thr3ads.net - CentOS - [CentOS] Pb with bounding [Feb 2019]

If this information is useful, please help other people find it:
Share via:

Pierre Malard

2019-Feb-06 17:47 UTC

[CentOS] Pb with bounding

Hi,

We have a Dell server with 4 Ethernet interface. I would to aggregate them in a
bond. Everything work but the default gateway doesn?t work on the ? bond0 ?
interface and I have no links.

My configuration:
- CentOS 7:
:/etc/sysconfig/network-scripts# uname -a
Linux nas-mtd2 3.10.0-957.5.1.el7.x86_64 #1 SMP Fri Feb 1 14:54:57 UTC 2019
x86_64 x86_64 x86_64 GNU/Linux

- NetworkManager disabled:
:/etc/sysconfig/network-scripts# systemctl list-unit-files | grep NetworkManager
NetworkManager-dispatcher.service             disabled
NetworkManager-wait-online.service            disabled
NetworkManager.service                        disabled

- ifcfg-em1 to em4 (they are identical):
:/etc/sysconfig/network-scripts# cat ifcfg-em2
NAME=bond0-slave[1 to 4]
DEVICE=em[1 to 4]
TYPE=Ethernet
BOOTPROTO=none
ONBOOT=yes
MASTER=bond0
SLAVE=yes
NM_CONTROLLED=no

- the bond0 declaration:
:/etc/sysconfig/network-scripts# cat ifcfg-bond0
DEVICE=bond0
TYPE=Ethernet
BONDING_MASTER=yes
BOOTPROTO=none
ONBOOT=yes
NM_CONTROLLED=no

# IP
IPADDR=172.16.1.128
PREFIX=16
GATEWAY=172.16.0.254

# Bonding
BONDING_OPTS="mode=802.3ad ad_select=stable use_carrier=1 miimon=100
arp_ip_target=172.16.1.128 updelay=100 lacp_rate=fast"
USERCTL=no

After a ? systemctl restart network ?:
:/etc/sysconfig/network-scripts# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group
default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
2: em1: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:0e brd ff:ff:ff:ff:ff:ff
3: em2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
4: em3: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
5: em4: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
6: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue
state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
    inet 172.16.1.128/16 brd 172.16.255.255 scope global bond0
       valid_lft forever preferred_lft forever

And the route command:
:/etc/sysconfig/network-scripts# ip r
default via 172.16.0.254 dev bond0
169.254.0.0/16 dev bond0 scope link metric 1006
169.254.0.0/16 dev bond0.10 scope link metric 1008
172.16.0.0/16 dev bond0 proto kernel scope link src 172.16.1.128

but their is no network connection on ? bond0 ?!

The only way I found to have a network active connection is keeping my first
interface outside the bond and having the default route on this one!
In this case I have a specific IP to it and a bond one on bound.

As I must have also a VLAN connection on the bound, I have create a ?
ifcfg-bond0.10 ? file:
:/etc/sysconfig/network-scripts# cat ifcfg-bond0.10
NAME=bond0.10
DEVICE=bond0.10
#TYPE="Ethernet"
VLAN=yes
NM_CONTROLLED=no
ONPARENT=yes
DEFROUTE="no"
# IP
IPADDR=XXX.XXX.XXX.142
NETMASK=255.255.255.0
NETWORK=XXX.XXX.XXX.0
BROADCAST=XXX.XXX.XXX.255
PREFIX=24
GATEWAY=1XXX.XXX.XXX.126

With that and the first Ethernet interface on a specific address I have:
:/etc/sysconfig/network-scripts# ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group
default qlen 1000
    link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
    inet 127.0.0.1/8 scope host lo
       valid_lft forever preferred_lft forever
2: em1: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc mq state UP group
default qlen 1000
    link/ether d0:94:66:65:32:0e brd ff:ff:ff:ff:ff:ff
    inet 172.16.1.86/16 brd 172.16.255.255 scope global em1
       valid_lft forever preferred_lft forever
3: em2: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
4: em3: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
5: em4: <BROADCAST,MULTICAST,SLAVE,UP,LOWER_UP> mtu 1500 qdisc mq master
bond0 state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
6: bond0: <BROADCAST,MULTICAST,MASTER,UP,LOWER_UP> mtu 1500 qdisc noqueue
state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
    inet 172.16.1.128/16 brd 172.16.255.255 scope global bond0
       valid_lft forever preferred_lft forever
8: bond0.10 at bond0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc
noqueue state UP group default qlen 1000
    link/ether d0:94:66:65:32:10 brd ff:ff:ff:ff:ff:ff
    inet XXX.XXX.XXX.142/24 brd XXX.XXX.XXX.255 scope global bond0.10
       valid_lft forever preferred_lft forever
:/etc/sysconfig/network-scripts# ip r
default via 172.16.0.254 dev em1
169.254.0.0/16 dev em1 scope link metric 1002
169.254.0.0/16 dev bond0 scope link metric 1006
169.254.0.0/16 dev bond0.10 scope link metric 1008
172.16.0.0/16 dev em1 proto kernel scope link src 172.16.1.86
172.16.0.0/16 dev bond0 proto kernel scope link src 172.16.1.128
XXX.XXX.XXX.0/24 dev bond0.10 proto kernel scope link src XXX.XXX.XXX.142

And everything is working.

