Working Link Aggregation Config RT-AC66U

[mrengles]

Apologies. I guess I was a little confused, because earlier you said..

then

So I assumed you had not configured the switch properly.

@saintdev

I wish it was that simple, it's not the switch configuration. It works with other hardware, just can't get the Rt-ac66u aggregation IDs to matchup and work right. I PM-ed BS the main ddwrt dev who was not very helpful on the subject.

I'm still working on 802.3ad on my end with ddwrt and the rt-ac66u.

Mrengles

PS. I've been looking through all of the ddwrt source code and config files (when time permits) hoping something jumps out at me. Some builds of ddwrt have bonding supported right in the webgui. I'm thinking/hoping that remnants of a bonding configuration file is leftover causing the issue, but I cannot be 100% certain. I'll definitely keep everyone posted if I make any headway with LACP on the RT-AC66U with ddwrt.

 

[DjRavix]

I Got A AC68U That I am trying to get working with Link Aggregation however every time I run the script to bound the ports My LAN Ports all seem to get Bricked
This even happens when I don't have anything connected to the LAN Ports I wound to bound

Can you guys help me out with getting it to work ???

 

[KAD]

first I recommend you post what you want to bond the router to

what switch model or nas model etc

then you should post the output of LinkAgg -s

 

[DjRavix]

first I recommend you post what you want to bond the router to

what switch model or nas model etc

then you should post the output of LinkAgg -s

Okay Let me first give you all the info
It's A Synology DS414 running DSM 4.3-3810 Update 4
RT-AC68U Firmware: 3.0.0.4.374.38_2 (Merlin build)

So I used the run CMD page
/tmp/mnt/sda/LinkAgg 1 2

and than ....
/tmp/mnt/sda/LinkAgg -s

and now it says ....
"Bond Created Successfully"

HM ... Just noticed that it tells me It should work ...

however I am not able to connect to My DS414

And also notice that all other devices that are connected to ports 3 & 4 Lose there connection with the network

There is a Windows PC hooked to LAN Port 3 witch I have manually given an IP
However I changed it to automatic for now and it doesn't get a 192.168.1.x Address assigned

Well thats all for now ...
See you back after I have Rebooted my AC68U

 

[KAD]

well, if you read through this threads history

you'll find that most synology user's are stuck
there's some unknown incompatibility between these routers and synology nas when using LinkAgg

there was one case, where somebody claimed they got it working with synology, but then several weeks/months later the same person posted and their comments make me think it's not actually working even in that case

so sorry, but synology, have no clue what's wrong

 

[DjRavix]

well, if you read through this threads history

you'll find that most synology user's are stuck
there's some unknown incompatibility between these routers and synology nas when using LinkAgg

there was one case, where somebody claimed they got it working with synology, but then several weeks/months later the same person posted and their comments make me think it's not actually working even in that case

so sorry, but synology, have no clue what's wrong

I got that ...
But The weird thing is that the LAN ports also seem to brick When the NAS is not connected to it

I also looked into the synology DSM and found that it still gives a "Failed to establish IEEE 802.3ad Connection" Error
I Actually did this by setting up a hotspot on my DS414 (using A Cisco Wifi Dongle) and connecting a Netbook to it to see what DMS would say

Now I have Unpluged all the cables From the LAN ports ...
Turned Off my AC68U
just waited some time ....

Run the script from my wireless System

"LinkAgg -S" Shows
--- Bond Errors ---







--- Bond Status ---

Bond Status: bond0 UP
Bridge to WAN Status: Member of br0=bond0
Slave 1 Status: vlan=vlan3 Link=UP Port=1
Slave 2 Status: vlan=vlan4 Link=UP Port=2

Connected Back My Wired PC to check if the LAN Ports still work ...
and the ports were bricked again ...

For some reason I do not think it's the NAS thats the problem
Because it's not connected ...

 

[KAD]

I still don't know what's wrong

according to the output of LinkAgg -s
the router is set up and functioning correctly

if there was something else to check, I would add that to status checker function of that script

RMerlin mentioned all the support request for LinkAgg
I'd estimate that 90% of problems reported here are compatibility with synology nas

synology does have an officially supported device list
however there are only 4 specific switch models listed
3com 4200G
Allied Telesis AT-x510-28GTX
HP Procurve Switch 2824
Volktek NSH-2926

outside of those 4 models synology doesn't guarantee compatibility for Link Aggregation

earlier in this thread, we also looked at quite a few other advanced items
like transmit has policy

and all has failed to work with synology nas

at this point, I hate to turn somebody away
but I've come to just accept that it does not work with synology nas
and there's nothing I can do about it

now if anybody does have a valid idea, of what the incompatibility is
I'm certainly willing to look into it
but for now, I'm out of idea's

 

[RMerlin]

Dunno if Synology is as open as QNap (which has a telnet interface), but if it does, might be worth logging into it, and looking exactly how they are configuring their LAG, which kernel module / version they use, etc...

