I have a large feature class containing many thousands of polygons. For several reasons, at least part of the time in our development/deployment process, it lives in a file GDB.

It is part of a map service that we publish where there are different layers based on multiple (4-6) thematic versions of this same feature class (rendered, filtered, and labeled based on different attribute values).

Basically, I've been adding the feature class to the map, renaming the layer, setting the display attributes, then adding it again and doing the same things.

This works just fine, but I have a nagging feeling that it's inefficient, particularly for a map service. The consumer of the map service has limited abilities to change rendering.

  • Would be useful to see a screenshot of your MXD.
    – jakc
    Sep 30 '13 at 12:49

Shame that your client has limited abilities to change rendering as the perfect answer would be to make use of dynamic layers.

However, your approach does not sound like a bad one. The key question is, are you having performance issues, and if so this might well relate to other aspects such as the large number of vertices that are having to be rendered, regardless on if this is being served up from multiple layers.

I would suggest making use of definition queries over just not symbolising the polygons you do not want shown. I dont have the metrics to back this up, but if you do some of your own using mxdperfstat or even the in-built preview tools, I believe you will see a slight improvement. (In fact, running MXDperfstat and posting the results here would also be useful to diagnose further any issues).

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.