I have a problem with iSCSI on AMD64 Xen HVM

[Carsten Heesch]

>> 	int max = 24 /* MAX_SKB_FRAGS + (rx->status <= RX_COPY_THRESHOLD) */;

I've just recompiled XENHVM setting this for a quick test:

> int max = MAX_SKB_FRAGS;


Before, I was receiving said error message a lot; now it's gone. Also, throughput has massively increased!

My test setup was:

2x FreeBSD/amd64 8.2-RC2 (compiled about a week ago), XENHVM kernel.
Both DomU's were running on the same physical machine (CentOS 5.5 x86_64 with Xen 4.3.3 on Intel i7), their xn0 NICs are using the same bridge on Dom0.

I was able to trigger the error message on the receiving side by doing a simple "scp test.bin <target>:/dev/null" from one DomU to the other. Throughput dropped within seconds from ~40MB/sec below 2MB/sec.

Now, with that change in place, throughput remains well above 50MB/sec! (approximates actual disk read speed minus SSH compression/encryption overhead) 


Which would be the right value for max? This was obviously only a quick, dirty test. I haven't got a clue either, where the max=5 came from, but it doesn't seem to be a reasonable value.


Cheers
Carsten