer and CallbackStore that will
help in this respect.

You can fetch a newer Borges from CVS which has these fixes.

> 2) How can I disable old pages altogether? I can design my app around 
> this. Maybe it could help. Borges has many other cool features I just 
> can't let go of now.

You can reduce the number of saved pages by changing the value of
Borges::LRUCache::DEFAULT_CAPACITY.  Currently this is set to 20.

I'm not sure exactly where the bug is, so eliminating use of the
LRUCache completely in Borges::Session (@continuations) may or may not
stop the memmory growth, but that is also worth a shot.

> Vladimir Slepnev
> 
> P.S. Just tried the same test on the SushiNet example: simply hitting 
> the "Browse" link repeatedly. Increases memory about 2.5M per click. 
> Sad.

Yes, I'm well aware of the problem.

-- 
Eric Hodel - drbrain at segment7.net - http://segment7.net
All messages signed with fingerprint:
FEC2 57F1 D465 EB15 5D6E  7C11 332A 551C 796C 9F04

-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://rubyforge.org/pipermail/borges-users/attachments/20040410/dae1aa51/attachment.bin