Re: Polygraph version 3.0.0 availiability

On Mon, 2005/02/14 (MST), <charlie_at_xiplink.com> wrote:

> I have noticed that Polygraph version 3.0.0 documentation has started to
> show up on the polygraph web site. Any insight into when it will be
> released to the public? And if it's anytime soon, do you have a list of
> new features?

New features include HTTP compression[1], automated convertion of access
logs to PGL workloads[2], Cookies[search 3], and improved tracing
support[4]. I do not know when the new code will be available to nonpaying
public.

        [1] http://www.web-polygraph.org/docs/userman/compression.html
        [2] http://www.web-polygraph.org/docs/userman/access2poly.html
        [3] http://www.web-polygraph.org/docs/reference/pgl/types.html
        [4] http://www.web-polygraph.org/docs/userman/replay.html

> Second question: anyone have any pointers to examples of how to create
> custom Polygraph Content Database files?

You can use cdb program from Polygraph distribution to add files to
the content database:

         cdb test3.cdb add --format markup file1.html

you can run that many times, adding more and more files

         cdb test3.cdb add --format markup file2.html
         cdb test3.cdb add --format markup file3.html

or several files at once:

         cdb test3.cdb add --format markup file4.html file5.html

At any time, you can see the contents of a file by using "show"
command:

         cdb test3.cdb show

The "markup" format will tell cdb to split the file into HTML pieces.
Polygraph servers will assemble "semi-random" HTML pages from those
pieces runtime, based on PGL-configured content size distribution.
Image src links and such will be substituted to point to unique
objects for each page. We have not tested that mode recently, but I
hope it still works. It is useful for testing content adaptation and
prefetching services, among other things.

The new "linkonly" format will add files as is, without splitting into
HTML elements, but will still process the source links. You can, for
example, populate your database with thousands of HTML pages from
popular real web sites and test various content acceleration algorithms
using "real" HTML.

The "verbatim" format will add files as is and will not cause source
links to be processed. This format is useful for non-HTML sources such
as images or vireus samples. It is also useful for preserving original
file sizes.

It is always a good idea to start with a small, easy to understand and
debug database before creating large, realistic content collections.

More details are available at
http://www.web-polygraph.org/docs/userman/csm/
which we need to update with the new --format examples above. Please let me
know if the above does not answer your questions or if you need a different
functionality.

Thank you,

Alex.
Received on Tue Feb 15 2005 - 09:30:25 MST