Re: -710 repeated

In article <34F7635C.226C@seanet.com>, joanned@seanet.com wrote: > > > > > -Peter > > > > F.PALMITJAVILA wrote: > > > > > -710 repeated error without structure modification in 7.23. > > > (Table has been dropped or renamed etc...) > > > Informix service says we need to do an update statistics, but we can't > > > do it four times per day!!!!! > > > what happen? > > > > > > Thanks. We also have had a plague -710 errors, on both permanent and temporary tables. We opened an Informix case (ref. 142117, if you want to query them on it ...) and also did our own investigations. Firstly, we noted under 'onstat -g ses' that the user-created temporary tables sometimes listed the same table twice, duplicating the name and partnum. THIS IS A KNOWN BUG IN THE 7.23 ENGINE, REF. 43872. It occurs when a temporary table (explicit or implicit) is indexed AFTER data has been inserted. The trick is to create the table with an explicit CREATE TEMP TABLE statement and create indexes BEFORE any data is inserted (contrary to all known best practice). As regards permanent tables, we were told that any of the Data Definition statements : ALTER (FRAGMENT, INDEX, TABLE) CREATE (INDEX, PROCEDURE, ROLE, SCHEMA, SYNONYM, TRIGGER) DROP (INDEX, PROCDEURE, ROLE, SYNONYM, TRIGGER) RENAME (COLUMN, TABLE) etc ... can (not WILL) cause the error to occur if executed on a table whilst a cursor is defined against that table. MALC P (Tunbridge Wells Equitable) > One of my co-workers was told by Informix that this is a known bug. > It had something to do with attempts to alter a table. If you could > not get an exclusive lock on the table, your alter table would fail of > course....however, the catalogs where updated anyway. I wish I had > a bug number for you, but I wasn't part of the tech support call, I > just heard about it after the fact. > > Good luck! > > -----== Posted via Deja News, The Leader in Internet Discussion ==----- http://www.dejanews.com/ Now offering spam-free web-based newsreading