On 2005.02.06, Jim Wilcoxson <[EMAIL PROTECTED]> wrote:
> we're using ns_shares heavily and so are stuck with TCL 7.x, which is
> no longer supported.  (I know there are hacks in TCL 8 for ns_shares,
> but when we benchmarked them, they were much slower.  So no compelling
> reason to upgrade, for us, at least not yet.)

Jim,

Can you share the code and a description of the process you used to
benchmark ns_share performance?  Can you share the results of your
benchmarking, too?

(If you've already done this before, I apologize: could you point me to
where you've done so?)

If the performance of ns_share is the ONLY thing keeping you from moving
forward to AOLserver 4.x and Tcl 8.x, then I'd like to see what kind of
effort would be involved in making ns_share perform better.

The current ns_share implementation in 4.x (based on my understanding of
the code in nsd/tclshare.c) uses a single Tcl hashtable serialized
through a single critsec lock and Tcl traces to keep data synchronized
across threads.

I think we could get much better performance by defining a new Tcl obj
type that facilitates the thread-shared storage (read: this could
benefit NSV's too), but I don't want to start working on this until I
have a decent benchmark in place to judge whether the effort is worth
the investment or not.

If you've already done the initial work of creating a usable benchmark,
I'd like to reuse that as a head-start.  Otherwise, I might go ahead and
create one ... in my Copious Spare Time (tm).

-- Dossy

--
Dossy Shiobara                       mail: [EMAIL PROTECTED]
Panoptic Computer Network             web: http://www.panoptic.com/
  "He realized the fastest way to change is to laugh at your own
    folly -- then you can let go and quickly move on." (p. 70)


--
AOLserver - http://www.aolserver.com/

To Remove yourself from this list, simply send an email to <[EMAIL PROTECTED]> 
with the
body of "SIGNOFF AOLSERVER" in the email message. You can leave the Subject: 
field of your email blank.

Reply via email to