what is the difference between Ether channel and LAG in SG300 switch

[night-fury]

Hi,

I tried to look around but could not find a proper explanation on what is the difference between an ether channel and LAG in small business series switch SG300.

we have a few ESXi servers and want to increase the copy speed when one VM is moved from one ESXi host to another. NIC teaming is done on the ESXi server having 2 NICs. When i configured the corresponding ports on the SG300 switch via CLI, the host became unreachable.

interface gigabitethernet21
 channel-group 1 mode auto
!
interface gigabitethernet22
 channel-group 1 mode auto

interface Port-channel1
 description esxi
 switchport trunk allowed vlan add 30-33

If i remove one port from the ether channel say gi21, the server becomes reachable.

Then I checked the web console and found an option to configure LAG. did it through the web interface for the same ports, and the server was reachable. only difference that i saw in the configuration was a command (in bold below)

interface gigabitethernet21
 channel-group 1 mode auto
!
interface gigabitethernet22
 channel-group 1 mode auto

interface Port-channel1
 description esxi

flowcontrol auto
 switchport trunk allowed vlan add 30-33

I am a little confused as to are these same or differ in some way.

Is there any other way to increase throughput in such cases?

[Milos Megis]

EtherChannel and LAG should be same.

However you use mode auto, which enables PAgP protocol which is Cisco proprietary and probably it is not supported on ESX side.
Try use option ON, or you can use active which enables LACP protocol which should be supported on ESX side

From command help:
active - Enable LACP unconditionally
auto - Enable PAgP only if a PAgP device is detected
desirable - Enable PAgP unconditionally
on - Enable Etherchannel only
passive - Enable LACP only if a LACP device is detected

[night-fury]

Thank for the revert Milos.

I already tried the mode ON. In small business series switch, there are only two modes, auto and on.

When configured 2 ports with auto, the LAG shows one port as active member and one as standby. Thus they do not help with the throughput but may provide redundancy.

If I configure both ports with ON mode, both ports are active members but LACP gets disabled.

[Julio E. Moisa]

Hi

Link aggregation (LAG) is simply a means to bundle multiple ethernet connections between a pair of switches to share the load.

EtherChannel is Cisco's pre-standards model for doing so, which was originally inherited from the Kalpana acquisition.

IEEE 802.3ad describes the standard way to aggregate links, including the Link Aggregation Control Protocol (LACP).

The resulting link looks like a single interface to Layer 3.

You would use HSRP for protection between two (or more) physical routers so that Layer 3 devices could recover seamlessly from a loss of a default gateway. That protocol serves a totally different purpose than link aggregation; it supports the upstream consumers of the aggregated links.

There are mechanisms to execute link aggregation across multiple chassis, but the techniques are all proprietary. For Cisco, this technique is called Multichassis LACP

Link aggregation (LAG) is a partial implementation of the 802.3ad port aggregation standard. It bundles all of the controller’s distribution system ports into a single 802.3ad port channel, thereby reducing the number of IP addresses needed to configure the ports on your controller. When LAG is enabled, the system dynamically manages port redundancy and load balances access points transparently to the user.

Please rate the comment if it is useful

:-)