M4300-28G ip management port fails after restart in a stacked configuration.

[JosLaan]

Hello,

we have a stacked configuration from 6 times a M4300-28G switch (firmware version 12.0.2.9), the stack up and running.

The ip management  port (IP4) is set to vlan 1 and via the web interface all settings en configurations can be seen.

After a restart of the stack (all the swithes go down) the web interface fails and de ip address of the ip management port can't be accessed any more.

After setting the ip address from the ip management port again (via CLI command line interface port)  the ip management port is available again.

What can be the reason that the ip management port isn't available after a restart?

best regards,

Jos Laan

[DanielZhang · Accepted solution]

Hi Jos,

We have received the TechSupport file just you sent to us.

This error may cause by serviceport enhancement in 12.0.2.x which add a new default IP address  192.168.0.239/24.

We also find a similar issue just like yours which will only appear on slave stack unit.

Maybe this issue have same root cause with yours.(default serviceport IP address)

So We will share the new firmware to you when the similar issue fix.

Please use a workaround method to avoid the error for now if possible:

Set the an another IP address on management VLAN.

Such as:

192.168.1.x/24

192.168.10.x/24

Thanks for your understanding and patience.

[LaurentMa]

Hi JosLaan

Welcome to the community!

One possible reason would be: configuration lives in non-volatile memory, when it's saved. Until it is saved, it is applied in RAM only.
When you configure the switch, it is not saved automatically. Are you using Web Interface (GUI)? A Save button should do it.

I believe you had configured the Management VLAN IP address and clicked on "Apply" button on that GUI page.

By Applying the configuration, you save it in RAM and make it work in running config.

But the startup config remains untouched as this point: you should go to Maintenance -> Save Config and click SAVE button to write the running config into the startup config. Only the startup config in non-volatile memory is getting used after a reboot.

Let us know if this helps!

Regards,

[JosLaan]

Hello Laurent,

thank you for your explanation, I checked the settings for ip manangement after reboot and they seems ok:

(M4300-28G) (Config)#show ip management

IPv4 Interface Status.......................... Down
IPv4 Management Interface...................... vlan 1
IP Address..................................... 192.168.4.1
Subnet Mask.................................... 255.255.0.0
Method......................................... Manual
Routing Mode................................... Enable
Default Gateway................................ 0.0.0.0

Source Interface............................... vlan 1
Burned In MAC Address.......................... B0:7F:B9:42:C6:A4
Locally Administered MAC address............... 00:00:00:00:00:00
MAC Address Type............................... Burned In

When I change the ip-address (see below) to another address and back again to the original value then the Web Interface will working again.

(M4300-28G) (Config)#ip management vlan 1 192.168.4.2 255.255.0.0

(M4300-28G) (Config)#ip management vlan 1 192.168.4.1 255.255.0.0

Or disconnecting the ethernet cable from the switch and plug it in again, then also de Web Interface will working again.

Very strang!

When logging (via CLI USB port) on the master switch a few message appear (see below) but the stacking configuration looks working well after full startup:

U-Boot 2012.10-00216-g55eda36 (May 25 2016 - 15:31:45)VerNo=1.0.0.8

+ + Checking for application
Extracting application from .stk file...done.
Loading Operational code...done.
Uncompressing apps.lzma
DMA pool size: 8388608

started!

(Unit 1)>

Applying Global configuration, please wait ...

Applying Interface configuration, please wait ...
Trying to attach more units.....

Trying to attach more units.....

Trying to attach more units.....

Trying to attach more units.....

Trying to attach more units.....
disc ERR: Timeout occurred
ATP: TX timeout, seq 164. c6:d0 cli 778. to 4 tx cnt 21.
RPC - Timeout to CPU: b0:7f:b9:42:c6:fc.
RPC - Timeout to CPU: b0:7f:b9:42:c6:d0.

User:

After first installing the switches in a stack configuration and connected to our existing network we had a BPDU error message (every 10 sec.), after a firmware update (from 12.0.0.11 to 12.0.2.9) this message was fixed.