 

[saulos]

Hi Merlin,
yes synology have the telnet SSH enabled ( you can set it from web interface ),
I have a 713 with the latest FW.
I can check those for you but don't know where to check

 

[KAD]

Hi Merlin,
yes synology have the telnet SSH enabled ( you can set it from web interface ),
I have a 713 with the latest FW.
I can check those for you but don't know where to check

modprobe -v bonding.ko
or if modinfo exist, you could use that to obtain version info

also a couple of pages back, somebody already posted some source code synology is using for bonding, including their LAG set up script

what I haven't seen anybody do yet
is set up LAG with synology (even if in a failed state)
run cat /proc/net/bonding/bond0 on the synology side to see what synology's setup script actually did once LAG was enabled

 

[DjRavix]

modprobe -v bonding.ko
or if modinfo exist, you could use that to obtain version info

also a couple of pages back, somebody already posted some source code synology is using for bonding, including their LAG set up script

what I haven't seen anybody do yet
is set up LAG with synology (even if in a failed state)
run cat /proc/net/bonding/bond0 on the synology side to see what synology's setup script actually did once LAG was enabled

like this ...
DiskStation414> cat /proc/net/bonding/bond0
Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 100
Down Delay (ms): 0

802.3ad info
LACP rate: fast
Min links: 0
Aggregator selection policy (ad_select): stable
Active Aggregator Info:
Aggregator ID: 1
Number of ports: 1
Actor Key: 17
Partner Key: 1
Partner Mac Address: 00:00:00:00:00:00

Slave Interface: eth0
Speed: 1000
Duplex: full
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 00:11:32:22:a1:9d
Aggregator ID: 1
Slave queue ID: 0

Slave Interface: eth1
Speed: 1000
Duplex: full
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 0
Permanent HW addr: 00:11:32:22:a1:9e
Aggregator ID: 2
Slave queue ID: 0

 

[KAD]

yep that's exactly what I mean

synology is using a much newer version of the bonding driver

besides that the error I see is aggregator ID, it should match on both slaves but it doesn't

this is the same problem the guy trying to do this with DD-WRT was having
but really doesn't tell us what is wrong, just that it's not working correctly
and something is causing the bonding driver to think there should be 2 different bonds

this
Aggregator selection policy (ad_select): stable
is also a field not showing on my ac66u

so it's something that could be looked at
I wonder if it's possible on syno side to set value 1 bandwidth

the 3 options for that value are

Stable or 0 :
This is the original behaviour and is the default setting.
The active aggregator is chosen by largest aggregate bandwidth.
Re-selection of the active aggregator occurs only when all slaves of the active aggregator are down or the active aggregator has no slaves.
Bandwidth or 1 :
The active aggregator is chosen by largest aggregate bandwidth. Reselection occurs if:

A slave is added to or removed from the bond.
Any slave's link state changes.
Any slave's 802.3ad association state changes.
The bond's administrative state changes to up.

Count or 2 :
The active aggregator is chosen by the largest number of ports (slaves). Reselection occurs as described under the Bandwidth setting above.

under the stable option, aggregator id's are only changed if both links go down
by setting value 1, you could add/remove slaves, etc to try and get the aggregator ID's to change and match

not this option was added with driver version 3.4.0 and so does not exist in current firmware for these routers

 

[DjRavix]

Tomorrow I will get a Netgear GS108Tv2
Witch from what I have been reading seems to work pretty fine with A Synology NAS

Than Lets see if it will work with the AC68u

 

[dragnet]

LACP ac66u to Thecus N5550

Hello all,

Don't mean to hijack this thread but I thought I would bring Thecus into the mix.

Thanks to KAD I've configured my ac66u to successfully bond ports 3 and 4. Likewise my Thecus N5550 supports 802.3ad and I have bonded the two gigabit NICs successfully there. Unfortunately LAN2 on the Thecus (vlan4 or port 4 on the ac66u) shows no traffic. I have no Load Balancing and no Failover working. Here is the output from cat /proc/net/bonding/bond0 from each device. The only thing I see is that Aggregator ID does not match between devices. Does the ID need to match on both devices or is this ID local to the device in question?

