Skip to main content
Question

fme_feature_type missing (sometimes)

  • January 29, 2025
  • 7 replies
  • 41 views

pkno
Contributor
Forum|alt.badge.img+5
  • Contributor

Hello everyone,

I have a question regarding the fme_feature_type. I encountered a problem in my workspace regarding the SchemaMapper which (after working properly before) gave me the following error:

SchemaMapper_Mapper (SchemaMappingFactory): A required column was not found in the table. Row fme_feature_type will be skipped.

 

After doing some testing I realized the attribute fme_feature_type was <missing> across the board in this particular conversion which I believe is required for the SchemaMapper to work. 

I’m somewhat stumped because sometimes the workspace functions properly and the attribute has proper entries. I noticed especially with larger datasets that fme_feature_type just goes missing between transformers. Some culprints that I found were for example the SubstringExtractor and the AttributeKeeper. The data goes in with fme_feature_type properly filled and comes out with the data missing.

I don’t think this is expected behaviour but maybe someone can shine some light on what might be happening here.

Thanks all.

 

7 replies

nielsgerrits
VIP
Forum|alt.badge.img+54

What FME version do you use?


pkno
Contributor
Forum|alt.badge.img+5
  • Author
  • Contributor
  • January 29, 2025
nielsgerrits wrote:

What FME version do you use?

I am using FME Form 2024.1.2.1


nielsgerrits
VIP
Forum|alt.badge.img+54
pkno wrote:
nielsgerrits wrote:

What FME version do you use?

I am using FME Form 2024.1.2.1

I can’t reproduce what you describe using 2024.1.2.0 using an AttributeKeeper or a SubstringExtractor. It may be possible fme_feature_type is overwritten by an empty value when using something like a FeatureMerger?

Do you do production runs with or without FeatureCaching? Because if it works when running step by step using FeatureCaching, the order may be differen from when it is run in production?


pkno
Contributor
Forum|alt.badge.img+5
  • Author
  • Contributor
  • January 29, 2025

I am still testing at the moment, so feature cashing is enabled which is why I could pinpoint exactly where the entries went missing. In this case it looked something like this:

FeatureJoiner (inspect cashed values: fme_feature_type has entries) outputs into SubstringExtractor input (inspect cashed values: fme_feature_type missing)

Unfortunately I can’t provide a screenshot right now as this error only occured after several hours of processing. I haven’t had this problem using smaller datasets.

   Edit: I have not tried without feature cashing yet, I might try that over night


nielsgerrits
VIP
Forum|alt.badge.img+54

Workspaces are generally quite a bit quicker when featurecaching is disabled. But when this does not occur when using a small sample set, and does happen with a bigger set, I would suspect it is data related...


pkno
Contributor
Forum|alt.badge.img+5
  • Author
  • Contributor
  • January 29, 2025

I am not sure how this could be data related as there is nothing else happening between the two transformers, but maybe I am missing something. Will surely do some further testing. Just now I was able to produce the error once again.  
 

This is the FeatureJoiner Output (only fme_feature_type column shown)

And this is the AttributeKeeper which comes right after

 


pkno
Contributor
Forum|alt.badge.img+5
  • Author
  • Contributor
  • January 30, 2025

This was somewhat solved by moving the AttributeKeeper behind the SchemaMapper or alternatively recreating fme_feature_type entries using an AttributeCreator. I still can’t explain the behaviour though. 


Reply


Cookie policy

We use cookies to enhance and personalize your experience. If you accept you agree to our full cookie policy. Learn more about our cookies.

 
Cookie settings