The question is 'what are the two sessions involved in the block doing?'

If, for example, the TOAD user runs an UPDATE, all the affected records will be locked until that use eventually gets around to doing a COMMIT. If an application user then tries to UPDATE one of these locked records, they will be stuck. This is completely expected.

However, I will assume that we are actually dealing with completely different records in the same or different tables. In that case, there are only 2 possibilities:

- There is a BITMAP index on the table.
- You have an FK indexing issue. If you don't have an index on an FK, then if someone changes a parent PK or deletes a parent, it will lock all the related rows in the child. See http://oradoc.photo.net/ora8doc/DOC/...1/ch6.htm#1574 for more info.

HTH,

- Chris