Updated version of utils2_ak posted.

What would we do without you Art. Thanks for all your work Peter >From: "Art S. Kagel" <kagel@bloomberg.net> >To: informix-list@iiug.org >Subject: Updated version of utils2_ak posted. >Date: Tue, 10 Aug 1999 09:49:47 -0400 >Organization: Bloomberg LP >Reply-To: kagel@bloomberg.net >Sender: owner-informix-list@iiug.org > > ** NOTICE ** NOTICE ** NOTICE ** NOTICE ** NOTICE ** NOTICE ** NOTICE ** > >To all users of myschema.ec and dbcopy.ec: > >I have submitted, and Walt has posted, a new version of utils2_ak which >contains several updated utilities including new features in the ul.ec >utility. > >HOWEVER, there is a SERIOUS and insidious bug in previously posted >versions of dbcopy.ec. If you use the -F option to speed processing in >combination with the -i option to ignore duplicate key errors dbcopy fails >to re-PUT any rows in the PUT buffer which follow the row in error. This >will result in the loss of data as these rows will have NOT been inserted >by the FLUSH that detected the error. The new release, dated August 9, >1999 in the README.1st file, contains a fixed version that repairs the >problem. The fix cleaned up the error handling code and should result in >a slight improvement in throughput for all runs as a bonus. > >Also included is an update to myschema.ec that fixes two bugs related to >fragmented tables and indexes which may either cause the generated schema >to fail to execute with syntax errors or to be syntactically valid but >incorrect. One bug causes parts of a table's fragmentation expression to >be appended to subsequent fragmented table's expression if the latter's >expression is shorter, this may give unexpected results or fail to >execute. The other bug caused all but the last fragmented index in the >schema to be output as a non-fragmented index. Again the August 9 version >has the fixed version. > >Art S. Kagel > >