I have an application for which data is being written to many disks simultaneously. I would like to use a MySQL table space on each disk. If one of the disks crashes it is tolerable to lose that data, however, I must continue to write to the other disks. My specific concerns are around the redo/undo log.
It is my understanding that there is a separate undo log for each table space and a single redo log for the CLUSTER of all of the table spaces. Is that correct? 1. What happens if a disk containing a table space ( but not the redo log) crashes. Will MySql continue writing to the other disks. Will the redo log be corrupted at this point or would it continue to be valid with the understanding that there is no more data being added for the crashed disk. Could the redo log be used to restore the data on the crashed disk to the point of the crash? 2. What happens if the disk containing the signal redo log crashes. Does that halt the system until the disk is returned. Is there any way of allowing the system to run without a redo log in this situation? Is there any way to distribute the redo logs to the location of the tablespaces? If I were to put the redo log on a raid disk, I assume that would allow the system to continue running in these scenarios. I would prefer a solution that does not add this cost to the system. Thanks for your help Paul