RE: In-place alter table

See "Informix Guide to SQL: Syntax", section on "ALTER TABLE" stmt, subsection "Scope of the In-Place Alter Algorithm" (p. 1-49 in my copy). This is for 7.2x. I understand that 7.3x has much fewer restrictions for getting an in-place alter to occur. ================================ Paul A. Mosser, Open Systems DBA Informix Certified DBSA Wells Fargo & Co. Tempe, Arizona paul.mosser@wellsfargo.com ================================ -----Original Message----- From: Carlson@WHSmith [mailto:carlson1@bellsouth.net] Sent: Wednesday, February 03, 1999 3:49 AM To: informix-list@iiug.org Subject: Re: In-place alter table Just started and learned that the in-place alter table will not work as expected when you modify a column definition (change from char(10) to char(22)). Looks like the ol' unload / mod schema / recreate tables / reload syndrome. Are there any guidelines concerning when an alter table will do an in-place alter? TIA John Carlson Informix DBA WHSmith USA Carlson@WHSmith wrote: > > IDS 7.23.UC11 > HPUX 10.2 > > Due to various requirements, I will need to change the length of a char > column in various tables in my system. I understand that the alter > table will make this change in-place on this version. Since I haven't > done this before, I'm wondering if there are any gotchas out there. > > 1. What are the log space requirements for a database using buffered > logging? > 2. Is is better to force completion of the physical updates to the > table? > > I've scanned Dejanews and haven't found much. > > Still sounds better than an unload, drop, recreate, reload . . . > anything you all could add would be great. > > TIA > > John Carlson > Informix DBA > WHSmith USA