Re: Proposal: B_ARRIER (addresses wapbl performance?)

[Jason Thorpe <thorpej%shagadelic.org@localhost>]

On Dec 9, 2008, at 11:00 AM, Manuel Bouyer wrote:

> But I don't get why we want to have the journal on stable storage  
> *now*.
> > From what I understand, what we want is to have the journal entry for
> this transaction to stable storage *before* the metadata changes  
> start to hit
> stable storage, and the journal cleanup hits stable storage *after*
> all metadata changes are on stable storage. I can't see what FUA  
> brings
> us here, as long as writes to stable storage are properly ordered.

You want your journal to remain self-consistent, otherwise you can't  
trust it to replay it.

Also, without explicit cache management, you can't be sure when the  
data gets written out of the drive's cache.  Again, command completion  
has nothing to do with how the writes are ordered to the oxide.

-- thorpej