I successfully have two vlans running over one physical interface
connected to my managed switch (a PowerConnect 5224), but I can't get
the same two vlans to work when running over a trunk interface spanning
four physical interfaces.Before: (this works, but only uses one physical interface)
---------
# cat hostname.gem0
inet xx.xxx.xxx.x 255.255.255.0 NONE [public ip masked]# cat hostname.hme0
up# cat hostname.vlan2
inet 10.0.2.1 255.255.255.0 NONE vlan 2 vlandev hme0# cat hostname.vlan3
inet 10.0.3.1 255.255.255.0 NONE vlan 3 vlandev hme0After: (this doesn't work)
--------
# cat hostname.gem0
inet xx.xxx.xxx.x 255.255.255.0 NONE [public ip masked]# cat hostname.hme0
up# cat hostname.hme1
up# cat hostname.hme2
up# cat hostname.hme3
up# cat hostname.trunk0
trunkproto roundrobin trunkport hme0 trunkport hme1 trunkport hme2
trunkport hme3# cat hostname.vlan2
inet 10.0.2.1 255.255.255.0 NONE vlan 2 vlandev trunk0# cat hostname.vlan3
inet 10.0.3.1 255.255.255.0 NONE vlan 3 vlandev trunk0Assuming all looks good above, the changes I made on the switch were:
- create a bogus vlan (id 1111)
- make the four ports untagged members of vlan id 1111
- remove the four ports as members of vlan id 1
- make the four ports be members of trunk 1
- make trunk 1 tagged member of vlans 2 and vlan 3But no traffic gets through. For instance, ssh-ing through the firewall
and switch to target results in "Network is unreachable" during which
pflog shows a match for "pass out on vlan2"; `tcpdump -n -i trunk0`
shows nothing; `tcpdump -n -i vlan2` shows nothing; heck, even `tcpdump
-n -i hme0` shows nothing. Likewise, `tcpdump` on the target shows no
traffic.Any ideas?
Thanks,
Kent
Looking at the output from `ifconfig` (see below), I notice that the
trunk0 doesn't show that its "UP" - why wouldn't it be up?Thanks,
Kent# ifconfig
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33168
groups: lo
inet 127.0.0.1 netmask 0xff000000
inet6 ::1 prefixlen 128
inet6 fe80::1%lo0 prefixlen 64 scopeid 0x8
gem0: flags=8863<UP,BROADCAST,NOTRAILERS,RUNNING,SIMPLEX,MULTICAST> mtu 1500
lladdr 00:03:ba:0f:34:09
groups: egress
media: Ethernet autoselect (100baseTX full-duplex)
status: active
inet xx.xxx.xxx.x netmask 0xffffff00 broadcast 96.231.191.255
[public ip masked]
inet6 fe80::203:baff:fe0f:3409%gem0 prefixlen 64 scopeid 0x1
gem1: flags=8822<BROADCAST,NOTRAILERS,SIMPLEX,MULTICAST> mtu 1500
lladdr 00:03:ba:0f:34:0a
media: Ethernet autoselect (100baseTX full-duplex)
status: active
hme0:
flags=8b63<UP,BROADCAST,NOTRAILERS,RUNNING,PROMISC,ALLMULTI,SIMPLEX,MULTICAST>
mtu 1500
lladdr 08:00:20:f2:e3:d4
trunk: trunkdev trunk0
media: Ethernet autoselect (100baseTX full-duplex)
status: active
inet6 fe80::a00:20ff:fef2:e3d4%hme0 prefixlen 64 scopeid 0x3
hme1:
flags=8b63<UP,BROADCAST,NOTRAILERS,RUNNING,PROMISC,ALLMULTI,SIMPLEX,MULTICAST>
mtu 1500
lladdr 08:00:20:f2:e3:d4
trunk: trunkdev trunk0
media: Ethernet autoselect (100baseTX full-duplex)
status: active
inet6 fe80::a00:20ff:fef2:e3d5%hme1 prefixlen 64 scopeid 0x4
hme2:
flags=8b63<UP,BROADCAST,NOTRAILERS,RUNNING,PROMISC,ALLMULTI,SIMPLEX,MULTICAST>
mtu 1500
lladdr 08:00:20:f2:e3:d4
trunk: trunkdev trunk0
media: Ethernet autoselect (100baseTX full-duplex)
status: active
inet6 fe80::a00:20ff:fef2:e3d6%hme2 prefixlen 64 scopeid 0x5
hme3:
flags=8b63<UP,BROADCAST,NOTRAILERS,RUNNING,PROMISC,ALLMULTI,SIMPLEX,MULTICAST>
mtu 150...
On a lark I just executed `ifconfig trunk0 up` and now my trunk is
working! And, to make it come up automatically, I just added the single
line "up" to hostname.trunk0...BTW, the "trunk" interface is not documented in hostname.if(5)
Thanks anyways,
Kent
well, neither is, say, bge(4). there's no need to specifically mention
trunk - you can configure it like any other interface. anything specific
to trunk is covered in trunk(4) and ifconfig(8).jmc
| Natalie Protasevich | [BUG] New Kernel Bugs |
| Greg KH | [GIT PATCH] driver core patches against 2.6.24 |
| Bart Van Assche | Integration of SCST in the mainstream Linux kernel |
| Andi Kleen | [PATCH x86] [0/16] Various i386/x86-64 changes |
git: | |
| Gerrit Renker | [PATCH 27/37] dccp: Integration of dynamic feature activation - part 2 (server side) |
| Jarek Poplawski | [PATCH] pkt_sched: Destroy gen estimators under rtnl_lock(). |
| Linus Torvalds | Re: [GIT]: Networking |
| Jeff Kirsher | [net-next PATCH 1/7] e1000e: enable CRC stripping by default |
| Jukka Andberg | ata/wdc vs gcc3 on amiga |
| YAMAMOTO Takashi | Re: wd.c patch to reduce kernel stack usage |
| Jason Thorpe | Re: ksyms patches. |
| rick | NFS transport |
