Message-ID: <562F7940.3020800@cumulusnetworks.com>
Date:	Tue, 27 Oct 2015 14:16:48 +0100
From:	Nikolay Aleksandrov <nikolay@...ulusnetworks.com>
To:	Arjun Pandey <apandepublic@...il.com>
Cc:	netdev@...r.kernel.org
Subject: Re: 802.3 ad bonding mode min-links doesn't work correctly once slave
 port link status is up again

On 10/27/2015 01:55 PM, Arjun Pandey wrote:
> Hi Nikolay
> 
> But based on this output i think they are part of same aggregator.
> cat  /sys/class/net/bond/bonding/slaves
> eth2 eth1
> I am adding slave ports via ifenslave bond eth1 eth2
> 
Hi again,
Please don't top post. You've shown that they're part of 1 bond
interface but that doesn't mean they're part of a single aggregator.
Look below,

> Each individual slave port seems to be getting a different aggr id. I
> confirmed this by adding an additional port
> [root@foo /]# cat /proc/net/bonding/bond
> 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: down
> MII Polling Interval (ms): 100
> Up Delay (ms): 0
> Down Delay (ms): 0
> 
> 802.3ad info
> LACP rate: fast
> Min links: 2
> Aggregator selection policy (ad_select): stable
> Active Aggregator Info:
> Aggregator ID: 1
^^^
Bond active aggregator id is 1

> Number of ports: 1
^^^
_1 port_

> Actor Key: 17
> Partner Key: 1
> Partner Mac Address: 00:00:00:00:00:00
> 
> Slave Interface: eth1
> MII Status: up
> Speed: 1000 Mbps
> Duplex: full
> Link Failure Count: 0
> Permanent HW addr: 08:00:27:0a:cd:2c
> Aggregator ID: 1
^^^
This port is part of aggregator id 1 (this is the port of
the active aggregator)

> Slave queue ID: 0
> 
> Slave Interface: eth2
> MII Status: up
> Speed: 1000 Mbps
> Duplex: full
> Link Failure Count: 0
> Permanent HW addr: 08:00:27:b0:4d:7e
> Aggregator ID: 2
^^^
This port is a part of a different aggregator (id 2)

> Slave queue ID: 0
> 
> Slave Interface: eth3
> MII Status: down
> Speed: Unknown
> Duplex: Unknown
> Link Failure Count: 0
> Permanent HW addr: 08:00:27:e7:dd:6b
> Aggregator ID: 3
> Slave queue ID: 0
> 
> Regards
> Arjun
> 
> On Tue, Oct 27, 2015 at 6:02 PM, Nikolay Aleksandrov
> <nikolay@...ulusnetworks.com> wrote:
>> On 10/27/2015 01:17 PM, Arjun Pandey wrote:
>>> Hi
>>>
>>> I have configured a bond with 2 slave bond ports in 802.3 ad mode.
>>> I have also set min-links=2 and miimon=100
>>>
>>>
>>> root@foo bonding]# cat /proc/net/bonding/bond
>>> 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: down
>>> MII Polling Interval (ms): 100
>>> Up Delay (ms): 0
>>> Down Delay (ms): 0
>>>
>>> 802.3ad info
>>> LACP rate: fast
>>> Min links: 2
>>> Aggregator selection policy (ad_select): stable
>>> Active Aggregator Info:
>>> Aggregator ID: 2
>>> Number of ports: 1
>>> Actor Key: 17
>>> Partner Key: 1
>>> Partner Mac Address: 00:00:00:00:00:00
>>>
>>> Slave Interface: eth1
>>> MII Status: up
>>> Speed: 1000 Mbps
>>> Duplex: full
>>> Link Failure Count: 7
>>> Permanent HW addr: 08:00:27:0a:cd:2c
>>> Aggregator ID: 1
>>> Slave queue ID: 0
>>>
>>> Slave Interface: eth2
>>> MII Status: up
>>> Speed: 1000 Mbps
>>> Duplex: full
>>> Link Failure Count: 4
>>> Permanent HW addr: 08:00:27:b0:4d:7e
>>> Aggregator ID: 2
>>> Slave queue ID: 0
>>>
>>> I tried the following steps :
>>> 1. Bring up bond with slaves as eth1 and eth2
>>> 2. Bring down eth1 down via ip link set down eth1
>>> 3. Check bond link status which now shows up as down
>>> 4. Restore eth1 link status up.
>>> 5. Bond link status is still set as down.
>>>
>>> I can't get bond link status up unless i make min-links 0/1 and bring
>>> bond status manually up via ip link set bond down and ip link set bond
>>> up
>>> [root@foo /]# ip link show
>>> 1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN
>>>     link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
>>> 2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast
>>> state UP qlen 1000
>>>     link/ether 08:00:27:f8:38:ad brd ff:ff:ff:ff:ff:ff
>>> 3: eth1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc
>>> pfifo_fast master bond state UP qlen 1000
>>>     link/ether 08:00:27:0a:cd:2c brd ff:ff:ff:ff:ff:ff
>>> 4: eth2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc
>>> pfifo_fast master bond state UP qlen 1000
>>>     link/ether 08:00:27:0a:cd:2c brd ff:ff:ff:ff:ff:ff
>>> 23: bond: <NO-CARRIER,BROADCAST,MULTICAST,MASTER,UP> mtu 1500 qdisc
>>> noqueue state DOWN
>>>     link/ether 08:00:27:0a:cd:2c brd ff:ff:ff:ff:ff:ff
>>>
>>> Am i missing something ?
>>>
>>>
>>> This is on Centos 6.5 and kernel 3.10.27.1
>>>
>>> Regards
>>> Arjun
>>
>> Hi Arjun,
>> I think your slaves are in different aggregators (judging by their agg ids) and min_links
>> checks if there're min_links number of active ports in the active aggregator (in your case
>> that's agg id 2 which has 1 slave in it).
>>
>> Cheers,
>>  Nik
>>
>>

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Powered by blists - more mailing lists