|
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] RE: Status summary on multiple connections> This may be a reasonable means of operation for the disk world. > It is woefully inadequate for the tape world, as follows: > > - In general, command ordering is crucial - out of order command > processing will lead to data corruption. > - This would require the initiator backup application to block on > completion of every single write command of a backup operation before > issuing the next command. > > If this blocking were performed, both the throughput and capacity of a tape > device/media would be negatively impacted by an order of magnitude or more. > This would occur even assuming an instantaneous transport. Sorry Joe, You folks already have given up on ordering. Your buffering structure is designed to perform lying writes. You gather in a command, gather in its data, post the completion and then pick up the next command. No present day tape drive except a few of the highly specialized Fibre Channel implementations allows ordered command queuing to be invoked from the host adapter and no driver exploits it because it is so rarely implemented. It is only with Rob Basham's new tape command set proposal that you are really going to get going full speed ahead into command queuing, and that is because the ordering information is contained in the command and the data can be pulled down as desired. The other major tape feature is copy management. That works perfectly, because the recipient or source of data is always a disk drive that is being managed by a tape drive. The tape can manage any re-ordering necessary. Bob
Home Last updated: Tue Sep 04 01:06:55 2001 6315 messages in chronological order |