Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 1 | |
| 2 | |
| 3 | HOWTO for the linux packet generator |
| 4 | ------------------------------------ |
| 5 | |
| 6 | Date: 041221 |
| 7 | |
| 8 | Enable CONFIG_NET_PKTGEN to compile and build pktgen.o either in kernel |
| 9 | or as module. Module is preferred. insmod pktgen if needed. Once running |
Matt LaPlante | 3f6dee9 | 2006-10-03 22:45:33 +0200 | [diff] [blame] | 10 | pktgen creates a thread on each CPU where each thread has affinity to its CPU. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 11 | Monitoring and controlling is done via /proc. Easiest to select a suitable |
| 12 | a sample script and configure. |
| 13 | |
| 14 | On a dual CPU: |
| 15 | |
| 16 | ps aux | grep pkt |
| 17 | root 129 0.3 0.0 0 0 ? SW 2003 523:20 [pktgen/0] |
| 18 | root 130 0.3 0.0 0 0 ? SW 2003 509:50 [pktgen/1] |
| 19 | |
| 20 | |
Matt LaPlante | 2fe0ae7 | 2006-10-03 22:50:39 +0200 | [diff] [blame] | 21 | For monitoring and control pktgen creates: |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 22 | /proc/net/pktgen/pgctrl |
| 23 | /proc/net/pktgen/kpktgend_X |
| 24 | /proc/net/pktgen/ethX |
| 25 | |
| 26 | |
| 27 | Viewing threads |
| 28 | =============== |
| 29 | /proc/net/pktgen/kpktgend_0 |
| 30 | Name: kpktgend_0 max_before_softirq: 10000 |
| 31 | Running: |
| 32 | Stopped: eth1 |
| 33 | Result: OK: max_before_softirq=10000 |
| 34 | |
Matt LaPlante | 3f6dee9 | 2006-10-03 22:45:33 +0200 | [diff] [blame] | 35 | Most important the devices assigned to thread. Note! A device can only belong |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 36 | to one thread. |
| 37 | |
| 38 | |
| 39 | Viewing devices |
| 40 | =============== |
| 41 | |
| 42 | Parm section holds configured info. Current hold running stats. |
| 43 | Result is printed after run or after interruption. Example: |
| 44 | |
| 45 | /proc/net/pktgen/eth1 |
| 46 | |
| 47 | Params: count 10000000 min_pkt_size: 60 max_pkt_size: 60 |
| 48 | frags: 0 delay: 0 clone_skb: 1000000 ifname: eth1 |
| 49 | flows: 0 flowlen: 0 |
| 50 | dst_min: 10.10.11.2 dst_max: |
| 51 | src_min: src_max: |
| 52 | src_mac: 00:00:00:00:00:00 dst_mac: 00:04:23:AC:FD:82 |
| 53 | udp_src_min: 9 udp_src_max: 9 udp_dst_min: 9 udp_dst_max: 9 |
| 54 | src_mac_count: 0 dst_mac_count: 0 |
| 55 | Flags: |
| 56 | Current: |
| 57 | pkts-sofar: 10000000 errors: 39664 |
| 58 | started: 1103053986245187us stopped: 1103053999346329us idle: 880401us |
| 59 | seq_num: 10000011 cur_dst_mac_offset: 0 cur_src_mac_offset: 0 |
| 60 | cur_saddr: 0x10a0a0a cur_daddr: 0x20b0a0a |
| 61 | cur_udp_dst: 9 cur_udp_src: 9 |
| 62 | flows: 0 |
| 63 | Result: OK: 13101142(c12220741+d880401) usec, 10000000 (60byte,0frags) |
| 64 | 763292pps 390Mb/sec (390805504bps) errors: 39664 |
| 65 | |
Matt LaPlante | 5d3f083 | 2006-11-30 05:21:10 +0100 | [diff] [blame] | 66 | Configuring threads and devices |
| 67 | ================================ |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 68 | This is done via the /proc interface easiest done via pgset in the scripts |
| 69 | |
| 70 | Examples: |
| 71 | |
| 72 | pgset "clone_skb 1" sets the number of copies of the same packet |
| 73 | pgset "clone_skb 0" use single SKB for all transmits |
| 74 | pgset "pkt_size 9014" sets packet size to 9014 |
| 75 | pgset "frags 5" packet will consist of 5 fragments |
| 76 | pgset "count 200000" sets number of packets to send, set to zero |
