# linux arp 参数整理 Date: 2024-06-22 #### **linux arp相关参数** 在 Linux 中，有多个内核参数可以控制 ARP（地址解析协议）的行为。这些参数可以通过 `/proc/sys/net/ipv4/conf/*/` 目录下的文件来配置，或者使用 `sysctl` 命令来设置。以下是一些常用的 ARP 内核参数及其含义： **arp\_ignore**： - - 控制系统对不同网络接口上收到的 ARP 请求的响应行为。 - 取值范围： - 0：在任何网络接口上收到 ARP 请求时，系统都会进行响应（默认值）。 - 1：只有当目标 IP 地址在收到请求的网络接口上配置时，系统才会响应。 - 2：只有当目标 IP 地址是最接近请求源的地址时，系统才会响应 **arp\_announce**： - - 控制系统在发送 ARP 请求时，源 IP 地址的选择策略。 - 取值范围： - **0 (默认):** 允许使用任意网卡上的 IP 地址作为 ARP 请求的源 IP 地址。通常情况下，会使用 IP 数据包的源 IP。 - **1:** 尽量避免使用不属于该发送网卡子网的本地地址作为发送 ARP 请求的源 IP 地址。 - **2:** 忽略 IP 数据包的源 IP 地址，选择该发送网卡上最合适的本地地址作为 ARP 请求的源 IP 地址。 **arp\_accept**： - - 控制系统GARP处理，当IP不在ARP表中。（GARP的源地址和目标地址一样）， 如果存在直接更新不会判断这个标志 - 取值范围： - **0 (默认值):** 始终接受 GARP 数据包并相应地更新 ARP 表。即使 GARP 数据包包含未分配给其任何接口的 IP 地址，系统也会更新其 ARP 缓存。 - **1:** 仅当数据包中的源 IP 地址与分配给系统接口之一的 IP 地址匹配时，才接受 GARP 数据包。这是一种更保守的方法，因为它只更新系统实际使用的地址的 ARP 缓存。 - **2:**（自 Linux 内核 v6.0起）仅当源 IP 地址与接收数据包的接口位于同一子网时，才接受 GARP 数据包。这通过确保仅从同一网络上的设备接受 GARP 数据包来增加另一层安全性。 **arp\_notify；** - - Linux 中的 arp\_notify 参数控制在网络设备启动或其硬件地址（MAC 地址）更改时是否自动发送 Gratuitous ARP (GARP) 请求。GARP 请求是一种特殊的 ARP 数据包，设备发送该数据包是为了通知其 IP 地址到 MAC 地址的映射，即使没有收到 ARP 请求。 - 取值范围： - - - **0 (默认值):** 不自动发送 GARP 请求。 - **1:** 当网络设备启动或其硬件地址更改时，自动发送 GARP 请求。 **drop\_gratuitous\_arp** - - drop\_gratuitous\_arp 是 Linux 内核中的一个参数，用于控制是否丢弃 Gratuitous ARP (GARP) 数据包。GARP 数据包是设备发送的特殊 ARP 数据包，用于宣布其 IP 地址到 MAC 地址的映射，即使没有收到 ARP 请求。 - **取值范围：** - **0 (默认值):** 接受 GARP 数据包并相应地更新 ARP 表。 - **1:** 丢弃所有 GARP 数据包。 #### **验证：** 之前在做LVS实验搭建DR环境的时候用到前面两个参数；**arp\_ignore， arp\_announce** 环境搭建可以参考 https://garlicspace.com/2024/06/02/%e8%b4%9f%e8%bd%bd%e5%9d%87%e8%a1%a1/ lb 10.1.1.10  5a:86:a3:16:85:0 rs1 10.1.1.11  56:ff:71:bd:27:40  veth-rs1 rs2 10.1.1.12  46:15:5c:b7:68:35 client 10.1.1.20 ee:86:01:ae:45:b0   **arp\_ignore设置1，只有目标地址在收到请求的网卡上才回应。** ``` ip netns exec client arp Address HWtype HWaddress Flags Mask Iface 10.1.1.11 ether 56:ff:71:bd:27:40 C veth-client 10.1.1.10 ether 5a:86:a3:16:85:08 C veth-client 10.1.1.12 ether 46:15:5c:b7:68:35 C veth-client ``` 使用 arp -d 清除arp记录， 并且将rs1 的 **arp\_ignore标识设置为1， 由于之前仅设置了net.ipv4.conf.all，修改这个参数即可。** ``` ip netns exec client arp Address HWtype HWaddress Flags Mask Iface 10.1.1.10 ether 56:ff:71:bd:27:40 C veth-client ``` 看下rs1地址信息 ``` # rs1 #ip netns exec rs1 ip a 1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 10.1.1.10/32 scope global lo valid_lft forever preferred_lft forever 264: veth-rs1@if263: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 56:ff:71:bd:27:40 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.11/24 scope global veth-rs1 valid_lft forever preferred_lft forever inet6 fe80::54ff:71ff:febd:2740/64 scope link valid_lft forever preferred_lft forever # client # ip netns exec client ip a 1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 268: veth-client@if267: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether ee:86:01:ae:45:b0 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.20/24 scope global veth-client valid_lft forever preferred_lft forever inet6 fe80::ec86:1ff:feae:45b0/64 scope link valid_lft forever preferred_lft forever ``` 由于**arp\_ignore设置为0， client客户端在做arp请求查询 10.1.1.10时 ，rs1将10.1.1.10在lo配置了VIP，rs1的veth-rs1，应答了arp请求， 于是更新arp信息。