I created an SDE view between a point feature class and a table (stored in SDE) - the join is done on a text field called RFRNC. I changed the view so that it is now a LEFT OUTER JOIN in line with this doc:


Problem is - when create a row in my flat table in the RFRNC column with a value I know exists in the featureclass - it never shows up or updates in the view - what is going on here? Any ideas greatly appreciated.

  • How are you creating a row in your flat table? If you recreate the view after you add the data does it show up?
    – cwhite
    Sep 1, 2010 at 1:39

3 Answers 3


Apparently you just need to work with unversioned featureclasses - then it you see the values - elementary mistake! You need to change the Editor toolbar to work with unversioned datasets too.


Just a guess, but the following information from here may be why you're having an issue:

If you choose to replace views created by ArcSDE by altering the view afterward, you are taking responsibility for not only the query's correctness and efficiency but also its relevance and suitability for use with ArcSDE clients. For example, creating one-to-many views involving feature classes results in duplicate ObjectIDs. This leads to unexpected behavior in applications such as ArcMap and ArcCatalog, because the ObjectID is supposed to contain unique values. See About ObjectID fields for a definition of ObjectIDs.

  • Yes - I read that, but going from the article I showed the link for - it should work without a problem.
    – Vidar
    Sep 1, 2010 at 8:35

The first thing I would do is compress your SDE database. In the past when I have had this issue it has been because the features I was using for the join were not in the base table of the feature class. From past research, I have found that you won't see records in a spatially enabled view unless the features are in the "base tables" for that particular feature class. When you compress your SDE database you will migrate all the records in the A and D tables to the base tables. If that turns out to be the problem, script a task to compress your SDE database on nightly basis to ensure that all edits are in the base table and available to the spatial view.

  • The data is held in a remote SDE (sql server) database - does compressing apply in this instance? or is that just for file based data?
    – Vidar
    Aug 31, 2010 at 20:58
  • Compress only applies to SDE data; Compact is used for Access databases and there's no file-based "compress" or "compact" that I'm familiar with. Aug 31, 2010 at 21:06
  • Michael is correct, compression applies to SDE data only. I used the term specifically for that reason. You can open ArcCatalog, right click on the the database instance and find the Compress button in the context menu. It can also be found in ArcToolbox and by using sdeversion -o compress at the command line
    – Zachary
    Aug 31, 2010 at 21:43
  • There is a Compact command in ArcToolbox that is used with File and Personal GDB.
    – Vidar
    Sep 1, 2010 at 8:32
  • 1
    To clear up the terminology a bit, the term "compression" is used not only on SDE data (where it, simply put, denotes moving records from A_ and D_ tables to the base business table), but also for file geodatabases, where it means compression in the more traditional sense, i.e. reduction of the occupied disk space. Zachary here meant the SDE compression process.
    – Petr Krebs
    Sep 1, 2010 at 11:35

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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