Adrian Bunk | d0f19d8 | 2006-06-30 18:28:43 +0200 | [diff] [blame] | 77 | for continuous sends until explicitly stopped. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 78 | |
| 79 | pgset "delay 5000" adds delay to hard_start_xmit(). nanoseconds |
| 80 | |
| 81 | pgset "dst 10.0.0.1" sets IP destination address |
| 82 | (BEWARE! This generator is very aggressive!) |
| 83 | |
| 84 | pgset "dst_min 10.0.0.1" Same as dst |
| 85 | pgset "dst_max 10.0.0.254" Set the maximum destination IP. |
| 86 | pgset "src_min 10.0.0.1" Set the minimum (or only) source IP. |
| 87 | pgset "src_max 10.0.0.254" Set the maximum source IP. |
| 88 | pgset "dst6 fec0::1" IPV6 destination address |
| 89 | pgset "src6 fec0::2" IPV6 source address |
| 90 | pgset "dstmac 00:00:00:00:00:00" sets MAC destination address |
| 91 | pgset "srcmac 00:00:00:00:00:00" sets MAC source address |
| 92 | |
| 93 | pgset "src_mac_count 1" Sets the number of MACs we'll range through. |
| 94 | The 'minimum' MAC is what you set with srcmac. |
| 95 | |
| 96 | pgset "dst_mac_count 1" Sets the number of MACs we'll range through. |
| 97 | The 'minimum' MAC is what you set with dstmac. |
| 98 | |
| 99 | pgset "flag [name]" Set a flag to determine behaviour. Current flags |
| 100 | are: IPSRC_RND #IP Source is random (between min/max), |
| 101 | IPDST_RND, UDPSRC_RND, |
| 102 | UDPDST_RND, MACSRC_RND, MACDST_RND |
Francesco Fondelli | f0e82fd | 2006-09-27 16:33:05 -0700 | [diff] [blame] | 103 | MPLS_RND, VID_RND, SVID_RND |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 104 | |
| 105 | pgset "udp_src_min 9" set UDP source port min, If < udp_src_max, then |
| 106 | cycle through the port range. |
| 107 | |
| 108 | pgset "udp_src_max 9" set UDP source port max. |
| 109 | pgset "udp_dst_min 9" set UDP destination port min, If < udp_dst_max, then |
| 110 | cycle through the port range. |
| 111 | pgset "udp_dst_max 9" set UDP destination port max. |
| 112 | |
Steven Whitehouse | ca6549a | 2006-03-23 01:10:26 -0800 | [diff] [blame] | 113 | pgset "mpls 0001000a,0002000a,0000000a" set MPLS labels (in this example |
| 114 | outer label=16,middle label=32, |
| 115 | inner label=0 (IPv4 NULL)) Note that |
| 116 | there must be no spaces between the |
| 117 | arguments. Leading zeros are required. |
| 118 | Do not set the bottom of stack bit, |
Matt LaPlante | fa00e7e | 2006-11-30 04:55:36 +0100 | [diff] [blame] | 119 | that's done automatically. If you do |
Steven Whitehouse | ca6549a | 2006-03-23 01:10:26 -0800 | [diff] [blame] | 120 | set the bottom of stack bit, that |
| 121 | indicates that you want to randomly |
| 122 | generate that address and the flag |
| 123 | MPLS_RND will be turned on. You |
| 124 | can have any mix of random and fixed |
| 125 | labels in the label stack. |
| 126 | |
| 127 | pgset "mpls 0" turn off mpls (or any invalid argument works too!) |
| 128 | |
Francesco Fondelli | f0e82fd | 2006-09-27 16:33:05 -0700 | [diff] [blame] | 129 | pgset "vlan_id 77" set VLAN ID 0-4095 |
| 130 | pgset "vlan_p 3" set priority bit 0-7 (default 0) |
| 131 | pgset "vlan_cfi 0" set canonical format identifier 0-1 (default 0) |
| 132 | |
| 133 | pgset "svlan_id 22" set SVLAN ID 0-4095 |
| 134 | pgset "svlan_p 3" set priority bit 0-7 (default 0) |
| 135 | pgset "svlan_cfi 0" set canonical format identifier 0-1 (default 0) |
| 136 | |
| 137 | pgset "vlan_id 9999" > 4095 remove vlan and svlan tags |
| 138 | pgset "svlan 9999" > 4095 remove svlan tag |
