Open Peter-JanGootzen opened 1 year ago
Latency has now also regressed somehow.
statfs
50000 runs numactl pinned to right NUMA node and interrupt core
XLIO + libnfs -O2 used to be: 82usec
NFS: 48usec
Host<>DPU: 34usec
Now: Without XLIO: 160usec With XLIO + libnfs -O2: 137usec So we lost 55usec somewhere! Further testing: With XLIO + libnfs -O3: 126usec With XLIO + libnfs -O3 + 2 Virtio queues: 112usec With XLIO + libnfs -O3 + 2 Virtio queues + DPU pinning to core 6,7: 100usec
--enable-latency-measuring Reports that with the last configuration: NFS: 68usec Host<>DPU: 32usec
TODO:
It would be interesting to test whether doing the NFS polling on the Virtio queue polling thread would improve performance. This requires a non-existent libnfs patch.
TODO disable c-states p-states
Currently with XLIO sequential write performance with
bs=4k, iodepth=16, numjobs=1
is ~244MB/s, while sequential read only gives ~20MB/s.Furthermore blocksizes larger than the page size (4k) don't do anything (i.e. no performance increase).