Large sym enumeration file

Hi All, I noticed something kind of troubling today.  The sym enumeration file in my hdb seems very large.  Its about 400mb and contains 4.5mill entries.  When I inspected it I noticed I have items in there that I may not want including orderIds. When I use .Q.dpft to store tables all the symbol columns got enumerated there.

Couple questions,

This will cause slower queries correct?

This doesn’t seem like an easy/quick fix.  I think I would have to change all columns I don’t want enumerated there to either strings or their own enumeration.  Then create a new sym file.  Can someone help me think about a correct fix for this, if its needed?

HI Roni

We hit exactly same problem and turnout the most simple to solve this for us is create a new HDB, copy the data from the old HDB to new old day by day, with some column converted to string ( character array ) instead of sym

In fact, I feel if the column is not usually used as filtering criteria, it should never been using datatype sym.

re: This will cause slower queries correct? 

Theoretically it should but not by a lot as those columns are enumerated.
The concern would be more for time needed to reload the table, as well as memory it occupies for the server.

And as above suggest it’s gonna be painful for the fix…

X fix x cb ok r???