On Thu, Oct 30, 2008 at 08:17:04PM -0400, Thor Lancelot Simon wrote: > On Thu, Oct 30, 2008 at 01:28:21PM -0700, Bill Stouder-Studenmund wrote: > > On Wed, Oct 29, 2008 at 05:51:09PM -0400, Thor Lancelot Simon wrote: > > > > The problem is that this won't help. Ordered tags will relate the > > sequencing of commands relative to each other. The journal, however, > > doesn't care about the relative ordering of operations, it wants to know > > when the writes to the journal have hit stable storage. > > > > The key problem is that, on SCSI disks with the write cache enabled, a > > write command can complete by writing to the cache. > > But SCSI disks don't lie like this unless explicitly configured to, and > with proper use of tags, there is no need to configure them that way; > there is no performance benefit. Ok, you want to have the write cache off. How exactly then does the barrier help? The kernel will know when the journal write(s) are done, and it can then issue the asynchronous writes to the blocks. With the cache off, we won't get status until the bits hit the media, so we can maintain correct two-stage commit semantics. My biggest concern with the tag use is that your barrier mixes in with ALL the outstanding i/o. Even reads. Take care, Bill
Attachment:
pgpX_3aCHHKUi.pgp
Description: PGP signature