tag:blogger.com,1999:blog-5237676719536842801.post9006951480510751971..comments2024-03-28T08:25:10.999+04:00Comments on Dive in Oracle: Commit After n UpdatesMahmoud A. Elsayedhttp://www.blogger.com/profile/11349719211814162336noreply@blogger.comBlogger3125tag:blogger.com,1999:blog-5237676719536842801.post-63491764169467369982012-07-01T09:34:02.273+04:002012-07-01T09:34:02.273+04:00You ca use "FOR UPDATE" in cursor & ...You ca use "FOR UPDATE" in cursor & current of clause in update statement then after one record update it automaically commit & also hole table not in lock state only one record at a time in lock state.<br /><br /><br /><br /><br />Pushpendra ChoubeyAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-5237676719536842801.post-39454940045788095442012-05-25T15:59:44.224+04:002012-05-25T15:59:44.224+04:00i think for loop have implicit close cursor, so we...i think for loop have implicit close cursor, so we don't need to close cursor explicit in for loop, am i right?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-5237676719536842801.post-54537612064322530522012-05-17T23:25:30.163+04:002012-05-17T23:25:30.163+04:00You are still ignoring the problem caused by keepi...You are still ignoring the problem caused by keeping your cursor open.<br /><br />Unless you close/reopen the cursor, you still need the UNDO space to hold the cursor's initial state. This is required by Oracle's read-consistency mechanisms. If you do not have the UNDO space for a single statement, you are sure to run into an ORA-1555 error (snapshot too old).<br /><br />Also, you mention "overloading redo log file". Your code will produce a lot more REDO because you have to track the UNDO (redo and undo are not the same thing).Anonymousnoreply@blogger.com