Cat 9300 not registering 8 port 10Gb module

[slfashley]

Hi Everyone,

We have a C9300-24T switch, (switch2) in a 2 switch stack that wont initialize a c9300-nm-8x module.

The c9300-nm-8x module works fine in other switches.

Switch 1 in the stack has a c9300-nm-8x module working perfectly.

After removing the c9300-nm-8x 10Gb module and re-installing the previously working c9300-nm-4g module, it then also would not initialize via hot swap.

It was necessary to power off and cold boot the switch to reinitialize the 4 x 1Gb module to get it working again.

Cold boot did not work for the 8port 10Gb module at all.

Basic Stack details

Switch Ports Model SW Version SW Image Mode
------ ----- ----- ---------- ---------- ----
* 1 41 C9300-24T 17.06.04 CAT9K_IOSXE INSTALL
2 41 C9300-24T 17.06.04 CAT9K_IOSXE INSTALL

Switch# Role Mac Address Priority Version State
-------------------------------------------------------------------------------------
*1 Active 04eb.403d.b800 15 V03 Ready
2 Standby 482e.728d.e500 1 V05 Ready

The only thing i notice is switch 2 is HW ver 05 and active switch 1 is HW ver 03. this is the only C9300 stack we have in service that has different HW versions, tho all the posts i have seen on Community suggest HW version is irrelevant ?

Has anyone encountered a similar issue.

The problem we have is this switch stack is in other part of the country making troubleshooting a bit complicated .. perhaps we are missing something that can be fixed remotely ?

many thanks

[Leo Laohoo · Accepted solution]

---

@slfashley wrote:
It was necessary to power off and cold boot the switch to reinitialize the 4 x 1Gb module to get it working again.

---

Cold reboot the stack several times. 

And no, I am not joking.

[marce1000]

 -  @Leo Laohoo  As usual I appreciate your eternal knowledge forever (LOL) , but then it would be added value to explain why this can help.

 M.

[Leo Laohoo]

---

@marce1000 wrote:
 -  @Leo Laohoo  As usual I appreciate your eternal knowledge forever (LOL) , but then it would be added value to explain why this can help.

---

The process that manages the NM is stuck.  "reload" command will not change anything.  Cold reboot is the only way. I suspect two or three cold-reboot will "unlock" this process back to enable the NM to re-initialize.  

Cold-Reboot is going to be everyone's favorite "workaround" because IOS-XE is very buggy.

[slfashley]

thank you ... we will definitely try this when we visit the site again.

[marce1000]

   >...The problem we have is this switch stack is in other part of the country making troubleshooting a bit complicated 
 - It would be advisable to scrutinize the boot process (e.g.). Perhaps someone could connect a console and keep a PuTTY log and send it to you  , also check the logs just after boot , in both cases look for : https://bst.cisco.com/bugsearch/bug/CSCvh85987 and or if related info's come up. In that context you could have a go with :
                  https://software.cisco.com/download/home/286313829/type/282046477/release/Bengaluru-17.6.5

 M.

[slfashley · Accepted solution]

thank you .. will checkout that bug

[Leo Laohoo]

Just a sh*ts-n-giggles, try upgrading to 17.10.1.

[twalsh1]

Encountered this running 17.09.05. Both modules in a two switch stack disappeared after power outage. Power cycles cleared one but only removing the module between power cycles finally cleared the second.