On 2/28/2016 11:50 AM, Sagi Grimberg wrote:
I think you're right. Do you have any good suggestion how to trigger
iSER first burst data, as apparently my normal testing didn't hit
this?
The first burst should happen by default (against LIO which support
ImmediateData and UnsolicitedDataOut). But it won't make a difference
for the initiator which registers the entire buffer, sends the first
burst and let the target read the rest accordingly...
Perhaps if you change the iscsi tpg FirstBurstLength to be subpage say
3k (default is 64k) you can get the isert (when using MRs over iwarp)
hit this...
Steve can you help?
I can try this out. Did this discussion result in needing a code change
though?
Also, one acid test I do that produced non-zero initial offsets for
nfsrdma server was mounting a large ramdisk to the client and build the
kernel with 'make -j 16 O=blah' with O= pointing to the mounted remote
ramdisk. nfsrdma is very different from iser/block protocols but its
still a good (evil) test. :)
--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