But? I am not satisfy. Is it normal?
Is it a way to have all the Ethernet link on a bond?

Your help is needed :-)

Best wishes


--
Pierre Malard

  ? Si l'on veut croire en l'humanit?,
     il faut voir et comprendre l'inhumanit? ?
   |\      _,,,---,,_
   /,`.-'`'    -.  ;-;;,_
  |,4-  ) )-,_. ,\ (  `'-'
 '---''(_/--'  `-'\_)   ?r

perl -e '$_=q#: 3|\ 5_,3-3,2_:
3/,`.'"'"'`'"'"' 5-.  ;-;;,_:  |,A- 
) )-,_. ,\ (  `'"'"'-'"'"':
'"'"'-3'"'"'2(_/--'"'"'
`-'"'"'\_):
24?r::#;y#:#\n#;s#(\D)(\d+)#$1x$2#ge;print'
- --> Ce message n?engage que son auteur <--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL:
<http://lists.centos.org/pipermail/centos/attachments/20190206/eec5cf4d/attachment.sig>

Gordon Messmer

2019-Feb-07 00:36 UTC

head link

[CentOS] Pb with bounding

On 2/6/19 9:47 AM, Pierre Malard wrote:> We have a Dell server with 4 Ethernet interface. I would to aggregate them
in a bond.
> ...
> :/etc/sysconfig/network-scripts# cat ifcfg-bond0
> ...
> BONDING_OPTS="mode=802.3ad

What type of managed switch are you using for those interfaces? Can you 
also include the configuration that you gave the switch? Do the switch 
logs indicate any errors with the bonded ports?

Pierre Malard

2019-Feb-07 15:19 UTC

head link

[CentOS] Pb with bounding

Sorry, my answer take a while? but I would test before write.

Effectively, the problem became from switch configuration.
Our switch are a stack of Avaya ERS4870 and I have difficulties to understand
the interface. I had configure all ports concerned by my configuration like that
for all ports:
In the ? VLAN ? Configuration:
  - VLAN tagging: ? tagAll(trunk) ?
  - ? PrivateVlanPortType ?: ? trunk ?
  - ? DefaultVlanID ? : to a LAN not distribued. I don?t understand why they ask
for that even I choose ? tagAll ? (!!!)
For the ? LACP ? one:
  - activate ? AdmiEnabled ?
  - ? ActorAdminState ? structure: activation of ? lacpActive ?, ? aggregation ?
and ? shortTimeout ?
  - ? actorAdminKey ?: chose one

In my case the problem come that I have gave one of my real VLAN as ?
DefaultVlanID ?. With this configuration no paquet can pass. Now I have choose
the VLAN 1 in this case and everything is Ok.

Best regards and thanks?
> Le 7 f?vr. 2019 ? 01:36, Gordon Messmer <gordon.messmer at gmail.com>
a ?crit :
> 
> On 2/6/19 9:47 AM, Pierre Malard wrote:
>> We have a Dell server with 4 Ethernet interface. I would to aggregate
them in a bond.
>> ...
>> :/etc/sysconfig/network-scripts# cat ifcfg-bond0
>> ...
>> BONDING_OPTS="mode=802.3ad
> 
> 
> What type of managed switch are you using for those interfaces? Can you
also include the configuration that you gave the switch? Do the switch logs
indicate any errors with the bonded ports?
> 
> 
> _______________________________________________
> CentOS mailing list
> CentOS at centos.org
> https://lists.centos.org/mailman/listinfo/centos
--
Pierre Malard

   ? Tant que les lions n?auront pas leurs propres historiens, les histoires
   de chasse tourneront toujours ? la gloire du chasseur ?
                                                  Proverbe africain
   |\      _,,,---,,_
   /,`.-'`'    -.  ;-;;,_
  |,4-  ) )-,_. ,\ (  `'-'
 '---''(_/--'  `-'\_)  ?r

perl -e '$_=q#: 3|\ 5_,3-3,2_:
3/,`.'"'"'`'"'"' 5-.  ;-;;,_:  |,A- 
) )-,_. ,\ (  `'"'"'-'"'"':
'"'"'-3'"'"'2(_/--'"'"'
`-'"'"'\_):
24?r::#;y#:#\n#;s#(\D)(\d+)#$1x$2#ge;print'
- --> Ce message n?engage que son auteur <--

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: Message signed with OpenPGP
URL:
<http://lists.centos.org/pipermail/centos/attachments/20190207/7cfc28f5/attachment-0002.sig>

Reasonably Related Threads

Search for more reasonably related threads

CentOS - Feb 2019 - Pb with bounding

[CentOS] Pb with bounding

[CentOS] Pb with bounding

[CentOS] Pb with bounding

Reasonably Related Threads