hi, actually my boss wanted me to use SeaStar framework because of the good benchmark results they have achieved ... but as you guessed I am just trying to understand it's methods and make its tools regarding my project needs up and running ... that I have n't succeeded so far for memcached tool. I have to mention that one of this project objectives is setup a server to listen to a Udp port and read and process data packets as fast as 5 million packets per second. I have reached to 250,000 packets per second so far using seastar udp-server/client code with DPDK enabled and network-stack = native parameters.
On Tuesday, October 31, 2017 at 10:33:38 AM UTC+3:30, Dormando wrote: > > No idea what that is. > > Do you really need something that fast (which doesn't seem to be > supported?). Thread scalability's gotten better in mainline since they did > their bench, it seems. > > If you're just playing around, sorry. I'm not familiar with the thing. > > On Mon, 30 Oct 2017, [email protected] <javascript:> wrote: > > > Hi, I asked this question in seastar group but no answer yet ... I > thought maybe I can get my answer in here: > > I have stock at the begining trying to use memcached app in seastar: > > I compiled memcache.cc code file in /seastar/apps/memcached folder and I > ran the executable file with these parametes: > > ./a.out --dpdk-pmd --network-stack native --host-ipv4-addr 192.168.1.37 > --gw-ipv4-addr 192.168.1.1 --netmask-ipv4-addr 255.255.255.0 -c 1 > --max-slab-size > > 1 > > just to check how it will behave ... I ran a simple cpp code from > another computer in the network to send random small strings (as UDP > packets) to > > ip:port (192.168.1.37:11211) but the result was this : > > > > k-ipv4-addr 255.255.255.0 -c 1 --max-slab-size 1 > > EAL: Detected 4 lcore(s) > > EAL: Probing VFIO support... > > EAL: PCI device 0000:05:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10fb net_ixgbe > > EAL: PCI device 0000:05:00.1 on NUMA socket -1 > > EAL: probe driver: 8086:10fb net_ixgbe > > EAL: PCI device 0000:0a:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0a:00.1 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0b:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0b:00.1 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0c:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0c:00.1 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0d:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:0d:00.1 on NUMA socket -1 > > EAL: probe driver: 8086:10a7 net_e1000_igb > > EAL: PCI device 0000:12:00.0 on NUMA socket -1 > > EAL: probe driver: 8086:10d3 net_e1000_em > > ports number: 1 > > Port 0: max_rx_queues 4 max_tx_queues 4 > > Port 0: using 1 queue > > LRO is off > > RX checksum offload supported > > TX ip checksum offload supported > > TSO is supported > > TX TCP&UDP checksum offload supported > > Port 0 init ... done: > > Creating Tx mbuf pool 'dpdk_pktmbuf_pool0_tx' [1024 mbufs] ... > > Creating Rx mbuf pool 'dpdk_pktmbuf_pool0_rx' [1024 mbufs] ... > > Port 0: Enabling HW FC > > > > Checking link status > > Created DPDK device > > .........................done > > Port 0 Link Up - speed 1000 Mbps - full-duplex > > seastar memcached v1.0 > > > > > > ======> > > I checked /mnt/hugepages folder and nothing were in there ... > > when I passed this address (that is the address I specify to mount huge > pages to be used by DPDK) in front of --hugepage parameter, I get this > error: > > > > k-ipv4-addr 255.255.255.0 -c 1 --hugepages /dev/hugepages/ > > terminate called after throwing an instance of 'std::system_error' > > what(): ftruncate: Invalid argument > > Aborting. > > Backtrace: > > 0x0000000000422e84 > > 0x0000000000422faf > > 0x0000000000423083 > > /lib/x86_64-linux-gnu/libpthread.so.0+0x000000000001138f > > /lib/x86_64-linux-gnu/libc.so.6+0x0000000000035427 > > /lib/x86_64-linux-gnu/libc.so.6+0x0000000000037029 > > /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x000000000008f84c > > /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x000000000008d6b5 > > /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x000000000008d700 > > /usr/lib/x86_64-linux-gnu/libstdc++.so.6+0x000000000008d918 > > 0x00000000004da1bf > > 0x00000000004da27a > > 0x00000000004d8e7d > > 0x00000000004dbeef > > 0x00000000004779be > > 0x0000000000503870 > > 0x000000000041b375 > > /lib/x86_64-linux-gnu/libc.so.6+0x000000000002082f > > 0x000000000041b598 > > Aborted (core dumped) > > > > --------------- > > can any one please help me on this issue to learn how to set up > memcached server in seastar ? I also tried to get the possible data entry > (cached) using > > echo "stats cachedump 15 4" | nc 192.168.1.37 11211 > > that I found after googling reading from memcache (to see if data is > stored or not?) > > > > Thank you so much in advance! > > > > -- > > > > --- > > You received this message because you are subscribed to the Google > Groups "memcached" group. > > To unsubscribe from this group and stop receiving emails from it, send > an email to [email protected] <javascript:>. > > For more options, visit https://groups.google.com/d/optout. > > > > -- --- You received this message because you are subscribed to the Google Groups "memcached" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
