06-30-2018 05:12 AM - edited 03-01-2019 05:35 AM
I ran into an issue where the APIC/nodes were using an old Infra VLAN. Tried to reset the devices, but it still kept coming back with the old Infra VLAN.
On version 3.x what is the correct procedure for both APIC and nodes to factory reset the devices to wipe any existing Infra VLAN info?
I believe we tried "acidiag touch clean", but that did not help wipe the Infra VLAN info. Where in the documentation does it mention the steps to factory reset in version 3.x? I was only able something in the troubleshooting guide here - https://www.cisco.com/c/en/us/td/docs/switches/datacenter/aci/apic/sw/1-x/troubleshooting/b_APIC_Troubleshooting/b_APIC_Troubleshooting_chapter_01001.html but looks like this is for version 1.x.
Solved! Go to Solution.
06-30-2018 03:07 PM - edited 07-31-2019 04:19 AM
Hi Manuel,
The document you referred to is still relevant for v3.x. Last week I did three successul complete resets using the method below: My method requires that you have OOB management or console access to each switch and CIMC management or console access to each APIC.
On each leaf and spine switch, run the setup-clean-config.sh script. DO NOT reload the switch yet - if you do, the switch will re-learn its config from the APIC and you'll need to return to Step 1 again later.
I was able to power down each switch after the setup-clean-config.sh script had run, however this is not strictly recessary, but it does make the process more failsafe.
Spine201# setup-clean-config.sh
In progress
In progress
In progress
In progress
In progress
Done
Spine201#
Run the following commands on each APIC:
apic# acidiag touch clean This command will wipe out this device. Proceed? [y/N] y
apic# acidiag touch setup
This command will reset the device configuration, Proceed? [y/N] y
DO not reload the APICs until the above has been completed on each APIC, otherwise if one APIC reloads before you have reloaded the others, it may re-join the cluster and re-learn its configuration.
Run the following command simultaneously (or within a couple of minutes) on each APIC:
apic# acidiag reboot
This command will restart the this device, Proceed? [y/N] y
While the APICs are reloading, you can run the reload command on the switches, but be prepared for a lot of noise from the spines. In my case where I was able to power off the switches, I did not power on the switches until I had completed the initial setup on each APIC, and then I powered them on in the order I wanted them to be discovered - that way I KNEW which switch was being discovered in the fabric discovery process.
When the APICs have reloaded, go through the initial setup on each one, making sure you are consistent with the values you use for the infrastructure VLAN and TEP pools.
I hope this helps
Don't forget to mark answers as correct if it solves your problem. This helps others find the correct answer if they search for the same problem
06-30-2018 01:47 PM
To remove all config and wipe the APIC, login to console with monitor and keyboard and use:
acidiag touch setup
acidiag reboot
If you do this on all three APICs you will have a completely fresh environment to work with
06-30-2018 03:07 PM - edited 07-31-2019 04:19 AM
Hi Manuel,
The document you referred to is still relevant for v3.x. Last week I did three successul complete resets using the method below: My method requires that you have OOB management or console access to each switch and CIMC management or console access to each APIC.
On each leaf and spine switch, run the setup-clean-config.sh script. DO NOT reload the switch yet - if you do, the switch will re-learn its config from the APIC and you'll need to return to Step 1 again later.
I was able to power down each switch after the setup-clean-config.sh script had run, however this is not strictly recessary, but it does make the process more failsafe.
Spine201# setup-clean-config.sh
In progress
In progress
In progress
In progress
In progress
Done
Spine201#
Run the following commands on each APIC:
apic# acidiag touch clean This command will wipe out this device. Proceed? [y/N] y
apic# acidiag touch setup
This command will reset the device configuration, Proceed? [y/N] y
DO not reload the APICs until the above has been completed on each APIC, otherwise if one APIC reloads before you have reloaded the others, it may re-join the cluster and re-learn its configuration.
Run the following command simultaneously (or within a couple of minutes) on each APIC:
apic# acidiag reboot
This command will restart the this device, Proceed? [y/N] y
While the APICs are reloading, you can run the reload command on the switches, but be prepared for a lot of noise from the spines. In my case where I was able to power off the switches, I did not power on the switches until I had completed the initial setup on each APIC, and then I powered them on in the order I wanted them to be discovered - that way I KNEW which switch was being discovered in the fabric discovery process.
When the APICs have reloaded, go through the initial setup on each one, making sure you are consistent with the values you use for the infrastructure VLAN and TEP pools.
I hope this helps
Don't forget to mark answers as correct if it solves your problem. This helps others find the correct answer if they search for the same problem
07-26-2019 04:30 PM
Thanks !
it is quiete importan reload times make sure any switch syncs with old state from another.
07-28-2019 08:02 PM
07-28-2019 09:59 PM
Not quite sure why you are repeating information straight from the solution above. Was Step 1 not written clearly enough?
11-07-2022 08:07 AM
Also you can break into rommon and clean up with 'cmdline clear_config'
loader>cmdline clear_config
Discover and save your favorite ideas. Come back to expert answers, step-by-step guides, recent topics, and more.
New here? Get started with these tips. How to use Community New member guide