On Thu, Mar 11, 2004 at 12:59:46AM -0500, Nicholas Sturm wrote:
> 
> Would you possibly agree with this, "There is no known way to meaningfully
> evaluate the performance of freenet?"

No. There are several ways to evaluate it. My favourite is "push/pull
tests". Insert a file on one node, and fetch it from another unrelated
node. How long does it take? How many retries? Etc. That is one
performance measure. Another one is "does streaming work?". Nobody has
been crazy enough to try recently to the best of my knowledge :).
> 
> 
> > [Original Message]
> > From: Conrad Sabatier <[EMAIL PROTECTED]>
> > To: vinyl1 <[EMAIL PROTECTED]>
> > Cc: <[EMAIL PROTECTED]>
> > Date: 3/9/2004 7:28:55 AM
> > Subject: Re: [freenet-support] freenet commitment settings
> >
> >
> > On 09-Mar-2004 vinyl1 wrote:
> > > OK, I have the latest build (5074), the latest seed nodes, the latest
> > > everything I could find as of today, 3/8/2004
> > > 
> > > On the Web interface, I can't load The Freedom Engine, Dolphin's Free
> Index,
> > > and Content of Evil.  I can load the Freenet Help Index and YoYo.  This
> is
> > > possibly due to unreachable, out-of-date nodes.  The stuff that does
> load
> > > seems to work better than before.
> >
> > Basing your evaluation of your node's performance on what's appearing
> and/or
> > reachable via the main web interface is not a good idea.  Most of the
> gateway
> > sites have been horribly unreliable for quite some time now (with the
> exception
> > of DFI).  I really don't know why this is the case; I've had no trouble
> at all
> > inserting DFI daily.  To ensure that the next DBR update is inserted in
> time, I
> > always start the update process (an automated, scheduled job) at least 1
> 1/2
> > hours prior to the rollover time (12:00 am GMT).  I'm wondering if other
> > gateway site maintainers are not allowing enough time in advance for their
> > inserts to complete on time.
> >
> > DFI's insert got a little screwed up yesterday, due to the fact that I
> was in
> > the process of running a portupgrade of my JDK under FreeBSD.  Probably
> due to
> > the additional load on the system from running the build, FIW somehow
> ended up
> > inserting DFI one day further into the future than it was supposed to.  I
> didn't
> > discover this and finally get the problem corrected until about 8:30 pm
> CST
> > (2:30 am GMT).
> >
> > I saw someone else here basing assumptions about their node's behavior on
> the
> > fact that they couldn't reach YoYo!  Unfortunately, Yoyo! is one of the
> more
> > unreliable sites lately (no idea why).
> >
> > Anyway, the gist of the idea is this: don't assume that non-functioning
> gateway
> > sites mean your node is not working.
> >
> > -- 
> > Conrad Sabatier <[EMAIL PROTECTED]> - "In Unix veritas"
> >
> > _______________________________________________
> > Support mailing list
> > [EMAIL PROTECTED]
> > http://news.gmane.org/gmane.network.freenet.support
> > Unsubscribe at
> http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support
> > Or mailto:[EMAIL PROTECTED]
> 
> 
> 
> _______________________________________________
> Support mailing list
> [EMAIL PROTECTED]
> http://news.gmane.org/gmane.network.freenet.support
> Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support
> Or mailto:[EMAIL PROTECTED]

-- 
Matthew J Toseland - [EMAIL PROTECTED]
Freenet Project Official Codemonkey - http://freenetproject.org/
ICTHUS - Nothing is impossible. Our Boss says so.

Attachment: signature.asc
Description: Digital signature

_______________________________________________
Support mailing list
[EMAIL PROTECTED]
http://news.gmane.org/gmane.network.freenet.support
Unsubscribe at http://dodo.freenetproject.org/cgi-bin/mailman/listinfo/support
Or mailto:[EMAIL PROTECTED]

Reply via email to