What triggers the behavior of the BTRee cleaner thread?

IDS 7.24UC7, Solaris 2.6 This is in relation to my earlier post about the BTree Cleaner thread interfering with user processing. I thought the BTRee thread only wakes up like once every 60 seconds. It seems that it senses if an index entry has been deleted and proceeds to re-arrange the indexes. Anyway, that is exaclty what's happening right now in our production system. What our process does: (Lock mode has not been set) 1. unload about 200 records 2. delete about 200 records 3. begin work 4. lock table in exclusive mode 5. drop and recreate 3 indexes 6. commit work (Table has only about 2000 rows and only occupies 8 pages) We are getting errors SQLCODE: -289 ISAM ERROR CODE: -113, could not lock table. Looking at the logical log entries, the btree cleaner thread kicks of right after the HDELETE and DELITEM entries. Any thoughts? Lyzander * Sent from RemarQ http://www.remarq.com The Internet's Discussion Network * The fastest and easiest way to search and participate in Usenet - Free!