I'm an entry level FME user and I'm struggling trying to transform a file geodatabase to kml. The feature class consists of 67k culvert line features with 38 attributes. Both FME and ArcGIS seem to generate the kml but Google Earth encounter an error whey trying to open it. I thought the tutorial could help me discover what I need to do to fix this.
I'll attempt using the information and workbench file provided in this link https://knowledge.safe.com/articles/887/optimizing-large-datasets-for-kml-and-google-earth.html
Hi @danellelaurin
1. May I know your FME version.
2. Could you share with us a portion of the data?
3. May I know the content of the error you receive on GE?
Thank you.
67k features, with 38 attributes each, may be too big for a single KML file for Google Earth to process. In that case, chopping it up into tiles (or regions) is the answer, although in FME this is a bit tricky to get your head around.
But there are a few tips and tricks that could work having just one KML (or KMZ) file:
- Do you really need all 38 attributes? Remove any attributes you don't really need.
- In the KMLStyler, make sure you set the "Allow Unique Styles Per Feature" to "No". By default, it's set to "Yes", and creates a style block in your KML for each of your 67k features, whereas when it's set to "No", you only create one style block.
Also, have a look at this KML tutorial: https://cdn.safe.com/training/tutorials/KML-Pathway-Tutorial.pdf
When creating a large-scale KML file through FME, if Google Earth encounters an 'out of memory' error. A best practice may be to create several KML files rather than a single file. This will allow you to keep the data as you currently have it while reducing the size/memory of each independent KML file. The KML files can be created and linked using a Network Link KML. A tutorial can be found that provides an in-depth look at this process: https://knowledge.safe.com/articles/887/optimizing-large-datasets-for-kml-and-google-earth.html
Lastly, if trying to style the overview KML, ensure the KMLStyler that contains the polygon/boundary features is styled to how you wish it to be, for example: transparent. If you leave the boundary KMLstyler blank, you will have an overview KML file that is opaque.
I hope that helps!
Thanks for your help. The Optimizing Large Datasets for kml workbench template worked as expected.