** ``` #ip netns exec rs1 tcpdump -i veth-rs1 arp -netvv dropped privs to tcpdump tcpdump: listening on veth-rs1, link-type EN10MB (Ethernet), capture size 262144 bytes ee:86:01:ae:45:b0 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.10 tell 10.1.1.20, length 28 56:ff:71:bd:27:40 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.10 is-at 56:ff:71:bd:27:40, length 28 56:ff:71:bd:27:40 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.20 tell 10.1.1.11, length 28 ee:86:01:ae:45:b0 > 56:ff:71:bd:27:40, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.20 is-at ee:86:01:ae:45:b0, length 28s ``` 设置为1后， 查询10.1.1.10地址，lb 返回5a:xxx地址。后面rs1，rs2查询客户端client地址。 ``` # ip netns exec client tcpdump -i veth-client arp -netvv dropped privs to tcpdump tcpdump: listening on veth-client, link-type EN10MB (Ethernet), capture size 262144 bytes ee:86:01:ae:45:b0 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.10 tell 10.1.1.20, length 28 5a:86:a3:16:85:08 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.10 is-at 5a:86:a3:16:85:08, length 28 10.1.1.20 - - [03/Jun/2024 05:28:21] "GET / HTTP/1.1" 200 - 10.1.1.20 - - [03/Jun/2024 05:28:23] "GET / HTTP/1.1" 200 - 56:ff:71:bd:27:40 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.20 tell 10.1.1.11, length 28 ee:86:01:ae:45:b0 > 56:ff:71:bd:27:40, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.20 is-at ee:86:01:ae:45:b0, length 28 46:15:5c:b7:68:35 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.20 tell 10.1.1.12, length 28 ee:86:01:ae:45:b0 > 46:15:5c:b7:68:35, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.20 is-at ee:86:01:ae:45:b0, length 28 ```   **arp\_announce设置2，发送arp请求时使用发送arp请求网卡的地址， 如果设置0，则可能不是** rs1的网卡配置 ``` 264: veth-rs1@if263: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 56:ff:71:bd:27:40 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.11/24 scope global veth-rs1 valid_lft forever preferred_lft forever inet6 fe80::54ff:71ff:febd:2740/64 scope link valid_lft forever preferred_lft forever 271: veth-rs1-extra@if270: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 66:82:27:71:76:15 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 172.168.1.2/24 scope global veth-rs1-extra valid_lft forever preferred_lft forever inet6 fe80::6482:27ff:fe71:7615/64 scope link valid_lft forever preferred_lft forever ```   通过给rs1添加一个网卡 ``` ip link add veth-rs1-extra type veth peer name veth-rs1-extra-br ip link set veth-rs1-extra netns rs1 ip link set veth-rs1-extra-br master br0 ip link set veth-rs1-extra-br up ip netns exec rs1 ip addr add 172.168.1.2/24 dev veth-rs1-extra ip netns exec rs1 ip link set veth-rs1-extra up ``` 在6.19内核版本是接口名称需要小于15 ``` # ip link add veth-rs1-extra type veth peer name veth-rs1-extra^C # ip a|grep extr # ip link add veth-rs1-extra type veth peer name veth-rs1-extra-br Error: argument "veth-rs1-extra-br" is