在 2017/2/13 下午5:08, panxinhui 写道: > > > 在 2017/2/7 下午2:46, Eric Dumazet 写道: >> On Mon, Feb 6, 2017 at 10:21 PM, panxinhui <xin...@linux.vnet.ibm.com> wrote: >> >>> hi all >>> I do some netperf tests and get some benchmark results. >>> I also attach my test script and netperf-result(Excel) >>> > HI, all > I use loopback interface to run netperf tests, > #tc qd add dev lo root pfifo limit 10000 > #ip link > 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc pfifo state UNKNOWN mode > DEFAULT group default qlen 1000 > link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 > > and put the result in netperf.xlsx(excel) > > It is a 32 vcpus P8 machine, with 32Gib memory. > > This time spinlock is the best one, qspinlock > pvqspinlock. So sad. > This time, I have appiled some optimising patches on pvqspinlock. When there is a high contention, the performance has a good improvement ans is very similar to spinlock.
Result is attached in netperf.xlsx thanks xinhui > thanks > xinhui >>> There are two machine. one runs netserver and the other runs netperf >>> benchmark. 1000Mbps network is connected with them. >>> >>> #ip link infomation >>> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state >>> UNKNOWN mode DEFAULT group default qlen 1000 >>> link/ether ba:68:9c:14:32:02 brd ff:ff:ff:ff:ff:ff >>> >>> According to the results, there is not much performance gap with each other. >>> And as we are only testing the throughput, the pvqspinlock shows the >>> overhead of its pv stuff. but qspinlock shows a little improvement than >>> spinlock. My simple summary in this testcase is >>> qspinlock > spinlock > pvqspinlock. >>> >>> when run 200 concurrent netperf, I paste the total throughput here. >>> >>> concurrent runners| total throughput | variance >>> ------------------------------------------- >>> spinlock | 199 | 66882.8 | 89.93 >>> ------------------------------------------- >>> qspinlock | 199 | 66350.4 | 72.0239 >>> ------------------------------------------- >>> pvqspinlock | 199 | 64740.5 | 85.7837 >>> >>> You could see more data in nerperf.xlsx >>> >>> thanks >>> xinhui >> >> >> Hi xinhui >> >> 1Gbit NIC is too slow for this use case. I would try a 10Gbit NIC at least... >> >> Alternatively, you could use loopback interface. (netperf -H 127.0.0.1) >> >> tc qd add dev lo root pfifo limit 10000 >>
netperf.xlsx
Description: MS-Excel 2007 spreadsheet
