I didn't sent out this patch after recent discussion at
http://www.open-fcoe.org/pipermail/devel/2009-April/002276.html, related
excerpt from that email.
"So if you set cmd_per_lun to 3, you want to also add a slave_configure
that increases this to a normal default. 3 commands is not normally the
queue depth we end up with for any FC driver. 3 commands is very low."
This patch was prepared before above discussion on mail list, so ignore
this patch for now unless you see this won't impact performance
adversely.
Vasu
On Wed, 2009-05-06 at 16:36 -0700, Robert Love wrote:
> From: Vasu Dev <[email protected]>
>
> The current cmd_per_lun value is 32 while other FC HBAs such
> as lpfc and qlax are using 3, so reduced this to 3 in fcoe also.
>
> I noticed thrash in using 32 cmd_per_lun in stressing 256 luns
> causing lots of retries, so it seems cmd_per_lun=3 is more
> appropriate value as in other existing FC HBA.
>
> Signed-off-by: Vasu Dev <[email protected]>
> ---
>
> drivers/scsi/fcoe/fcoe.c | 2 +-
> 1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/scsi/fcoe/fcoe.c b/drivers/scsi/fcoe/fcoe.c
> index e69d125..284984e 100644
> --- a/drivers/scsi/fcoe/fcoe.c
> +++ b/drivers/scsi/fcoe/fcoe.c
> @@ -125,7 +125,7 @@ static struct scsi_host_template fcoe_shost_template = {
> .change_queue_depth = fc_change_queue_depth,
> .change_queue_type = fc_change_queue_type,
> .this_id = -1,
> - .cmd_per_lun = 32,
> + .cmd_per_lun = 3,
> .can_queue = FCOE_MAX_OUTSTANDING_COMMANDS,
> .use_clustering = ENABLE_CLUSTERING,
> .sg_tablesize = SG_ALL,
>
> _______________________________________________
> devel mailing list
> [email protected]
> http://www.open-fcoe.org/mailman/listinfo/devel
_______________________________________________
devel mailing list
[email protected]
http://www.open-fcoe.org/mailman/listinfo/devel
