The PowerChute Network Shutdown 5.2 manual’s UPS Details example shows HTTPS, port 443 and two illustrative NMC addresses. Enter only the cards that protect the actual load, then complete the trusted-certificate and registration checks. The pictured addresses are vendor examples, not values to copy or proof of a working shutdown path.[3] View full-size image
Secure the APC Network Management Card and prove the PowerChute shutdown path before relying on either during an outage. Match the card, UPS application and PowerChute version to the equipment actually protected, restrict management access, and verify trusted HTTPS communication. Then test the approved sequence from event receipt through application and host shutdown with battery reserve remaining. A working web login or a selected protocol is not evidence that the correct servers will stop safely.
Schneider Electric’s current Network Management Card 3 user guide covers AP9640, AP9641, and AP9643 models, UPS and outlet control, IPv4 and IPv6, notifications, logs, supported management protocols, environmental monitoring, firmware, and PowerChute integration. Its NMC3 security handbook treats authentication, encrypted services, certificates, access control, event records, and service reduction as explicit security controls. PowerChute Network Shutdown 5.2 adds its own supported platforms, UPS topologies, event actions, shutdown sequences, logging, SSH actions, SNMP integration, and environment-specific behavior.[1]
This runbook joins those layers without publishing live infrastructure details. IP addresses, hostnames, UPS names, authentication phrases, certificates, community strings, account lists, firewall rules, server inventory, event logs, and shutdown commands belong in restricted operational systems. Public documentation should describe control objectives and verified procedure, not hand an attacker a management map. Changes should be tested on the exact firmware, PCNS version, platform, UPS topology, and workload before broad deployment.
ALLMSP's in-house team helps organizations in Lawrenceville, Suwanee, Gwinnett County, Metro Atlanta, and across Georgia secure APC network management cards and PowerChute shutdown, then validates the result and documents the support path.
Key decisions at a glance
- Inventory every AP9640, AP9641, AP9643, embedded NMC3, UPS application, firmware version, UPS association, IP and DNS record, certificate, account, protocol, notification, integration, and support owner in a protected system.
- Place NMC access on a restricted management plane, use trusted certificates and unique roles, prefer HTTPS, SSH, and SNMPv3, disable unused legacy services, synchronize time, centralize logs, and test break-glass access.
- Model PowerChute Network Shutdown from the actual single, redundant, parallel, outlet-group, virtualization, and application topology rather than copying settings from an unrelated UPS or site.
- Test events and shutdown timing from alert receipt through workload stop, host shutdown, reserve, UPS turnoff, utility return, and controlled startup without forcing an uncontrolled production discharge.
- Treat NMC firmware, PCNS releases, operating systems, certificates, credentials, firewall rules, configuration backups, subscriptions, and retirement as one governed lifecycle with staged change and rollback.
Build a Protected NMC3 and PowerChute Dependency Inventory
Discover NMCs through purchasing, UPS SmartSlot inventories, DHCP and IP address management, management VLANs, certificate records, DNS, monitoring, SNMP managers, configuration systems, firewall rules, Schneider Electric subscriptions, and physical rack inspection. For each device, record exact card or embedded controller model, serial, hardware platform, boot monitor, APC operating system and application firmware, UPS model and UPS ID, physical slot, MAC, protected management address, DNS name, certificate subject and expiry, enabled protocols, local and remote users, authentication source, roles, SNMP managers and trap receivers, email or syslog destinations, environmental sensors, outlet-group permissions, PCNS registrations, time source, configuration backup, firmware entitlement, warranty, site, service owner, security owner, and break-glass custodian. Keep sensitive values in a secret manager rather than the asset record.
Map the dependency graph around each PowerChute instance. Name the protected operating system or virtual appliance, supported PCNS release, licensing state, Java or appliance baseline where applicable, firewall path, NMC endpoints, UPS configuration type, outlet group, virtualization manager, clusters, hosts, virtual machines, storage, network services, identity, DNS, time, scripts, application owners, shutdown priority, startup owner, and maximum tolerated outage. Schneider Electric publishes separate PCNS guidance for standard, VMware, Hyper-V, and other environments because the actions and prerequisites differ. Do not apply a generic server script to a hyperconverged cluster or assume one PCNS node protects equipment connected to another UPS.
