Today I change the switch and test again used LACP and trunk group mode .
[b]1、config LACP mode in switch[/b]
the result is same ,four client just 45Mp/s 、65 Mp/s 、54 Mp/s 53Mp/s ,the
total throughout still like two nics.
r...@opensolaris:~# dladm show-aggr -P
LINK POLICY ADDRPOLICY LACPACTIVITY LACPTIMER FLAGS
aggr1 L4 auto off short -----
r...@opensolaris:~# dladm show-aggr -s -i 4
LINK PORT IPACKETS RBYTES OPACKETS OBYTES IPKTDIST OPKTDIST
aggr1 -- 478228 707101154 310367 23287550 -- --
-- e1000g2 181495 265767728 116603 8743344 38.0 37.6
-- e1000g3 160776 237541650 89539 6721486 33.6 28.8
-- e1000g5 2 256 104221 7822208 0.0 33.6
-- e1000g6 135955 203791520 4 512 28.4 0.0
aggr1 -- 539723 796045522 351091 26407694 -- --
-- e1000g2 196996 287809424 127266 9574078 36.5 36.2
-- e1000g3 181492 267981462 105621 7944376 33.6 30.1
-- e1000g5 2 256 118200 8888728 0.0 33.7
-- e1000g6 161233 240254380 4 512 29.9 0.0
[b]2、Use trunk group in switch[/b]
the result is like lacp mode.
r...@opensolaris:~# dladm show-aggr -P
LINK POLICY ADDRPOLICY LACPACTIVITY LACPTIMER FLAGS
aggr1 L4 auto off short -----
r...@opensolaris:~# dladm show-aggr -s -i 4
LINK PORT IPACKETS RBYTES OPACKETS OBYTES IPKTDIST OPKTDIST
aggr1 -- 452171 668378924 290798 21884474 -- --
-- e1000g2 129003 186725146 81900 6155716 28.5 28.2
-- e1000g3 179157 267807716 0 0 39.6 0.0
-- e1000g5 2 128 93986 7062478 0.0 32.3
-- e1000g6 144009 213845934 114912 8666280 31.8 39.5
aggr1 -- 585917 866832346 376202 28322054 -- --
-- e1000g2 176536 255595534 111973 8416924 30.1 29.8
-- e1000g3 235355 352934638 0 0 40.2 0.0
-- e1000g5 5 416 113316 8516214 0.0 30.1
-- e1000g6 174021 258301758 150913 11388916 29.7 40.1
It same more nics aggregation and get more bad performance .
Does this is a konwn Bug ?
Can networking guys give us some advise? It's really very important for us.
--
This message posted from opensolaris.org
_______________________________________________
networking-discuss mailing list
[email protected]
