Commit a2327b08 authored Sep 13, 2022 by Shuhei Matsumoto Committed by Tomasz Zawadzki Sep 21, 2022

doc/nvme_multipath: Suggest transport_ack_timeout for fast error detection



Signed-off-by: Shuhei Matsumoto <smatsumoto@nvidia.com>
Change-Id: Icbdac9102920c004c2b88ef371d11488208727df
Reviewed-on: https://review.spdk.io/gerrit/c/spdk/spdk/+/14482


Reviewed-by: Ben Walker <benjamin.walker@intel.com>
Reviewed-by: GangCao <gang.cao@intel.com>
Reviewed-by: Aleksey Marchuk <alexeymar@nvidia.com>
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>

parent 2826ef13

doc/nvme_multipath.md

+3 −0

Original line number	Diff line number	Diff line
		@@ -47,6 +47,9 @@ If the step 3, 4, or 5 fails, the reset reverts to the step 3 and then it is ret
		recovered within `ctrlr_loss_timeout_sec` seconds. If `ctrlr_loss_timeout_sec` is -1, it retries
		indefinitely.

		By default, error detection on a qpair is very slow for TCP and RDMA transports. For fast error
		detection, a global option, `transport_ack_timeout`, is useful.

		### Path Selection

		Multipath mode supports two path selection policies, active-passive or active-active.