Reconcile power and data dependencies. PCNS cannot contact the NMC if the management switch, firewall, DNS, or authentication service loses power too early. A storage array cannot stop safely after the hypervisor hosting its management path is gone. Keep critical network, time, identity, and logging services available for the intended sequence, and document local fallback when centralized dependencies are unavailable. The inventory is complete only when each UPS, NMC, PCNS instance, protected host, outlet group, event action, credential owner, and recovery path has an accountable relationship.
Harden APC NMC3 Access, Protocols, Certificates, and Logging
Place NMC3 devices on a dedicated infrastructure management segment with explicit routes from approved administration, monitoring, logging, and PowerChute systems. Deny direct user, guest, wireless, and Internet access. Use source restrictions, firewall logging, and an administrative jump path appropriate to the organization, do not depend on obscurity or a nonstandard port.
Replace initial credentials immediately, create named administrator and read-only roles for accountable use, remove or disable stale users, restrict powerful device actions, and keep a tested break-glass account whose secret is vaulted, monitored, and rotated after use. The current NMC3 security handbook should decide which authentication features and cryptographic options the installed firmware supports. Prefer HTTPS with a trusted, verified certificate, SSH for command access, and SNMPv3 configured with both authentication and privacy. SNMPv3 by itself is not an encryption setting: check the supported privacy protocol and matching manager configuration rather than selecting a no-privacy profile. Disable HTTP, Telnet, FTP, SNMPv1/v2c, unused discovery, unused industrial protocols, and other unneeded services after dependencies are migrated. If a legacy monitoring platform temporarily requires an older protocol, document exact source, read-only scope where possible, compensating network restriction, owner, expiry, and replacement project. Do not export configuration files casually, they can reveal network and integration state even when passwords are obscured.
Establish certificate governance. Use a hostname that matches the certificate, protect the private key, record issuing authority and expiry, renew before the operational freeze window, and test every browser, API, PCNS, monitoring, and automation client that validates the endpoint. Self-signed certificates may encrypt traffic but do not establish centrally trusted identity.
Configure authoritative time and verify timezone so event chronology can be reconciled across UPS, NMC, PCNS, hypervisor, operating system, firewall, and ticketing systems.
Send security and power events to resilient central logging and alert on authentication failure, account change, configuration change, protocol enablement, certificate problem, firmware action, communication loss, on-battery, low runtime, bypass, overload, battery fault, and control operations. Review event destinations after personnel, mail, logging, or monitoring changes, an alert configured to a retired system is no alert. Finally, back up a sanitized, access-controlled configuration and prove local console or recovery access before disabling the last known management path.[2]
For PowerChute Network Shutdown 5.2, treat the NMC certificate import as an explicit trust decision. The standard guide says to add only NMC certificates you recognize and trust to the PowerChute-keystore. It also warns that PowerChute continues trusting certificates in that keystore after expiry or revocation; the operator must remove certificates that are no longer needed. Verify identity and fingerprint through the approved channel, record the trusted material, and include keystore cleanup and a connection retest in certificate rotation. Do not solve an unrecognized certificate by blindly accepting the prompt.[3]
Configure and Test PowerChute Network Shutdown 5.2 by Topology
Select the PCNS edition and installation path from Schneider Electric’s compatibility and installation guidance for the exact platform. Record operating system or appliance build, PCNS 5.2 release, license, NMC support, firewall requirements, certificate behavior, virtualization manager, service account, storage and cluster prerequisites, and upgrade path before installation. Register only the NMC addresses and UPS topology that physically protect the host or cluster: single UPS, redundant UPS, parallel system, advanced configuration, or outlet group. Use a unique protected authentication phrase and ensure the NMC and PCNS values agree without exposing it in tickets or scripts. Separate NMC administrative credentials from application authentication and give virtualization or operating-system service accounts the minimum permissions documented for the selected actions.
Translate business priorities into event configuration. Define which on-battery, low-runtime, communication-loss, UPS-off, bypass, battery, overload, or environmental events start notifications, commands, migration, workload shutdown, host shutdown, or UPS turnoff. Time the sequence backward from usable battery reserve. Allow for event confirmation, scripts, application drain, database checkpoints, virtual-machine prioritization, storage stop, host shutdown, command timeout, communication variance, and battery aging. A long delay that consumes all reserve is not resilience, an aggressive delay that stops production during a brief transfer may also be wrong.
