|
ipvs local node setup is broken since kernel 2.6.28
test setup is as following:
MASTER/RS1 (IP 10.0.0.20/24, VIP 10.0.0.1):
eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state
UNKNOWN qlen 2000
link/ether 00:e0:4c:b5:c7:a2 brd ff:ff:ff:ff:ff:ff
inet 10.0.0.20/24 brd 10.0.1.255 scope global eth0
inet 10.0.0.1/32 scope global eth0
FWM 5 wlc
-> 10.0.0.20:3306 Local 1 0 0
-> 10.0.0.30:3306 Route 1 0 0
RS1 ~ # sysctl -a | egrep "\.(rp_fi|ip_forw)"
net.ipv4.conf.all.rp_filter = 0
net.ipv4.conf.all.arp_filter = 0
net.ipv4.conf.all.arp_announce = 2
net.ipv4.conf.all.arp_ignore = 1
net.ipv4.conf.all.arp_accept = 0
net.ipv4.conf.default.rp_filter = 0
net.ipv4.conf.default.arp_filter = 0
net.ipv4.conf.default.arp_announce = 2
net.ipv4.conf.default.arp_ignore = 1
net.ipv4.conf.default.arp_accept = 0
net.ipv4.conf.lo.rp_filter = 1
net.ipv4.conf.lo.arp_filter = 0
net.ipv4.conf.lo.arp_announce = 2
net.ipv4.conf.lo.arp_ignore = 1
net.ipv4.conf.lo.arp_accept = 0
net.ipv4.conf.eth0.rp_filter = 1
net.ipv4.conf.eth0.arp_filter = 0
net.ipv4.conf.eth0.arp_announce = 2
net.ipv4.conf.eth0.arp_ignore = 1
net.ipv4.conf.eth0.arp_accept = 0
net.ipv4.ip_forward = 0
RS1 ~ # iptables -L -nvt mangle
Chain PREROUTING (policy ACCEPT 34M packets, 3959M bytes)
pkts bytes target prot opt in out source
destination
0 0 MARK tcp -- * * 0.0.0.0/0
10.0.0.1 tcp dpt:3306 MARK xset 0x5/0xffffffff
RS2 (IP 10.0.0.30/24):
lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 10.0.0.1/32 scope global lo:1
eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state
UNKNOWN qlen 2000
link/ether 00:17:31:8d:58:3d brd ff:ff:ff:ff:ff:ff
inet 10.0.0.30/24 brd 10.0.0.255 scope global eth0
from MASTER trying to connect to VIP (local process tries to connect
to VIP on local nic):
telnet -b 10.0.0.20 10.0.0.1 3306 (no connect)
and tcpdump show (tcpdump -i eth0 -n -e -s0 -v host 10.0.0.1):
19:22:05.809708 00:e0:4c:b5:c7:a2 > 00:17:31:8d:58:3d, ethertype IPv4
(0x0800), length 74: (tos 0x10, ttl 64, id 57858, offset 0, flags [DF],
proto TCP (6), length 60) 10.0.0.20.42779 > 10.0.0.1.3306: S, cksum
0xc2bd (correct), 4115412076:4115412076(0) win 32792 <mss
16396,sackOK,timestamp 1955438 0,nop,wscale 8>
19:22:05.809811 00:17:31:8d:58:3d > 00:e0:4c:b5:c7:a2, ethertype IPv4
(0x0800), length 74: (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto
TCP (6), length 60) 10.0.0.1.3306 > 10.0.0.20.42779: S, cksum 0xbe1e
(correct), 920820514:920820514(0) ack 4115412077 win 5792 <mss
1460,sackOK,timestamp 1008245313 1955438,nop,wscale 8>
19:22:09.647727 00:17:31:8d:58:3d > 00:e0:4c:b5:c7:a2, ethertype IPv4
(0x0800), length 74: (tos 0x0, ttl 64, id 0, offset 0, flags [DF], proto
TCP (6), length 60) 10.0.0.1.3306 > 10.0.0.20.42779: S, cksum 0xba5e
(correct), 920820514:920820514(0) ack 4115412077 win 5792 <mss
1460,sackOK,timestamp 1008246273 1955438,nop,wscale 8>
as can be seen, packets goes through ipvs code and forwarded to RS2 and
REPLIED by RS2 but then dropped somewhere in iptables/ipvs code !
_______________________________________________
Please read the documentation before posting - it's available at:
http://www.linuxvirtualserver.org/
LinuxVirtualServer.org mailing list - lvs-users@xxxxxxxxxxxxxxxxxxxxxx
Send requests to lvs-users-request@xxxxxxxxxxxxxxxxxxxxxx
or go to http://lists.graemef.net/mailman/listinfo/lvs-users
|
