It's hard to make a comparison between the old MapGuide and
the new
MapGuide, as they have totally different data access
mechanisms. Even
the SDF format has been completely rewritten. In both
cases, I have
found SDF/SDF+ to give extremely rapid access to the data.
My slowdowns
with the old version had to do with the need to query an
external
datasource for additional attributes. This restriction has
been removed
in the new version.
I have also found that the standard mechanisms of view
scales and
generalization really help to increase speed of access, and
reduce load
on the server.
The only issues I have seen with large files so far are
related to
uploading the data via PHP. In these cases, I haven't seen
an problems
with packaging the data locally and uploading via
conventional methods.
It looks like Safe Software has introduced SDF+ in its most
recent
betas; I wonder if I can use it to manually create
packages...
Jason
-----Original Message-----
From: mse-users-bounces lists.mapserverfoundation.org
[mailto:mse-users-bounceslists.mapserverfoundation.org] On Behalf Of
Jeff McKenna
Sent: Thursday, January 19, 2006 1:08 PM
To: mse-users
Subject: [Mse-users] Efficiently Using Large Datasets
Hello everyone,
As a MapServer (traditional) user, I have a question for
both existing
MapGuide users and new-school MapGuide Open Source users.
Forgive me if I am ignorant of MapGuide's abilities, but
when you 'have
the blinders on' using one software program for a while
you're just not
aware of how other things work out there.
Let me start by explaining how we deal with large datasets
in MapServer
(traditional). Imagine trying to use a large dataset, such
as TIGER,
with several thousand 'lakes' files, remembering that the
web is all
about speed, and you get the picture of how important it is
to
efficiently display large datasets. MapServer's native
format is
shapefiles, so we would normally convert to shapefile, and
then use a
utility to create a spatial index of the files. MapServer
uses this
spatial index file to find the locations of all the
individual files
(such as 'lakes'). It works really well.
I have been asking around and searching on the list for how
to deal with
this for MapGuide Open Source. I have heard rumors that the
SDF format
is designed for this issue. I have also seen recent posts
about using
the 'Package Spatial Data' option in Studio. Another thing
that I heard
is that I may have to merge all of the 'lakes' files into
one massive
file (which scares me because I have also seen posts on
Studio/MapGuide
problems with large single files).
My question is how do MapGuide users handle this much data?
What is
recommended?
Thanks very much.
jeff
--
Jeff McKenna
DM Solutions Group Inc.
http://www.dmsolutions.ca
_______________________________________________
mse-users mailing list
mse-userslists.mapserverfoundation.org
http://lists.mapserverfoundation.org/mailman/list
info/mse-users
_______________________________________________
mse-users mailing list
mse-userslists.mapserverfoundation.org
http://lists.mapserverfoundation.org/mailman/list
info/mse-users
|