wrong: "name" not a valid ifname # ip link add veth-rs1-extra type veth peer name veth-rs1-extra RTNETLINK answers: File exists # ip link add veth-rs1-extra type veth peer name veth-rs1-ex-br # ip link set veth-rs1-extra netns rs1 # ip link set veth-rs1-ex-br master br0 # ip link set veth-rs1-ex-br up # ip netns exec rs1 ip addr add 172.168.1.2/24 dev veth-rs1-extra # ip netns exec rs1 ip link set veth-rs1-extra up ```   清除arp记录 ``` ip netns exec client ip -s -s neigh flush all ip netns exec rs1 ip -s -s neigh flush all ``` 设置**arp\_announce=0** ``` ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_announce=0 ``` 使用tcpdump抓包br0，查看发送的arp报文 ``` tcpdump -i br0 arp -netvv ``` 在rs1发包 ``` # ip netns exec rs1 ping -c 1 -I 172.168.1.2 10.1.1.20 PING 10.1.1.20 (10.1.1.20) from 172.168.1.2 : 56(84) bytes of data. --- 10.1.1.20 ping statistics --- 1 packets transmitted, 0 received, 100% packet loss, time 0ms ``` 抓到的包， 可以看到mac是56:ff:71:bd:27:40， veth-rs1， 对应地址172.168.1.2 ``` 56:ff:71:bd:27:40 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.20 tell 172.168.1.2, length 28 ee:86:01:ae:45:b0 > 56:ff:71:bd:27:40, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.20 is-at ee:86:01:ae:45:b0, length 28 ee:86:01:ae:45:b0 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.1 tell 10.1.1.20, length 28 1a:03:e9:42:52:6c > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.1 is-at 1a:03:e9:42:52:6c, length 28sshez ``` 设置2时，再次ping ``` # ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_announce=2 ``` ``` # ip netns exec rs1 ping -c 1 -I 172.168.1.2 10.1.1.20 ``` 这时抓到包，第三行发出的arp已经从172.168.1.2 改为10.1.1.11 ``` ee:86:01:ae:45:b0 > 1a:03:e9:42:52:6c, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.1 tell 10.1.1.20, length 28 1a:03:e9:42:52:6c > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.1 is-at 1a:03:e9:42:52:6c, length 28 56:ff:71:bd:27:40 > ee:86:01:ae:45:b0, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.20 tell 10.1.1.11, length 28 ee:86:01:ae:45:b0 > 56:ff:71:bd:27:40, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Reply 10.1.1.20 is-at ee:86:01:ae:45:b0, length 28 ```   与GARP相关的三个参数 **arp\_accept， arp\_notify， drop\_gratuitous\_arp** 是否接收并更新arp table， 是否发送，是否抛弃。 GARP（Gratuitous ARP): 针对自己的IP和MAC进行广播的ARP，如果普通ARP是查找其他的主机地址，那么这个类型ARP就是主动推销自己。 在刚刚配置完网卡或者重启机器会发送这个GARP， 用于检测地址冲突。 另外在一些协议中如VRRP， HSRP中当主设备发送故障，备用设备接管也是通过发送GARP来更新虚拟IP的地址信息。 sudo arping -U 下面验证一下 ``` # ip netns exec rs1 ip a 1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 10.1.1.10/32 scope global lo valid_lft forever preferred_lft forever 264: veth-rs1@if263: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 56:ff:71:bd:27:40 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.11/24 scope global veth-rs1 valid_lft forever preferred_lft forever inet6 fe80::54ff:71ff:febd:2740/64 scope link valid_lft forever preferred_lft forever 271: veth-rs1-extra@if270: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 66:82:27:71:76:15 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 172.168.1.2/24 