Have you any suggestions what we can do?

best regards,

Jos Laan

Protonic Holland

[JosLaan]

Hello Laurent,

are there know issues with firmware version 12.0.2.9 M4300-28G switch?

Have you any idea why the ip management port fails to response after a reboot?

best regards,

Jos Laan

Protonic Holland

[LaurentMa]

Hi JosLaan

No, there is no known issue on 12.0.2.9 so I am very perplex on this Management IP issue.

I originally thought your startup config wasn't properly saved, but that's not the case as we see the IPv4 Interface Status.......................... Down with the right IP address after a reboot. So there is another conflict affecting the VLAN 1 IP interface in startup config, since this interface is down.

Do you have conflicting configuration on OOB port, simultaneously? Can I ask you to follow indications from this post, in order to remove the configuration from the OOB Service Port? 

Regards,

[JosLaan]

Hello Laurent,

OOB port is set:

(M4300-28G) (Config)#show serviceport

Interface Status............................... Down
IP Address..................................... 0.0.0.0
Subnet Mask.................................... 0.0.0.0
Default Gateway................................ 0.0.0.0
IPv6 Administrative Mode....................... Enabled
Configured IPv4 Protocol....................... None
Configured IPv6 Protocol....................... None
IPv6 AutoConfig Mode........................... Disabled
Burned In MAC Address.......................... B0:7F:B9:42:C6:A5

ip-managment shows:

(M4300-28G) (Config)#show ip management

IPv4 Interface Status.......................... Up
IPv4 Management Interface...................... vlan 1
IP Address..................................... 192.168.4.1
Subnet Mask.................................... 255.255.0.0
Method......................................... Manual
Routing Mode................................... Enable
Default Gateway................................ 0.0.0.0

Source Interface............................... vlan 1
Burned In MAC Address.......................... B0:7F:B9:42:C6:A4
Locally Administered MAC address............... 00:00:00:00:00:00
MAC Address Type............................... Burned In

after reboot the show ip-managment command shows:

(M4300-28G) (Config)#show ip management

IPv4 Interface Status.......................... Up
IPv4 Management Interface...................... vlan 1
IP Address..................................... 192.168.4.1
Subnet Mask.................................... 255.255.0.0
Method......................................... Manual
Routing Mode................................... Enable
Default Gateway................................ 0.0.0.0

Source Interface............................... vlan 1
Burned In MAC Address.......................... B0:7F:B9:42:C6:A4
Locally Administered MAC address............... 00:00:00:00:00:00
MAC Address Type............................... Burned In

But I can't ping to it!

When switching from ip management address:

(M4300-28G) (Config)#ip management vlan 1 192.168.5.1 255.255.0.0

Ping works! (to the new address 192.168.5.1).

We have 6 stacked M4300-28G switches with glass and copper connections: 

(M4300-28G) (Config)#show stack-port

Configured Running
Stack Stack Link Link
Unit Interface Type Product Name Mode Mode Status Speed (Gb/s)
---- ---------------- ------- ----------------- ---------- ---------- ---------- ------------
1    0/25             Ethernet Ethernet Link Down 10
1    0/26             Ethernet Ethernet Link Down 10
1    0/27 SFP+ AXM764 Stack    Stack    Link Up   10
1    0/28 SFP+ AXM764 Stack    Stack    Link Up   10
2    0/25             Ethernet Ethernet Link Down 10
2    0/26             Ethernet Ethernet Link Down 10
2    0/27 SFP+ AXM764 Stack    Stack    Link Up   10
2    0/28 SFP+ AXC761 Stack    Stack    Link Up   10
3    0/25             Ethernet Ethernet Link Down 10
3    0/26             Ethernet Ethernet Link Down 10
3    0/27 SFP+ AXC761 Stack    Stack    Link Up   10
3    0/28 SFP+ AXM764 Stack    Stack    Link Up   10
4    0/25             Ethernet Ethernet Link Down 10
4    0/26             Ethernet Ethernet Link Down 10
4    0/27 SFP+ AXM764 Stack    Stack    Link Up   10
4    0/28 SFP+ AXM764 Stack    Stack    Link Up   10
5    0/25             Ethernet Ethernet Link Down 10
5    0/26             Ethernet Ethernet Link Down 10
5    0/27 SFP+ AXM764 Stack    Stack    Link Up   10
5    0/28 SFP+ AXC761 Stack    Stack    Link Up   10
6    0/25             Ethernet Ethernet Link Down 10
6    0/26             Ethernet Ethernet Link Down 10
6    0/27 SFP+ AXC761 Stack    Stack    Link Up   10
6    0/28 SFP+ AXM764 Stack    Stack    Link Up   10

What can it be?

