We have the following configuration. I have major speed (slowness) issues no matter if my files are all on a network or all local on my PC. The file based GeoDatabase we use consists of all of our campus buildings. In the GeoDatabase we have 4 basic feature classes and 2 data tables.
1) Feature class/layer Room (~80,000 polygons)
(all room polygons for all buildings on campus)
2) Feature class/layer Building (~1,500 polygons)
(all building outline polygons for all buildings on campus)
3) Feature class/layer Door (~67,000 polygons)
(all 45 degree doors polygons for all buildings on campus)
4) Feature class/layer Reference (~5,000,000 elements)
(all other line work from our CAD floor plans for stairs, columns, shafts, walls, doors, windows etc. All defined as lines. This is our problem layer. There are almost 5 million small lines that make up this layer.
ArcMap 10.4.1 - Map with one Data source attached (file based GeoDatabase) 500 MB in total size.
There are 6 layers associated with this data source. All graphic layers are filtered by a definition query by building and floor. One building and floor displayed per map. (ex. Room.BldRecNbr = '1000135' AND Room.Floor = '05')
1) Room polygons (~60 polygons)
2) Building polygon (1 polygons)
3) Door polygon (~51 polygons)
4) Reference lines (~1,500 lines)
5) Room Data table (~114,000 records)
6) Building Data table (~1150 records)
The map works fine as long as the Reference layer is not checked on. When ON, it appears to search the entire database for every line in that feature class before displaying onto the screen. Thus just zooming in our out takes upwards of 5 minutes each time. In my query there are only approx 1500 line elements selected in this feature class but in the GeoDatabase this feature class has over 5 million lines.
So the question is. Is there a setting or something that I can configure to read and display only the elements in the query or some other configuration setting that I am missing to speed up the display and operations of this layer?? If not, what other options do I have to speed up performance when using this layer.
We have looked into using FME to import and combining all reference elements into a single polyline. I have accomplished this but at this time I would simply like to know if I can speed up my performance without manipulating the data itself.
Thank you in advance for any help you can give me. I have attached a copy of my mxd file in case that helps.