vMX

hi team,

I am able to connect two vMX with LAG without using LACP. Once I enable the LACP, the ae interface status change to down.

Anybody can suggest if LACP is supported on vMX or not?

my vMX version 14.1R1.10

Just found something

Problem Report

LACP is not supported with virtIO.

You are not running the proper code for vMX, the code you have is from an internal engineering program.  please download the production code from: https://www.juniper.net/support/downloads/?p=vmx#sw

the proper code version fro vMX is 14.1R5

That defect applies to the production code fro vMX. 

Thanks

Hi,

Thank you very much for your reply.

Any ETA for the fix for this LACP defect?

The reason why LACP does not work in virtio mode is due to the fact that we use linux bridges to steer traffic to vmx. The inbuilt behavior of linux bridge is to consume (and not forward) all link-local multicast packets (like LACP). You can workaround the problem by 2 ways:

1) use SRIOV instead of virtio

2) use OVS instead of linux bridges in case of virtio

Hello,

Did anyone managed to configure vMX with lacp and OVS?

I tried but the bond interface will never come up.

Thanks.

I managed to configure LACP using virtio interfaces and linux bridge.

You have to patch the linux kernel and modify the BR_GROUPFWD_RESTRICTED from net/bridge/br_private.h in order to allow you to set /sys/class/net/<bridge iface>/bridge/group_fwd_mask.

like this.

/usr/src/linux/net/bridge # more patch.br_private.h
--- br_private.h	2018-05-29 17:18:19.408679820 +0900
+++ br_private.h	2018-05-29 17:19:31.177539982 +0900
@@ -35,7 +35,7 @@
 /* Control of forwarding link local multicast */
 #define BR_GROUPFWD_DEFAULT	0
 /* Don't allow forwarding of control protocols like STP, MAC PAUSE and LACP */
-#define BR_GROUPFWD_RESTRICTED	0x0007u
+#define BR_GROUPFWD_RESTRICTED	0x0000u
 /* The Nearest Customer Bridge Group Address, 01-80-C2-00-00-[00,0B,0C,0D,0F] */
 #define BR_GROUPFWD_8021AD	0xB801u

 but this way is out of linux  specification. 

 After modify br_private.h. it need to compile kernel.

After restart system. then give following command.

echo 255 > /sys/class/net/t1_br1/bridge/group_fwd_mask

 you can see this.

cat /sys/class/net/t1_br1/bridge/group_fwd_mask
0xff

Hi All,

Could anybody suggest why lacp enabled interfaces are not coming up in Juniper Virtual lab and Juniper cloud lab vMX device.

FYI, I am using Juniper virtual lab and Juniper cloud lab to build LACP for MC-LAG but here lacp enabled interfaces are not coming up. However, i used "force-up" command to make the lacp enabled interface up but its not communicating each other for which the IRB vlan is passing over the lacp interface. If "force-up" command is not used then interface will not come up.

Please let me know if anybody overcome this issue in juniper virtual/ Cloud lab.

Below is the vMX config:

vMX1:-

jcluser@Core-A> show configuration |display set | match ae0
set interfaces ge-0/0/0 gigether-options 802.3ad ae0
set interfaces ae0 description "ICL-LINK b/w Core A & B"
set interfaces ae0 flexible-vlan-tagging
set interfaces ae0 encapsulation flexible-ethernet-services
set interfaces ae0 aggregated-ether-options minimum-links 1
set interfaces ae0 aggregated-ether-options link-speed 1g
set interfaces ae0 aggregated-ether-options lacp active
set interfaces ae0 aggregated-ether-options lacp periodic fast
set interfaces ae0 aggregated-ether-options lacp force-up
set interfaces ae0 unit 0 family bridge interface-mode trunk
set interfaces ae0 unit 0 family bridge vlan-id-list 1-4094

jcluser@Core-A> show interfaces terse |match ae0
ge-0/0/0.0 up up aenet --> ae0.0
ge-0/0/0.32767 up up aenet --> ae0.32767
ae0 up up
ae0.0 up up bridge
ae0.32767 up up multiservice

jcluser@Core-A> show bridge domain

Routing instance Bridge domain VLAN ID Interfaces
default-switch ICCP 1
ae0.0--->> This interface should be indicated as ae0.0* if the lacp ineterface working properly, Here no * indication even after configured the "force-up" command on ae0.

jcluser@Core-A>

vMX2:-

jcluser@Core-B# show |display set |match ae0
set interfaces ge-0/0/0 gigether-options 802.3ad ae0
set interfaces ae0 description "ICL-LINK b/w Core A & B"
set interfaces ae0 flexible-vlan-tagging
set interfaces ae0 encapsulation flexible-ethernet-services
set interfaces ae0 aggregated-ether-options minimum-links 1
set interfaces ae0 aggregated-ether-options link-speed 1g
set interfaces ae0 aggregated-ether-options lacp active
set interfaces ae0 aggregated-ether-options lacp periodic fast
set interfaces ae0 aggregated-ether-options lacp force-up
set interfaces ae0 unit 0 family bridge interface-mode trunk
set interfaces ae0 unit 0 family bridge vlan-id-list 1-4094

[edit]
jcluser@Core-B# run show interfaces terse |match ae0
ge-0/0/0.0 up up aenet --> ae0.0
ge-0/0/0.32767 up up aenet --> ae0.32767
ae0 up up
ae0.0 up up bridge
ae0.32767 up up multiservice

[edit]
jcluser@Core-B# run show bridge domain

Routing instance Bridge domain VLAN ID Interfaces
default-switch ICCP 1
ae0.0----->Here also ae0 interface is not indicated by start (*).

[edit]
jcluser@Core-B#

Iam_Chethan,

Did you initialize the aggregated-devices?

On eveng for it to work, you need to configure a MAC address for the aex (system-id) and force it up.

The images have the same MAC for the ae0 interface and by ricochet the same system-id for LACP… It doesn't seem to work that way, unless you change the macs / system id and  force the lag on up… at least in the virtual environment…