logical log space consumption

Folks, Just want to make sure. The move-table job, alter fragment on table mytable init in mynewdbs, would generate lot less logical log data compared to using unload, load, or dbload ....... types of data migration jobs , Correct? Thanks, Frank --089e0149cede1ccc2a04dbd6bc6f

Nope. all will use about the same log space. dbload will use the fewest locks and be least likely to trigger a long transaction rollback. But the ALTER will be much faster. Art Art S. Kagel Advanced DataTools (www.advancedatatools.com) Blog: http://informix-myview.blogspot.com/ Disclaimer: Please keep in mind that my own opinions are my own opinions and do not reflect on my employer, Advanced DataTools, the IIUG, nor any other organization with which I am associated either explicitly, implicitly, or by inference. Neither do those opinions reflect those of other individuals affiliated with any entity with which I am affiliated nor those of the entities themselves. On Fri, May 3, 2013 at 5:07 PM, FRANK <yunyaoqu@gmail.com> wrote: > Folks, > > Just want to make sure. > > The move-table job, > > alter fragment on table mytable init in mynewdbs,> > would generate lot less logical log data compared to using > > unload, load, or dbload ....... types of data migration jobs , Correct? > > Thanks, > Frank > > --089e0149cede1ccc2a04dbd6bc6f > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --001a11c20c9adfca7404dbd7189f

Art, For load or dbload...., I think the logical log will have each row insertion recorded in it to be able to roll back later. So the logical log growing is proportional to the number of rows inserted.( We are talking about big table, like, millions, hundred millions or more of rows) But, for "alter fragment on table mytable init in mynewdbs", the logical log does not have to record each row info . If rollback needed, It can just drop the new data created so far there and keep the original table( I assume the original table will be touched and dropped only after the whole ALTER FRAGMENT is finished successfully) . So, the rollback should be fast, no need to walk through row by row as undoing load or dbload.... Missing some thing ? Thanks, Frank On Fri, May 3, 2013 at 5:33 PM, Art Kagel <art.kagel@gmail.com> wrote: > Nope. all will use about the same log space. dbload will use the fewest > locks and be least likely to trigger a long transaction rollback. But the > ALTER will be much faster. > > Art > > Art S. Kagel > Advanced DataTools (www.advancedatatools.com) > Blog: http://informix-myview.blogspot.com/ > > Disclaimer: Please keep in mind that my own opinions are my own opinions > and do not reflect on my employer, Advanced DataTools, the IIUG, nor any > other organization with which I am associated either explicitly, > implicitly, or by inference. Neither do those opinions reflect those of > other individuals affiliated with any entity with which I am affiliated nor > those of the entities themselves. > > On Fri, May 3, 2013 at 5:07 PM, FRANK <yunyaoqu@gmail.com> wrote: > > > Folks, > > > > Just want to make sure. > > > > The move-table job, > > > > alter fragment on table mytable init in mynewdbs,> > > > would generate lot less logical log data compared to using > > > > unload, load, or dbload ....... types of data migration jobs , Correct? > > > > Thanks, > > Frank > > > > --089e0149cede1ccc2a04dbd6bc6f > > > > > > > > > > ******************************************************************************* > > Forum Note: Use "Reply" to post a response in the discussion forum. > > > > > > --001a11c20c9adfca7404dbd7189f > > > > ******************************************************************************* > Forum Note: Use "Reply" to post a response in the discussion forum. > > --089e0149ccde6127fb04dbd94bee