On Mon, Dec 19, 2011 at 4:03 AM, David Dillow <[email protected]> wrote: > On Thu, 2011-12-01 at 20:13 +0100, Bart Van Assche wrote: >> Make it possible to disconnect via sysfs the IB RC connection used by the >> SRP protocol to communicate with a target. >> >> Let the SRP transport layer create a sysfs "delete" attribute for >> initiator drivers that support this functionality. > > This will be a nice feature to have, but I wonder about the proper > behavior. I had thought about having a 'drain' state, where we wouldn't > start new commands and would let existing ones complete or timeout > before killing the connection. That may be in addition to a 'kill it > now' variant. > > Though perhaps my gentler delete is best done by having a user script > walk all of the devices under the host and set them to offline, wait for > scsi_host/hostX/host_busy to drop to zero, then hit the SRP delete attr.
Sorry, but I disagree. One of the goals of this patch set is to make fast fail-over possible. Waiting for an error completion or time out contradicts with that goal. Bart. -- To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