scope global veth-rs1-extra valid_lft forever preferred_lft forever inet6 fe80::6482:27ff:fe71:7615/64 scope link valid_lft forever preferred_lft forever # ip netns exec rs1 arp -a ```   ``` ip netns exec rs1 sysctl -w net.ipv4.conf.veth-rs1-extra.arp_accept=0 net.ipv4.conf.veth-rs1-extra.arp_accept = 0 # ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_accept=0 net.ipv4.conf.all.arp_accept = 0 # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 2 probes (2 broadcast(s)) Received 0 response(s) # ip netns exec rs1 arp -a # ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_accept=1 net.ipv4.conf.all.arp_accept = 1 # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 3 probes (3 broadcast(s)) Received 0 response(s) # ip netns exec rs1 arp -a ? (10.1.1.20) at ee:86:01:ae:45:b0 [ether] on veth-rs1 ? (10.1.1.20) at ee:86:01:ae:45:b0 [ether] on veth-rs1-extra ``` 可以看到设置为1 veth-rs1， veth-rs1-extra两个接口上都增加了arp记录， 内核在6.0上支持设置值为2 ， 只在veth-rs1会增加记录。上面验证环境内核是 5.10.134， 下面用 6.1.9内核验证。   ``` # ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_accept=2 net.ipv4.conf.all.arp_accept = 2 # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 3 probes (3 broadcast(s)) Received 0 response(s) # ip netns exec rs1 arp -a ? (10.1.1.20) at 2e:46:eb:2f:33:37 [ether] on veth-rs1 ? (10.1.1.20) at 2e:46:eb:2f:33:37 [ether] on veth-rs1 # ip netns exec rs1 ip a 1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 10.1.1.10/32 scope global lo:0 valid_lft forever preferred_lft forever 6: veth-rs1@if5: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 0e:58:7e:b6:3e:d8 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.11/24 scope global veth-rs1 valid_lft forever preferred_lft forever inet6 fe80::c58:7eff:feb6:3ed8/64 scope link valid_lft forever preferred_lft forever 15: veth-rs1-extra@if14: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 16:54:5c:45:63:31 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 172.168.1.2/24 scope global veth-rs1-extra valid_lft forever preferred_lft forever inet6 fe80::1454:5cff:fe45:6331/64 scope link valid_lft forever preferred_lft forever # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 3 probes (3 broadcast(s)) Received 0 response(s) # ip netns exec rs1 arp -a gateway (10.1.1.1) at 06:aa:f9:6f:4b:42 [ether] on veth-rs1 ? (10.1.1.20) at 2e:46:eb:2f:33:37 [ether] on veth-rs1 ``` > 相关内核提交 > > ``` > # ip netns exec rs1 tcpdump -i veth-rs1 -netvv > 。。。 > 00:11:22:33:44:55 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.13 tell 10.1.1.13, length 28 > 00:11:22:33:44:55 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.11 tell 10.1.1.11, length 28 > > ``` > >   > > https://github.com/torvalds/linux/commit/e68c5dcf0aacc48a23cedcb3ce81b8c60837f48c   arp\_notify；机器重启或mac地址更换是否发送arp 设置arp\_notify ``` # ip netns exec rs1 sysctl -w net.ipv4.conf.all.arp_notify=1 ``` 更换mac ``` # rs1的veth-rs1更换mac # ip netns exec rs1 ip link set dev veth-rs1 address 00:11:22:33:44:55 [root@iZ8vbd88lmglnbsnad85q3Z ~]# ip netns exec rs1 ip a 1: lo: mtu 65536 