← Back to team overview

drizzle-discuss team mailing list archive

Update on testing the innodb replication log



I started the work to evaluate the innodb trx log:

At present, we are stuck due to this bug:
transaction_reader utility dumping generated SQL out of order when using
innodb trx log

When we call the transaction_reader, the generated SQL is in craaaazy order-
we are altering tables that haven't been created in schemas that don't exist
yet : (

Once we have a means of extracting data from the log so that it can populate
a validation server, we can proceed with making sure that the log accurately
reflects the state of the master.  Until then, we are a bit stuck.


Follow ups