As well driver versions are quite different.

ac66u cat /proc/net/bonding/bond0

Ethernet Channel Bonding Driver: v3.1.3 (June 13, 2007)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 0
Down Delay (ms): 0

802.3ad info
LACP rate: slow
Active Aggregator Info:
        Aggregator ID: 2
        Number of ports: 2
        Actor Key: 5
        Partner Key: 17
        Partner Mac Address: 00:14:fd:17:6e:7e

Slave Interface: vlan3
MII Status: up
Link Failure Count: 0
Permanent HW addr: 60:a4:4c:68:c3:80
Aggregator ID: 2

Slave Interface: vlan4
MII Status: up
Link Failure Count: 0
Permanent HW addr: 60:a4:4c:68:c3:80
Aggregator ID: 2

cat /proc/net/bonding/bond0 from Thecus N5550

Ethernet Channel Bonding Driver: v3.7.1 (April 27, 2011)

Bonding Mode: IEEE 802.3ad Dynamic link aggregation
Transmit Hash Policy: layer2 (0)
MII Status: up
MII Polling Interval (ms): 100
Up Delay (ms): 300
Down Delay (ms): 0

802.3ad info
LACP rate: slow
Aggregator selection policy (ad_select): stable
Active Aggregator Info:
        Aggregator ID: 1
        Number of ports: 2
        Actor Key: 17
        Partner Key: 5
        Partner Mac Address: 60:a4:4c:68:c3:80

Slave Interface: eth0
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 1
Permanent HW addr: 00:14:fd:17:6e:7e
Aggregator ID: 1
Slave queue ID: 0

Slave Interface: eth1
MII Status: up
Speed: 1000 Mbps
Duplex: full
Link Failure Count: 1
Permanent HW addr: 00:14:fd:17:6e:7f
Aggregator ID: 1
Slave queue ID: 0

When transferring data between the NAS and a PC I can see traffic through bond0 via ifconfig on both devices. I just don't see the secondary connection vlan4 taking any traffic, only port 3 (vlan3). If I unplug port 3 from the router I lose all connectivity to the NAS I do not failover to port 4 (vlan4).

Cheers and TIA...

 

[KAD]

I don't think that's hi-jacking at all

aggregator id is device specific

so your output, shows it working correctly on both ends
ac66u and thecus

 

[dragnet]

Thanks KAD.

Yes. It looks to be configured correctly but still no traffic on the secondary interface in the bond.

Short of unplugging vlan3, which kills my NAS connection, is there a way to force vlan4 to be the primary connection in the bond?

It's my understanding that 802.3ad should alternate packets between each NIC. Does 802.3ad by default use round robin scheduling between both NICs or will it only use the secondary interface under certain conditions such as failover?

Cheers...

 

[saintdev]

It's my understanding that 802.3ad should alternate packets between each NIC. Does 802.3ad by default use round robin scheduling between both NICs or will it only use the secondary interface under certain conditions such as failover?

This is where the transmit hash policy comes into play. The default (later 2) uses a hash of the destination MAC to determine which slave to send on. So transferring data between the bonded device and one other device will only ever result in one of the slaves being used. There is no round robin mode for 802.3ad, because of the requirements for sequential packets.

 

[DjRavix]

So I got my GS108Tv2

It was A breeze to set it up with My DS414
However ... the AC68U ...
Well the ports just Bricked again ...
So I am concerned that the Synoligy NAS was not the problem

So Did Anybody get the AC68U to work with link aggregation ???

 

[KAD]

@dragnet

some other items that I notice in your output

LACP rate : slow

your link is working because both sides match this
but
the LinkAgg script only sets

echo fast > /sys/class/net/${1}/bonding/lacp_rate

fast

so that's another mystery

 

[Tweexter]

Hello all! I created an account here just to engage this thread.
In the near future I'm going to purchase the AC68U and by the sounds of it put the Merlin firmware on there first thing. I'm waiting on the new Synology 1514+ to come out this year, hopefully sooner than later, and we'll see if that newer hardware plays nice with the 802.3ad idea.

I'm also considering getting a dual NIC PCI-E card for my server box at home so I can team that, as I do a LOT of file transfers to/from that thing. I read all 20 pages of this thread and saw someone mention multiple bonds (say 1+2 and 3+4) but didn't see a confirmation if that functionality was added with this script. All other devices in the household are wireless, so not worried about using all 4 ports.

Having a robust network infrastructure at home is just fun for me, as an IT geek. At work I get the play with server clusters that have 16+ NICs each, EqualLogic SANs, Cisco catalyst routers... etc etc. Of course vSphere makes link teaming a piece of cake. But, when I get home I want to keep on playing

Cheers!!