James Bottomley wrote:
quoted text
>>> However, there's still devloss_tmo to consider ... even in
>>> multipath, I don't think you want to signal path failure until
>>> devloss_tmo has fired otherwise you'll get too many transient up/down
>>> events which damage performance if the array has an expensive failover
>>> model.
>>>
>> Yes. But currently we have a very high failover latency as we always have
>> to wait for the requeued commands to time-out.
>> Hence we're damaging performance on arrays with inexpensive failover.
> 
> If it's a either/or choice between the two that's showing our current
> approach to multi-path is broken.
> 
>>> The other problem is what to do with in-flight commands at the time the
>>> link went down.  With your current patch, they're still stuck until they
>>> time out ... surely there needs to be some type of recovery mechanism
>>> for these?
>>>
>> Well, the in-flight commands are owned by the HBA driver, which should
>> have the proper code to terminate / return those commands with the
>> appriopriate codes. They will then be rescheduled and will be caught
>> like 'normal' IO requests.
> 
> But my point is that if a driver goes blocked, those commands will be
> forced to wait the blocked timeout anyway, so your proposed patch does
> nothing to improve the case for dm anyway ... you only avoid commands
> stuck when a device goes blocked if by chance its request queue was
> empty.


How about my patches to use new transport error values and make the 
iscsi and fc behave the same.

The problem I think Hannes and I are both trying to solve is this:

1. We do not want to wait for dev_loss_tmo seconds for failover.

2. The FC drivers can hook into fast_io_fail_tmo related callouts and 
with that set that tmo to a very low value like a couple of seconds if 
they are using multipath, so failovers are fast. However, there is a bug 
with where when the fast_io_fail_tmo fires requests that made it to the 
driver get failed and returned to the multipath layer, but commands in 
the blocked request queue are stuck in there until dev_loss_tmo fires.

With my patches here (need to be rediffed and for FC I need to handle 
JamesS's comments about not using a new field for the fast_fail_timeout 
state bit):

http://marc.info/?l=linux-scsi&m=117399843216280&w=2
http://marc.info/?l=linux-scsi&m=117399544112073&w=2
http://marc.info/?l=linux-scsi&m=117399844316771&w=2
http://marc.info/?l=linux-scsi&m=117400203324693&w=2
http://marc.info/?l=linux-scsi&m=117400203324690&w=2

For FC we can use the fast_io_fail_tmo for fast failovers, and commands 
will not get stuck in a blocked queue for dev_loss_tmo seconds because 
when the fast_io_fail_tmo fires the target's queues are unblocked and 
fc_remote_port_chkready() ready kicks in (iSCSI does the same with the 
patches in the links). And with the patches if multipath-tools is 
sending its path testing IO it will get a DID_TRANSPORT_* error code 
that it can use to make a decent path failing decision with.
--
unsubscribe notice
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/