FUJITA Tomonori wrote:
It's not. tgt properly handles SCSI ordering rule. The patch is about I/O threads policy. I/O threads simultaneously performs multiple SCSI commands that can be performed in any order.
I see, I guess I have some gap here understanding why and when its
allowed to perform SCSI commands in any order.
I/O threads uses FILO so if I/O threads has tons of requests, the latency of the first request could be large.
Yes, Alex had a setup with large number of IOPS and as of the FILO
(LIFO) processing such some commands processing was delayed and delayed
till the initiator started to send aborts, merging your FIFO patch
solved it.
Or.
--
To unsubscribe from this list: send the line "unsubscribe stgt" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html