On 10/28/2014 07:13 PM, Keith Busch wrote: > No comments this time. Did I totally botch this in the wrong direction, > or was the timing bad with the linux con/plumbers that week? > Basically the latter. We (ie hch and myself) discussed this briefly at LPC, and thought it'd be the right step in the right direction. 'Real' multipath support w/ blk-mq would need some more thought, though. What we've been discussing was to redo the entire I/O submission path: a) pass down I/O _without_ cloning on the current path b) clone the request _for retry_ only. with b) covering retry on a different path and partial I/O submission on the current path. That would have the benefit doing away with the current request/bio cloning. Plus we could even model things like 'preferred' path. Now we only need someone to implement this ... Cheers, Hannes -- Dr. Hannes Reinecke zSeries & Storage hare@xxxxxxx +49 911 74053 688 SUSE LINUX Products GmbH, Maxfeldstr. 5, 90409 Nürnberg GF: J. Hawn, J. Guild, F. Imendörffer, HRB 16746 (AG Nürnberg) -- dm-devel mailing list dm-devel@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/dm-devel
