This document describes powering down all devices in an ACI fabric for power maintenance or a physical move, then powering up the fabric.
A high-level process is documented in Cisco APIC System Management Configuration Guide and expanded in the next sections:
1. Confirm that the fabric is fully fit before taking any action. Do not proceed if the cluster is not healthy.
apic1# show controller Fabric Name : Fabric-A Operational Size : 3 Cluster Size : 3 Time Difference : 2450 Fabric Security Mode : STRICT ID Pod Address In-Band IPv4 In-Band IPv6 OOB IPv4 OOB IPv6 Version Flags Serial Number Health ---- ---- --------------- --------------- ------------------------- --------------- ------------------------------ ------------------ ----- ---------------- ------------------ 1* 1 10.0.0.1 0.0.0.0 fd00::1 10.1.1.11 fd00:1::11 6.2(3g) crva- SAL1234AB01 fully-fit 2 1 10.0.0.2 0.0.0.0 fd00::2 10.1.1.12 fd00:1::12 6.2(3g) crva- SAL1234AB02 fully-fit 3 1 10.0.0.3 0.0.0.0 fd00::3 10.1.1.13 fd00:1::13 6.2(3g) crva- SAL1234AB03 fully-fit Flags - c:Commissioned | r:Registered | v:Valid Certificate | a:Approved | f/s:Failover fail/success (*)Current (~)Standby (+)AS
apic1# avread
Cluster:
-------------------------------------------------------------------------
operSize 3
clusterSize 3
fabricDomainName Fabric-A
version apic-6.2(3g)
discoveryMode STRICT
drrMode OFF
kafkaMode OFF
autoUpgradeMode OFF
standbyType warm
optimalLeaders yes
APICs:
-------------------------------------------------------------------------
APIC 1 APIC 2 APIC 3
version 6.2(3g) 6.2(3g) 6.2(3g)
address 10.0.0.1 10.0.0.2 10.0.0.3
oobAddress 10.1.1.11/24 10.1.1.12/24 10.1.1.13/24
oobAddressV6 fd00:1::11/64 fd00:1::12/64 fd00:1::13/64
routableAddress 0.0.0.0 0.0.0.0 0.0.0.0
tepAddress 10.0.0.0/16 10.0.0.0/16 10.0.0.0/16
podId 1 1 1
chassisId 11111111-.-11111111 22222222-.-22222222 33333333-.-33333333
cntrlSbst_serial (APPROVED,SAL1234AB01) (APPROVED,SAL1234AB02) (APPROVED,SAL1234AB03)
active YES YES YES
flags cra- cra- cra-
capabilities [1,2,3]-N [1,2]-N [1,2,3]-N
health 255 255 255
2. Confirm that Cisco Integrated Management Controller (CIMC) access to all APICs is available and working.
3. Confirm that console access to all switches is available and working. If the APICs and switches are being moved to a new location and have new IP addresses and console connections, only CIMC access to the APICs must be validated before the move.
Take a configuration snapshot with AES encryption before the power-down. Use this snapshot only in the rare, worst-case scenario that requires a fabric rebuild:
Run the pre-upgrade script.
Although the system is not being upgraded, the pre-upgrade script can be used as a health-check tool in order to identify faults or issues that must be addressed.
Issues such as VLAN overlap require attention because a power cycle can result in an indeterminate configuration push.
For more information, see the Pre-Upgrade Checklists.
This step consists of shutting down all APICs except one of the Apics used for the shutdown activity.

5. Repeat the APIC selection and shutdown steps for all other APICs except apic-1.
Step 5 consists of shutting down apic-1.
Shut down the last APIC apic-1 using the CIMC GUI. This is because this servet is in read-only mode and does not process a shutdown request through the APIC GUI.
On M1-M4/L1-L4, click Host Power in the toolbar at the top right of any page and choose Shut Down.

On APIC-SERVER-G5, there is no Host Power menu. Access the same actions from Actions in the upper-right corner, and then open the Power submenu.

After you shut down all APICs, power down the switches by turning off their power supplies.
This procedure powers up the fabric through the Cisco Integrated Management Controller (IMC) GUI.
Power on the APICs through the CIMC GUI.
Power on the leaf switches that are connected directly to the APICs.
Power on the spine switches approximately one minute after you power on the leaf switches.
Power on the remaining leaf switches in the fabric.
Confirm that the fabric is fully fit before taking any action.
apic1# show controller Fabric Name : Fabric-A Operational Size : 3 Cluster Size : 3 Time Difference : 2450 Fabric Security Mode : STRICT ID Pod Address In-Band IPv4 In-Band IPv6 OOB IPv4 OOB IPv6 Version Flags Serial Number Health ---- ---- --------------- --------------- ------------------------- --------------- ------------------------------ ------------------ ----- ---------------- ------------------ 1* 1 10.0.0.1 0.0.0.0 fd00::1 10.1.1.11 fd00:1::11 6.2(3g) crva- SAL1234AB01 fully-fit 2 1 10.0.0.2 0.0.0.0 fd00::2 10.1.1.12 fd00:1::12 6.2(3g) crva- SAL1234AB02 fully-fit 3 1 10.0.0.3 0.0.0.0 fd00::3 10.1.1.13 fd00:1::13 6.2(3g) crva- SAL1234AB03 fully-fit Flags - c:Commissioned | r:Registered | v:Valid Certificate | a:Approved | f/s:Failover fail/success (*)Current (~)Standby (+)AS
apic1# avread
Cluster:
-------------------------------------------------------------------------
operSize 3
clusterSize 3
fabricDomainName Fabric-A
version apic-6.2(3g)
discoveryMode STRICT
drrMode OFF
kafkaMode OFF
autoUpgradeMode OFF
standbyType warm
optimalLeaders yes
APICs:
-------------------------------------------------------------------------
APIC 1 APIC 2 APIC 3
version 6.2(3g) 6.2(3g) 6.2(3g)
address 10.0.0.1 10.0.0.2 10.0.0.3
oobAddress 10.1.1.11/24 10.1.1.12/24 10.1.1.13/24
oobAddressV6 fd00:1::11/64 fd00:1::12/64 fd00:1::13/64
routableAddress 0.0.0.0 0.0.0.0 0.0.0.0
tepAddress 10.0.0.0/16 10.0.0.0/16 10.0.0.0/16
podId 1 1 1
chassisId 11111111-.-11111111 22222222-.-22222222 33333333-.-33333333
cntrlSbst_serial (APPROVED,SAL1234AB01) (APPROVED,SAL1234AB02) (APPROVED,SAL1234AB03)
active YES YES YES
flags cra- cra- cra-
capabilities [1,2,3]-N [1,2]-N [1,2,3]-N
health 255 255 255
| Revision | Publish Date | Comments |
|---|---|---|
1.0 |
16-Oct-2026
|
Initial Release |