| 139 | |
| 140 | |
| 141 | pgset "tos XX" set former IPv4 TOS field (e.g. "tos 28" for AF11 no ECN, default 00) |
| 142 | pgset "traffic_class XX" set former IPv6 TRAFFIC CLASS (e.g. "traffic_class B8" for EF no ECN, default 00) |
| 143 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 144 | pgset stop aborts injection. Also, ^C aborts generator. |
| 145 | |
| 146 | |
| 147 | Example scripts |
| 148 | =============== |
| 149 | |
Matt LaPlante | fff9289 | 2006-10-03 22:47:42 +0200 | [diff] [blame] | 150 | A collection of small tutorial scripts for pktgen is in examples dir. |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 151 | |
| 152 | pktgen.conf-1-1 # 1 CPU 1 dev |
| 153 | pktgen.conf-1-2 # 1 CPU 2 dev |
| 154 | pktgen.conf-2-1 # 2 CPU's 1 dev |
| 155 | pktgen.conf-2-2 # 2 CPU's 2 dev |
| 156 | pktgen.conf-1-1-rdos # 1 CPU 1 dev w. route DoS |
| 157 | pktgen.conf-1-1-ip6 # 1 CPU 1 dev ipv6 |
| 158 | pktgen.conf-1-1-ip6-rdos # 1 CPU 1 dev ipv6 w. route DoS |
| 159 | pktgen.conf-1-1-flows # 1 CPU 1 dev multiple flows. |
| 160 | |
| 161 | Run in shell: ./pktgen.conf-X-Y It does all the setup including sending. |
| 162 | |
| 163 | |
| 164 | Interrupt affinity |
| 165 | =================== |
| 166 | Note when adding devices to a specific CPU there good idea to also assign |
| 167 | /proc/irq/XX/smp_affinity so the TX-interrupts gets bound to the same CPU. |
| 168 | as this reduces cache bouncing when freeing skb's. |
| 169 | |
| 170 | |
| 171 | Current commands and configuration options |
| 172 | ========================================== |
| 173 | |
| 174 | ** Pgcontrol commands: |
| 175 | |
| 176 | start |
| 177 | stop |
| 178 | |
| 179 | ** Thread commands: |
| 180 | |
| 181 | add_device |
| 182 | rem_device_all |
| 183 | max_before_softirq |
| 184 | |
| 185 | |
| 186 | ** Device commands: |
| 187 | |
| 188 | count |
| 189 | clone_skb |
| 190 | debug |
| 191 | |
| 192 | frags |
| 193 | delay |
| 194 | |
| 195 | src_mac_count |
| 196 | dst_mac_count |
| 197 | |
| 198 | pkt_size |
| 199 | min_pkt_size |
| 200 | max_pkt_size |
| 201 | |
Steven Whitehouse | ca6549a | 2006-03-23 01:10:26 -0800 | [diff] [blame] | 202 | mpls |
| 203 | |
Linus Torvalds | 1da177e | 2005-04-16 15:20:36 -0700 | [diff] [blame] | 204 | udp_src_min |
| 205 | udp_src_max |
| 206 | |
| 207 | udp_dst_min |
| 208 | udp_dst_max |
| 209 | |
| 210 | flag |
| 211 | IPSRC_RND |
| 212 | TXSIZE_RND |
| 213 | IPDST_RND |
| 214 | UDPSRC_RND |
| 215 | UDPDST_RND |
| 216 | MACSRC_RND |
| 217 | MACDST_RND |
| 218 | |
| 219 | dst_min |
| 220 | dst_max |
| 221 | |
| 222 | src_min |
| 223 | src_max |
| 224 | |
| 225 | dst_mac |
| 226 | src_mac |
| 227 | |
| 228 | clear_counters |
| 229 | |
| 230 | dst6 |
| 231 | src6 |
| 232 | |
| 233 | flows |
| 234 | flowlen |
| 235 | |
| 236 | References: |
| 237 | ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/ |
| 238 | ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/examples/ |
| 239 | |
| 240 | Paper from Linux-Kongress in Erlangen 2004. |
| 241 | ftp://robur.slu.se/pub/Linux/net-development/pktgen-testing/pktgen_paper.pdf |
| 242 | |
| 243 | Thanks to: |
| 244 | Grant Grundler for testing on IA-64 and parisc, Harald Welte, Lennert Buytenhek |
| 245 | Stephen Hemminger, Andi Kleen, Dave Miller and many others. |
| 246 | |
| 247 | |
Steven Whitehouse | ca6549a | 2006-03-23 01:10:26 -0800 | [diff] [blame] | 248 | Good luck with the linux net-development. |