qdisc noop state DOWN group default qlen 1000 link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00 inet 10.1.1.10/32 scope global lo valid_lft forever preferred_lft forever 264: veth-rs1@if263: mtu 1500 qdisc noqueue state UP group default qlen 1000 link/ether 00:11:22:33:44:55 brd ff:ff:ff:ff:ff:ff link-netnsid 0 inet 10.1.1.13/24 scope global veth-rs1 valid_lft forever preferred_lft forever inet 10.1.1.11/24 scope global secondary veth-rs1 valid_lft forever preferred_lft forever inet6 fe80::54ff:71ff:febd:2740/64 scope link valid_lft forever preferred_lft forever ``` ``` # ip netns exec rs1 tcpdump -i veth-rs1 -netvv 00:11:22:33:44:55 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.13 tell 10.1.1.13, length 28 00:11:22:33:44:55 > Broadcast, ethertype ARP (0x0806), length 42: Ethernet (len 6), IPv4 (len 4), Request who-has 10.1.1.11 tell 10.1.1.11, length 28 ``` 通过tcpdump可以看到garp报文， 如果设置为0则不发送。 > 如果修改ip这个参数是不起作用的, redhat 有个修复是在ifup脚本中增加arping的指令发送garp > > https://bugzilla.redhat.com/show\_bug.cgi?id=1320366 > > https://github.com/fedora-sysv/initscripts/pull/24/commits/e7b5d95cb5d9ad73ce7c37b06dbf058034a95c72 > > https://medium.com/opsops/arp-stale-cache-and-a-rather-peculiar-arp-notify-behavior-in-linux-9c5105d07a49   **drop\_gratuitous\_arp**  https://patchwork.kernel.org/project/linux-wireless/patch/1454589080-21354-2-git-send-email-johannes@sipsolutions.net/ 验证这个功能需要一个名字空间就可以   ``` #查看arp table ip netns exec rs1 arp -a #如果存在， 使用arp -d删除 ? (10.1.1.20) at 00:11:22:33:55:99 [ether] on veth-rs1 ? (10.1.1.20) at 00:11:22:33:55:99 [ether] on veth-rs1-extra #ip netns exec rs1 arp -d 10.1.1.20 -i veth-rs1 #ip netns exec rs1 arp -d 10.1.1.20 -i veth-rs1-extra ip netns exec rs1 sysctl net.ipv4.conf.all.drop_gratuitous_arp net.ipv4.conf.all.drop_gratuitous_arp = 0 ``` ``` # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 1 probes (1 broadcast(s)) Received 0 response(s) [root@iZ8vbd88lmglnbsnad85q3Z ~]# ip netns exec rs1 arp -a ? (10.1.1.20) at 00:11:22:33:55:99 [ether] on veth-rs1 ? (10.1.1.20) at 00:11:22:33:55:99 [ether] on veth-rs1-extras ``` 设置**drop\_gratuitous\_arp=1 ， 并清理arptable** ``` ip netns exec rs1 sysctl -w net.ipv4.conf.all.drop_gratuitous_arp=1 net.ipv4.conf.all.drop_gratuitous_arp = 1 # ip netns exec rs1 arp -d 10.1.1.20 -i veth-rs1 # ip netns exec rs1 arp -d 10.1.1.20 -i veth-rs1-extra # ip netns exec rs1 arp -a #再次发送garp， arp表未更新 # ip netns exec client arping -U 10.1.1.20 ARPING 10.1.1.20 from 10.1.1.20 veth-client ^CSent 2 probes (2 broadcast(s)) Received 0 response(s) # ip netns exec rs1 arp -a ```   #### **参考及引用** 关于arp 介绍 https://cizixs.com/2017/07/31/arp-protocol/ 关于GARP https://www.practicalnetworking.net/series/arp/gratuitous-arp/ 图片from[曾彥博](https://www.facebook.com/groups/474930279315210/user/100021216714168/?__cft__[0]=AZWq16JCilSacwd9qCrV9323ibGFl3e6_sAmVI5SQsl7dQ1K2FNZr60Gq_nztbgaXGmgVSOUfy35A1UiLSpcAM7Q18KP3KB8y5O4SZsFARITk62OoaODX1eQTPyh_cnDiIOoYnDWtCzppqgEeTizrssyou2lgHCC7Y2sPL4CLo5JrDJ_c0hLa9iPf2AxbuTE0Ie7d3sS-RpJo7BbIZWBVeLS&__tn__=-UC*F)