Validate with a tabletop review first, then a controlled functional test. Confirm NMC-to-PCNS authentication, certificate trust, event receipt, log timestamps, firewall behavior, command permissions, application and VM priority, outlet-group registration, and notification paths. Inject only approved test conditions or use vendor-supported test mechanisms, do not create an uncontrolled outage to prove software.
During a planned failover exercise, observe the full chain from UPS event through PCNS action, workload stop, host power state, remaining runtime, UPS behavior, utility return, and startup ownership.
Test failure cases too: one NMC unreachable in a redundant topology, DNS unavailable, logging offline, expired certificate in a lab, a shutdown command that returns nonzero, a host already down, and an operator cancel or stop condition. Preserve logs and timings without exposing infrastructure secrets. A passing test documents expected event, actual timestamps, variance, reserve, exceptions, corrective owner, and next rehearsal.[3][4]
Worked shutdown budget: a fictional twelve-minute runtime window
This arithmetic is a planning example, not measured UPS performance or recommended PowerChute field values. Suppose an approved rehearsal establishes 720 seconds of usable runtime at the test load. Compare the observed sequence and documented delays with the reserve requirement; topology-specific operations may overlap or have additional dependencies.
| Planning allowance | Example seconds | What to validate in the actual rehearsal |
|---|---|---|
| Event confirmation | 60 | The intended NMC event reaches the correct PowerChute instance and starts only the approved action. |
| Application drain and shutdown commands | 90 | Applications complete their safe stop, and command failure or timeout is visible to the operator. |
| Host shutdown | 120 | The protected host reaches the required power state before its supporting power/network path disappears. |
| Timing variance allowance | 60 | Recorded variance and communication delays remain within the approved budget. |
| Required reserve | 180 | Reserve remains after the action chain; do not spend it extending a delay. |
| Total / headroom | 510 / 210 | 60 + 90 + 120 + 60 + 180 = 510 seconds, leaving 210 of the assumed 720 seconds. Remeasure after load or topology changes. |
Do the tabletop and controlled rehearsal before altering a production shutdown policy. Confirm the exact UPS/outlet topology, the management network’s continued power, stop authority and utility-return startup owner; the numbers above must not be copied into a different installation.[3][4]
Patch, Recover, Audit, and Retire the APC Power Control Plane
Review Schneider Electric advisories, NMC3 firmware, the NMC application package for the attached UPS, PCNS releases, supported operating systems, certificates, cryptographic policy, and virtualization compatibility on a scheduled cadence. Version 3 and later NMC3 firmware uses the Secure NMC System tool and associated subscription route described by Schneider Electric, older 2.5-and-lower methods and files are not interchangeable with that lifecycle. Identify the correct hardware platform and application before every update. A firmware package for Smart-UPS is not a rack-PDU package, and a card can become unavailable if an upgrade is interrupted or mismatched.
Stage change in a representative lab or low-risk site, back up configuration, verify current alarms and power state, record hash or signed-source evidence where published, schedule an operational window, maintain local access, and monitor the NMC reboot separately from the attached UPS load. After change, verify firmware, application, UPS communication, certificates, users, roles, protocols, firewall access, SNMP, logs, alerts, PCNS registration, sensors, outlet groups, time, and configuration drift.
For PCNS upgrades, validate supported source version, license, platform, Java or appliance requirements, backup, service state, virtualization privileges, event settings, scripts, and end-to-end shutdown after the upgrade.
Keep a recovery package that includes exact model, console cable or access method, IP recovery procedure, trusted firmware source, current configuration, certificate and key custody, account recovery, PCNS reinstall media, license record, topology, shutdown settings, and contacts. Practice restoration without using production secrets in an uncontrolled lab.
Audit quarterly for shared or dormant accounts, legacy protocols, expired certificates, unexpected configuration changes, open firewall paths, missing logs, failed notifications, stale DNS, unsupported firmware, duplicate IPs, orphaned PCNS instances, unowned scripts, unused service accounts, and shutdown tests older than policy allows.
