[SOLVED] vmbr0: port 1(enp2s0) entered blocking state

[RobFantini]

Hello

on pve5 systems, not pve4 I see this :

# dmesg|grep   block|grep vmbr
[    4.407075] vmbr0: port 1(enp2s0) entered blocking state
[    7.616099] vmbr0: port 1(enp2s0) entered blocking state
[325383.846647] vmbr0: port 1(enp2s0) entered blocking state

# ssh sys20  "dmesg|grep   block|grep vmbr"
[    4.205986] vmbr0: port 1(eno1) entered blocking state
[    9.845410] vmbr0: port 1(eno1) entered blocking state
[325167.997139] vmbr0: port 1(eno1) entered blocking state

# ssh sys6  "dmesg|grep   block|grep vmbr"
[    4.193060] vmbr0: port 1(enp2s0) entered blocking state
[    7.044614] vmbr0: port 1(enp2s0) entered blocking state
[326565.756673] vmbr0: port 1(enp2s0) entered blocking state

is ' blocking state ' something I need to deal with or is it normal?

 

[Alwin]

These are informational messages that your bridge is in blocking state (not forwarding packets), as long as the bridge goes to forwarding mode you should have no problem.
http://www.microhowto.info/troubleshooting/troubleshooting_ethernet_bridging_on_linux.html#idp36608

 

[RobFantini]

thanks for the response.

the state is now forwarding per ` brctl showstp vmbr0` .

Is it unusual to have 3 blocks at each node? Or during startup do pve5 nodes usually have multiple blocks?

Just checking to see if we have an something to adjust on the layer 3 network switch.

sys13  ~ # brctl showstp vmbr0
vmbr0
 bridge id              8000.002590aa5e96
 designated root        8000.002590aa5e96
 root port                 0                    path cost                  0
 max age                  20.00                 bridge max age            20.00
 hello time                2.00                 bridge hello time          2.00
 forward delay             0.00                 bridge forward delay       0.00
 ageing time             300.00
 hello timer               0.00                 tcn timer                  0.00
 topology change timer     0.00                 gc timer                  50.18
 flags


enp2s0 (1)
 port id                8001                    state                forwarding
 designated root        8000.002590aa5e96       path cost                  4
 designated bridge      8000.002590aa5e96       message age timer          0.00
 designated port        8001                    forward delay timer        0.00
 designated cost           0                    hold timer                 0.00
 flags

 

[Alwin]

Any interface that is dis-/connecting to the bridge will generate these entries in syslog. Depending on the bridge_stp set or not it will check for loops or not.

 

[RobFantini]

last question - you mentioned stp. the pve systems are connected to a switch that has rstp set. or mstp .

in that case should 'bridge_stp' be set to on ?

 

[lynn_yudi]

how to solved?
must be set 'bridge_stp' to on ?

 

[Alwin]

STP on the bridge will behave the same as on a switch. If you build multiple paths connecting to your bridge, then it might be a good option to activate it. Though the option introduces a delay while connecting a interface to the bridge, as on a switch, it has to learn any possible new paths. It is off by default, in our setup to reduce that delay, as usually it connects VMs to a network through a interface directly connected to a switch.

@lynn_yudi, only the user who opened the thread can modify the subject line and I guess as the original question was answered, it was marked as solved.

 

[geortsir3]

How did you solve this?

 

[geortsir3]

@lynn_yudi

 

[PeteMox]

Sorry for the old thread revival, would W10 guest on pmox host going to sleep cause this?

Thanks ;-)

 

[Alwin]

@PeteMox, the Win10 is nested? With multiple layers of virtualization, the performance will drop. The result can be slow response times.

 

[PeteMox]

@Alwin : yes it is as I wanted to try Proxmox out before I am able to purchase new hardware.

Thanks for the help

Pierre

 

[alsanbar]

Verifique y cambien la mac de la tarjeta de red