Matthew Marcus (Customer) asked a question.

April 21, 2026 at 9:48 PM
Why is progress database so slow to dump the schema?

We have 1000 tables. Why does it take a 60s to generate a .df file, whereas if I used a SQL schema dump it would be done in seconds?

  • 5 answers
  • 136 views
Peter Judge and Mike Fechner like this.

  • Matthew Marcus (Customer)

    I believe it's

     

    -b -i -noinactiveidx -db name -H host -S port

  • Dmitri Levin (Customer)

    Compare this to shared memory connections. The schema dump uses the join queries. Does your version of Progress support server-side joins?

  • Matthew Marcus (Customer)

    I could, but that's not a restriction we'd like to introduce in our case. And to keep the apples to apples comparison, SQL is fast over the network too even :).

     

    I checked our DB: Database server side join support (-ssj): Enabled

     

    Like, why couldn't I just make a SQL based version to make the same DF?

  • Dmitri Levin (Customer)

    Joins for ABL and SQL are different mechanisms, but they produce the same results. As I mentioned before, test is simple: connect to the database without the -H/-S parameters.

     

    The ABL code can be fast for remote connections without using server-side joins, but the schema dump was not written with this goal in mind.