When a UPS, NMC, server, or site retires, remove PCNS registration and integrations before power disappears, revoke certificates and credentials, delete DNS and firewall objects, archive required logs and configuration, update monitoring, clear reusable NMC configuration through the supported process, remove subscription or license assignments, and reconcile physical and logical assets. Retirement is complete when no management route, trusted certificate, service account, alert destination, automation action, or shutdown dependency still points at the removed device.[5][6]
Official product documentation and ALLMSP resources
- ALLMSP APC Hardware Support
- ALLMSP Hardware Support
- ALLMSP IT Consulting
- ALLMSP Managed IT Services
- ALLMSP Cybersecurity Services
- Contact ALLMSP
Validate the usable runtime and load assumptions before selecting the PowerChute delay budget. Commission APC Smart-UPS for load, runtime and failover.
If switched outlets or rack power paths are involved, reconcile their ownership and control dependencies with the shutdown sequence. Operate APC NetShelter rack PDUs and outlet control.
Frequently Asked Questions
Why is an APC Network Management Card a privileged system?
It can disclose power and environmental state, change configuration, control outlet groups, send commands, and participate in shutdown. Protect it like other infrastructure control planes, not like a passive web sensor.
Should APC NMC3 devices be reachable from user networks?
Generally no. Place them on a restricted management segment and allow only approved administration, monitoring, logging, and PowerChute paths with source controls and resilient recovery access.
Which protocols should an APC NMC3 use?
Use the secure services supported by the exact NMC firmware and integrations: normally trusted HTTPS, SSH where required, and SNMPv3 with authentication and privacy enabled. SNMPv3 alone does not guarantee encrypted traffic. Verify the matching manager profile and source restrictions, then disable unused legacy services only after their dependencies have migrated and been tested.
Is a self-signed NMC certificate sufficient?
Only when its identity has been independently verified and it is trusted through the approved client procedure; clicking through an unknown warning is insufficient. In PCNS 5.2, import only recognized, trusted NMC certificates. Its guide warns that the PowerChute-keystore keeps trusting imported certificates after expiry or revocation, so explicitly maintain and remove obsolete trust entries as part of rotation.
Supporting documentation: [3]
Can one PowerChute configuration be copied to every site?
No. PCNS behavior depends on physical UPS topology, outlet groups, platform, virtualization, storage, network dependencies, event thresholds, application timing, reserve, and recovery ownership.
What should determine PowerChute shutdown timing?
Use the actual UPS topology, measured load/runtime and observed shutdown sequence. Include event confirmation, commands and application drain, host/storage dependencies, timing variance and required reserve. A fictional 720-second window with 330 seconds of action allowances plus 180 seconds of reserve uses 510 seconds and leaves 210 seconds; those numbers are a worksheet, not settings to copy.
Does an NMC firmware upgrade turn off the attached UPS?
An NMC firmware update and a UPS firmware update are different operations. The management interface can restart separately from the load, but verify the exact card, UPS application, firmware method and current power state before an approved change. Do not apply an NMC-only expectation to a UPS firmware procedure or turn off production power merely to test an update.
How is NMC3 firmware version 3 or later updated?
Schneider Electric directs NMC3 version 3-and-later firmware through the Secure NMC System tool and applicable subscription. Confirm the exact hardware and application package, older update methods are not a universal substitute.
What belongs in an APC management recovery kit?
Include exact model, console or address recovery, trusted firmware source, current protected configuration, certificate and key custody, account recovery, PCNS reinstall and license records, topology, shutdown settings, and contacts.
How can ALLMSP harden APC NMC and PowerChute operations?
ALLMSP can inventory dependencies, segment management access, govern certificates and roles, reduce protocols, centralize events, model PCNS sequencing, stage firmware, test failover, and close retirement objects.
Sources
- Schneider Electric: Network Management Card 3 user guide ↩
- Schneider Electric: Network Management Card 3 security handbook ↩
- Schneider Electric: PowerChute Network Shutdown 5.2 standard guide ↩
- Schneider Electric: PowerChute Network Shutdown 5.2 installation guide ↩
- Schneider Electric: NMC3 firmware for Smart-UPS ↩
- Schneider Electric: How to upgrade APC Network Management Card firmware ↩
























































