Mailing List Archive: Management stops working on Turboiron TI24X

Dear fellow Foundr^W^W Brocade-Users,

we have some issues with a Turboiron TI24X that is used as "stupid"
switch. After we have switched some traffic to this device (basically
passing trough a 20G LACP to a single 10G port over a tagged VLAN) the
box stops responding to all management traffic (in-band) from time to
time for up to 15 minutes. After the box has recovered the logs do not
show any bad signs (and traffic is flowing truth the TI all the time).
We are running latest switching code (TIS08001c.bin), but we have also
seen this issue with other software versions and devices in the past -
but always linked to that specific VLAN we have now moved to the new
TI24X.

Using "dm raw" i have not seen anything suspicious, just some IPv6 ND
and ARP stuff flying around:
RX [b9b1df]a8d0e55e4dc3 ->Broadcast ARP Req S=80.81.193.39
D=80.81.194.60 port: 5
RX [b9b1df]2001:7f8::1a27:5051:c19d ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1df]2001:7f8::1a27:5051:c19d ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e0]fe80::669e:f3ff:fe22:3e03 ->ff02::1:ff00:2 ICMP Type=135 port: 5
RX [b9b1e0]80711f7477c5 ->Broadcast ARP Req S=80.81.194.56
D=80.81.194.111 port: 5
RX [b9b1e0]fe80::8271:1fff:fe16:5f19 ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e0]00243895b600 ->Broadcast ARP Req S=80.81.194.66
D=80.81.194.22 port: 5
RX [b9b1e0]2001:7f8::15cc:0:1 ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e0]2001:7f8::15cc:0:1 ->ff02::1:ff00:2 ICMP Type=135 port: 5
RX [b9b1e0]2001:7f8::15cc:0:1 ->ff02::1:ff00:2 ICMP Type=135 port: 5
RX [b9b1e0]fe80::46d3:caff:fe63:ec10 ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e0]fe80::669e:f3ff:fee0:1580 ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e1]fe80::669e:f3ff:fee0:1580 ->ff02::1:ff00:1 ICMP Type=135 port: 5
RX [b9b1e1]fe80::669e:f3ff:fee0:1580 ->ff02::1:ff00:2 ICMP Type=135 port: 5
RX [b9b1e1]fe80::669e:f3ff:fee0:1580 ->ff02::1:ff00:1 ICMP Type=135 port: 5

(The traffic passing trough the device is traffic from a Internet
exchange point, so quite "clean" stuff).

Any ideas on that?

Thanks,
-chris
_______________________________________________
foundry-nsp mailing list
foundry-nsp@puck.nether.net
http://puck.nether.net/mailman/listinfo/foundry-nsp

On 27/10/2014 18:14, Chris Hellkvist wrote:
> we have some issues with a Turboiron TI24X that is used as "stupid"
> switch. After we have switched some traffic to this device (basically
> passing trough a 20G LACP to a single 10G port over a tagged VLAN) the
> box stops responding to all management traffic (in-band) from time to
> time for up to 15 minutes.

the management ports on these boxes can freeze up if they get spammed with
certain type of traffic - from incidents in the past, it may be related to
either broadcast traffic or looped traffic. Management on these boxes is
connected to the forwarding plane in a weird sort of way, so if you use
in-band management, you may be running into the same problem.

Nick

_______________________________________________
foundry-nsp mailing list
foundry-nsp@puck.nether.net
http://puck.nether.net/mailman/listinfo/foundry-nsp