tag:blogger.com,1999:blog-24359421.post8566295175611633226..comments2023-04-29T09:47:48.006+02:00Comments on PrimeBase XT: Making PBXT Fully DurableUnknownnoreply@blogger.comBlogger4125tag:blogger.com,1999:blog-24359421.post-81360202233297551582007-12-21T14:24:00.000+01:002007-12-21T14:24:00.000+01:00Hi Gigiduru,The engine's transaction log is not th...Hi Gigiduru,<BR/><BR/>The engine's transaction log is not the same as the MySQL binlog. This is the case for all storage engines. So all the data is still written to the binlog.<BR/><BR/>Certainly it would be more efficient if MySQL accessed the engine's transaction log directly for replication purposes. This is an improvement to be considered in the future.Paul McCullaghhttps://www.blogger.com/profile/10476275147692948727noreply@blogger.comtag:blogger.com,1999:blog-24359421.post-49439580597986798842007-12-20T18:14:00.000+01:002007-12-20T18:14:00.000+01:00Then how is this affecting replication? Obviously,...Then how is this affecting replication? Obviously, you need MySQL to be able to read from the binlog and ship the DML to the slaves. You cannot just ship the short DML statements, disregarding the big ones. Cheers!<BR/>GigiduruAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-24359421.post-37805419218074413252007-12-20T16:30:00.000+01:002007-12-20T16:30:00.000+01:00Hi Frank,This will probably not be a case of "one ...Hi Frank,<BR/><BR/>This will probably not be a case of "one size fits all", so I plan to make this a tunable parameter (an amount in bytes). :)Paul McCullaghhttps://www.blogger.com/profile/10476275147692948727noreply@blogger.comtag:blogger.com,1999:blog-24359421.post-19867166805597724002007-12-20T16:20:00.000+01:002007-12-20T16:20:00.000+01:00Very interesting. I am curious to know what criter...Very interesting. I am curious to know what criteria you will use to classify a transaction as "short" vs "long"?<BR/><BR/>FrankFrankhttps://www.blogger.com/profile/04086666898806120300noreply@blogger.com