Suggestions are welcome!

best regards,

Jos Laan
Protonic Holland

[LaurentMa]

Hi JosLaan

I'm sorry for the issue. At this point we're quite convinced there's mismatch configuration or conflicting settings that only Tech-Support export file will divulgate.

Please issue “show tech-support” CLI command and export the file for instance in GUI in Maintenance/Export/HTTP file/Tech-Support file. Please kindly send it to me using PM and we'll closely look at it.

Regards,

[Jedi_Exile]

Can you check for following instead.

Show arp | include B0:7F:B9:42:C6:A4

Verify the arp is there for the IP you configured after the ping stops.  If not, that may be reason for ping not working  Also, what is source IP from which you are pinging.  Is it the same VLAN as management on VLAN 1.

Also for clarification, I assume you are not using the OOB port since OOB status is down.  OOB should be seperated from management vlan due to seperation of those, unless you connect OOB port back into a switch port to get service port online. You are however configuring a management vlan which by default is vlan 1.

Also, can you output and PM me the tech support file.   To generate the tech support file, do following via CLI at enable prompt

show tech-support

Wait for file to be generated and then download it either via TFTP via CLI or use the Web GUI to download the Tech support file from the switch.

[LaurentMa]

Hi Jedi_Exile

Hopefully JosLaan already produced and sent tech-support files (before and after his issue) to me.

I am going to send them to you. 

Regards,

[DanielZhang]

Hi Jos,

We have reproduced a similar issue just like yours.

I want to confirm that which unit connected with management port when you faced this issue?

in other words, Did you see this issue when the management port connected to  master unit?

You can use following commands to check which unit is master.

1.  Login switch and input username&password

2.  Enable.

3.  Show switch

[JosLaan]

Hello Daniel,

in our situation it makes no sense of the managemement port is connected with the master unit or another unit in the stack, in both cases ip management fails after restart.

When I change the ip-address (see below) to another address and back again to the original value then the Web Interface will working again.

(M4300-28G) (Config)#ip management vlan 1 192.168.4.2 255.255.0.0
(M4300-28G) (Config)#ip management vlan 1 192.168.4.1 255.255.0.0

Or disconnecting the ethernet cable from the switch and plug it in again, then also de Web Interface will working again,

best regards,

[JosLaan]

Hello Laurent,

ip management port works fine with firmware release 12.0.0.11!

But with this firmware version we have a warning every two seconds:

"DOT1S[dtlTask]: dot1s_txrx.c(269) 839 %% dot1sBpduReceive(): Discarding the BPDU, since it is an invalid BPDU type"

That was the reason we switch from firmware version 12.0.0.11 to 12.0.2.9 (where above issue BPDU is fixed),

Is there are any respons about the tech support files I send?

best regards,

[DanielZhang · Accepted solution]

Hi Jos,

We have received the TechSupport file just you sent to us.

This error may cause by serviceport enhancement in 12.0.2.x which add a new default IP address  192.168.0.239/24.

We also find a similar issue just like yours which will only appear on slave stack unit.

Maybe this issue have same root cause with yours.(default serviceport IP address)

So We will share the new firmware to you when the similar issue fix.

Please use a workaround method to avoid the error for now if possible:

Set the an another IP address on management VLAN.

Such as:

192.168.1.x/24

192.168.10.x/24

Thanks for your understanding and patience.

[JosLaan]

Hello Daniel,

is there any prospect when the new firmware will be released and available?

best regards,

Jos Laan

[DanielZhang]

Hi Jos,

Our engineer team are working on it.

I will reply this topic when the fix get ready.