aggr not load balancing

[Steven Surdock]

I switched from trunk to aggr on a "OpenBSD 6.8 GENERIC.MP#5 amd64" and it isn't load balancing across the two configured links. The remote side is a Cisco ASR9k with the same configuration. Is that expected?


$ cat /etc/hostname.aggr0
trunkport bge0 trunkport bge1 description "BE2 to ASR9k"
inet 192.168.200.12 255.255.255.240
up

$ifconfig
bge0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
lladdr fe:e1:ba:d0:54:6d
index 1 priority 0 llprio 3
trunk: trunkdev aggr0
media: Ethernet autoselect (1000baseT full-duplex)
status: active
bge1: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
lladdr fe:e1:ba:d0:54:6d
index 2 priority 0 llprio 3
trunk: trunkdev aggr0
media: Ethernet autoselect (1000baseT full-duplex)
status: active
aggr0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
lladdr fe:e1:ba:d0:54:6d
description: BE2 to ASR9k
index 7 priority 0 llprio 7
trunk: trunkproto lacp
trunk id: [(8000,fe:e1:ba:d0:54:6d,0007,0000,0000),
(8000,70:e4:22:67:2a:1e,000C,0000,0000)]
bge0 lacp actor system pri 0x8000 mac fe:e1:ba:d0:54:6d, key 0x7, port pri 0x8000 number 0x1
bge0 lacp actor state activity,aggregation,sync,collecting,distributing
bge0 lacp partner system pri 0x8000 mac 70:e4:22:67:2a:1e, key 0xc, port pri 0x8000 number 0x2
bge0 lacp partner state activity,aggregation,sync,collecting,distributing
bge0 port active,collecting,distributing
bge1 lacp actor system pri 0x8000 mac fe:e1:ba:d0:54:6d, key 0x7, port pri 0x8000 number 0x2
bge1 lacp actor state activity,aggregation,sync,collecting,distributing
bge1 lacp partner system pri 0x8000 mac 70:e4:22:67:2a:1e, key 0xc, port pri 0x8000 number 0x1
bge1 lacp partner state activity,aggregation,sync,collecting,distributing
bge1 port active,collecting,distributing
groups: aggr egress
media: Ethernet autoselect
status: active
inet 192.168.200.12 netmask 0xfffffff0 broadcast 192.168.200.15
inet6 fe80::fce1:baff:fed0:546d%aggr0 prefixlen 64 scopeid 0x7


$ netstat -in
Name Mtu Network Address Ipkts Ifail Opkts Ofail Colls
bge0 1500 <Link> fe:e1:ba:d0:54:6d 32188 0 368 0 0
bge1 1500 <Link> fe:e1:ba:d0:54:6d 8266100 0 7852958 0 0
aggr0 1500 <Link> fe:e1:ba:d0:54:6d 8297364 0 7852590 0 0
aggr0 1500 192.168.200. 192.168.200.12 8297364 0 7852590 0 0
aggr0 1500 fe80::%aggr fe80::fce1:baff:f 8297364 0 7852590 0 0

[Brian R. Landy]

Post by Steven Surdock
I switched from trunk to aggr on a "OpenBSD 6.8 GENERIC.MP#5 amd64" and it isn't load balancing across the two configured links. The remote side is a Cisco ASR9k with the same configuration. Is that expected?

Hi, try 6.9. There is a bugfix to pf that I found also corrects load balancing with aggr.


Brian

[Steven Surdock]

Post by Brian R. Landy

Post by Steven Surdock
I switched from trunk to aggr on a "OpenBSD 6.8 GENERIC.MP#5 amd64" and it isn't load balancing across the two configured links. The remote side is a Cisco ASR9k with the same configuration. Is that expected?

Hi, try 6.9. There is a bugfix to pf that I found also corrects load balancing with aggr.
Brian

Upgrading to 6.9 did resolve the outbound load balancing issue (OBSD --> Switch), but then broke inbound (switch --> OBSD) load balancing. I suspect this is just a load balancing algorithm setting in the ASR9k as the LAG is build on a l2 interface and not a routed interface.