Spatial Join Slow

[JulianKatz-Samuels]

Hi,

I'm using arcmap 10.0 SP5 to do a spatial join with a file that has about 61,000 observations, many of which are null, but it is taking an extremely long time (around 1 day). I thought that it would take a shorter period because I've done spatial joins with bigger files that took a shorter amount of time. Is it typical for it to be taking so long? Is there something that I can do to speed it up?

Thanks,
Julian

[curtvprice]

These datasets aren't that big. I'm wondering if one could do better putting one or both of the datasets in_memory.

[AlesVanek]

Hello,

I found this old post and I am having the same problem as described above.

I am on the latest ArcMap 10.8 - yet with 290k points and 20k objects, the spatial join takes about 20 hours - running on Xeon E5 2680 based cloud. 

Appreciate any advice

Regards, 

Alex 

[curtvprice]

Shapefile or GDB? GDB feature classes have spatial indexes built in so I would expect it to run faster.

[AlesVanek]

GDB actually helped for the 290k case - it did process in less then 10 minutes. 

However my larger file with 600k points as GDB joint with the same 20k objects in GDB will still not complete.

Current run time is ~30 hours and it is stalling.  ArcMap allocated ~254 MB out of 64 GB of server memory. 

Thanks

[curtvprice]

Just some thoughts

• I'd try it in Pro, they have been doing a lot of tiling work to support multiprocessing (and distributed ! processing) and this has made large overlays much more robust and scalable. That and Pro does it all in 64 bit which solves a lot of memory management issues.
• If your workflow is tied to ArcMap, and you haven't installed background 64 bit processing, I'd try that - then you get ArcMap tools running in 64 bit which also can help with scalability.

Good luck!