According to the following, there is a time in the compact where it
creates a compacted copy of the original. Takes the original offline and
renames the copy to the original. Running a task during this may cause
Problem statement: How do I get adminp to run on a timely basis when it
Now we know how it aborts. How do I work around that?
- Never run compact.
- Bring down server and run compact in a dedicated mode.
- Check to see if adminp is running periodically. If not, then load it
I might be able to whip out the third option with api's. However, is
there some way to do that with a monitor or some such animal?