ObjectID field - what happens when it gets to high?

Jay_Gregory

If I have some reoccurring scripts that keep appending to a feature class, eventually the objectID field will get too high - for example, over 1000 new records every 30 minutes.  Some of the old records are deleted to limit the size of the feature class, but the ObjectID field keeps getting higher and higher.  Theoretically, eventually, the ObjectID field will max out at 2,147,483,648.  What happens then?  Is there any way to reset the field programatically?

VinceAngelo

At 2kF/h, you'd hit the 2^31-1 32-bit integer limit (...7, not ...8) in 122 years
(44739.24 days).  If only some of the features are deleted every hour, you'll
hit the effective use limit of 50M rows in a table in just over 3 months.  This is
probably of greater concern.

Sure, I've populated tables with 690M rows, and I've seen a table with over
a billion rows, but populating large tables (the former took 22 hours) isn't the
same as making effective use of them (the latter could only be queried for
the past 72 hours of data).

The key to making *any* use of very large tables is paying careful attention
to how they're populated, the indexing, and the use of partitioning.  Since
spatial fragmentation is very likely to occur, you'll also need to design a
defragmentation protocol. Once that's in place, the solution for rowid
numbering might be more acute (unless your defrag procedure doesn't
involve additional inserts).  But before that becomes a critical issue, Esri
will probably have implemented a 64-bit rowid option.

- V