My test restore went without a hitch so I’m pretty confident about that.

Bob,

Not sure I understand how splitting your log into two files helped you, since you didn’t have enough disk space. My guess is that, as an example, you needed 10GB and couldn’t get that on one disk, so you created one log file of 5GB on one disk and another 5GB log file on the second disk. Does that sound right?

Your log file is circular and written to sequentially. Your situation seems to be an interesting one, no question. I don’t want to put words (or anything else) into Paul’s mouth, but from my perspective I would find it trickier for a handful of reasons. First, I have logs in two different locations, making it slightly harder to know where all my database files are. Second, restores have an extra “gotcha” that could go wrong. Third, I wouldn’t know which log files was being used without going to look at it first, which is also extra overhead I wouldn’t have if I used just the one file.

I’m sure I can think of more, but those came to mind first. Thanks for your comment, I will see if Paul can leave a comment as well and let us know about other pitfalls for having more than one log file.

Tom

Chris,

That does sound like a nice feature, and I wouldn’t be surprised if it has been discussed at Microsoft previously. I can look to bring it up again.

Thanks for the comment.

Tom

I’m a DBA that specialises in both Oracle & SQL Server. You might be aware that Oracle allows you to multiplex your “online redo log files” (the equivalent of the transaction log). The Oracle database will therefore maintain all the copies of the redo log, will write to one if the other is unavailable & will improve recoverability if a disk fails. I suppose the bottom line is that it gives us DBAs another option – more flexibilty which is key when faced with disaster scenarios.

Any comments?

I’ve asked Microsoft if there were any plans to implement multiplexing bit no plans yet.

Thanks,

Chris.