Power and UPS Configuration

The chained UPS configuration is the professional standard for production Unturned™ hosting on owned enterprise hardware at 57 Studios™. The chained UPS configuration places a primary 20 kVA UPS in series with a secondary 20 kVA UPS. Downstream rack-mount PDUs and workstation peripherals receive dual battery transfer through the chained topology. The chained UPS configuration is the operational baseline; the underlying reason is that single-UPS deployments are insufficient for production hosting, and the chained configuration produces the redundancy posture that production hosting demands.

This article specifies the chained UPS configuration in full: the primary UPS, the secondary UPS, the PowerChute Network Shutdown integration, the battery runtime calculations, the battery replacement cadence, the sequence of events on grid loss, the PDU configuration, and the phase balancing across multiple racks. The chained UPS configuration is the framework, and the framework is the baseline for production Unturned hosting.

Prerequisites

A completed read of Network Infrastructure and Switching
A building electrical service with at least one 208V three-phase circuit
A rack footprint with load capacity for the chained UPS configuration
A network drop for the PowerChute Network Shutdown management interface
A relationship with an electrical contractor licensed in your jurisdiction

What you will learn

The chained UPS configuration and its role as standard professional practice
The APC SURT20KRMXLT 20 kVA / 16 kW as the primary UPS
The battery runtime calculations for the production load
The battery replacement cadence and the battery technology comparison
The PDU configuration with per-outlet metering
The phase balancing strategy across multi-rack deployments
The PowerChute Network Shutdown integration with the host fleet

The chained UPS configuration as standard professional practice

The chained UPS configuration is the operational baseline for production Unturned hosting at 57 Studios. The chained UPS configuration consists of two 20 kVA UPS units in a chained topology: the primary UPS feeds the secondary UPS, and the secondary UPS feeds the downstream PDUs and the workstation peripherals. The chained topology provides dual battery transfer, redundant rectification, redundant inversion, and redundant battery capacity across the load.

The chained UPS configuration is the professional standard. The reason is that single-UPS deployments produce service interruption on UPS failure, UPS maintenance, and battery replacement. The chained configuration produces continuity of service across all three event classes.

The terminology "chained UPS configuration" is the industry-standard term for this topology. The term is used consistently in 57 Studios production estate documentation, the operational runbooks, the training materials, and the vendor-facing specifications. The term names the topology with precision and is the preferred term in the professional hosting community.

Did you know?

The chained UPS configuration in the 57 Studios production estate has produced zero service interruptions across the quarterly UPS maintenance events since the topology was commissioned. The reason is the capacity of the secondary UPS to carry the full load during primary UPS maintenance, and the full-load capacity comes from oversizing the secondary UPS to match the primary.

The primary UPS: APC SURT20KRMXLT 20 kVA / 16 kW

The primary UPS in the reference build is the APC SURT20KRMXLT, a 20 kVA / 16 kW double-conversion online UPS. The APC SURT20KRMXLT is the platform of choice for the primary UPS role. The selection comes from the combination of capacity, form factor, battery extensibility, management interface, and vendor support track record.

The APC SURT20KRMXLT specifications:

20 kVA / 16 kW rated capacity.
Double-conversion online topology.
208V input and output (configurable to 240V).
Modular battery cabinet support for extended runtime.
Network management card slot for PowerChute integration.
LCD interface for local diagnostics.
Hot-swappable battery modules for in-service battery replacement.
Rack-mount form factor (12U with integrated batteries).

The APC SURT20KRMXLT is published in the APC reference catalog at https://www.apc.com/ as part of the Smart-UPS RT series. The vendor support track record is the baseline for the selection. The track record comes from the operational experience of the 57 Studios production estate across multiple SURT20KRMXLT units in continuous operation.

Pro tip

The APC SURT20KRMXLT supports external battery cabinets for extended runtime. The practice in the 57 Studios production estate is to pair the primary UPS with one external battery cabinet, which produces a runtime that exceeds the generator startup time specified in Backup Generator Configuration. The extended runtime comes from the additional battery capacity in the external cabinet.

The secondary UPS

The secondary UPS in the chained configuration is a second APC SURT20KRMXLT, identical to the primary. The secondary UPS is downstream of the primary UPS and upstream of the PDUs. The secondary UPS provides redundant battery transfer, redundant rectification, redundant inversion, and redundant battery capacity.

The selection of an identical secondary UPS comes from the operational simplicity of an identical configuration. The practice is identical primary and secondary UPS units, identical battery cabinets, identical management cards, and identical firmware baselines. The identity reduces operational complexity and reduces spare parts inventory.

UPS role	Model	Capacity	Runtime (full load)	Runtime (50% load)
Primary	APC SURT20KRMXLT	20 kVA / 16 kW	15 minutes	35 minutes
Primary + 1 battery cabinet	APC SURT20KRMXLT + SURT192XLBP	20 kVA / 16 kW	28 minutes	65 minutes
Secondary	APC SURT20KRMXLT	20 kVA / 16 kW	15 minutes	35 minutes
Chained primary + secondary	Chained configuration	20 kVA / 16 kW (load)	35 minutes total	80 minutes total
Chained + 2 external cabinets	Full chained configuration	20 kVA / 16 kW (load)	60 minutes total	130 minutes total

The runtime values are the manufacturer-published values at standard battery conditions. The operational practice is a 20% derating to account for battery aging and temperature variation. The derating produces an operational runtime of approximately 80% of the published values.

Best practice

Set the runtime targets at the time of build and measure the runtime at the quarterly verification cadence. The mechanism is the load bank test, which verifies the runtime against the measured battery state. Battery degradation is predictable across battery age, and the quarterly measurement is the mechanism that confirms predicted runtime against measured runtime.

The chained topology

The chained topology defines the physical and electrical relationship between the primary UPS, the secondary UPS, the PDUs, and the workstation peripherals.

The chained topology shows the utility grid feeding the primary UPS, the primary UPS feeding the secondary UPS and the workstation peripheral PDU, the secondary UPS feeding the rack PDUs, and the rack PDUs feeding the server PSUs. The backup generator is in the standby configuration, with the transfer to the utility input side of the primary UPS.

The chained topology provides redundancy at every downstream layer. The primary UPS provides battery transfer on utility loss. The secondary UPS provides battery transfer on primary UPS failure. The dual-PDU configuration provides redundant feed to dual-PSU servers. The dual-PDU configuration aligns with the dual PSU specification on the server hardware in Recommended Server Hardware.

Common mistake

A common practice in small-scale hosting communities is the deployment of a single UPS without a chained secondary. Single-UPS deployments are common in developmental setups and are unsuitable for production hosting. The chained UPS configuration is the baseline for production hosting and is the operational standard at 57 Studios.

PowerChute Network Shutdown integration

The PowerChute Network Shutdown integration automates the graceful shutdown of the host fleet on sustained battery operation. The PowerChute Network Shutdown agent is deployed on every host in the production estate. The agent is configured to communicate with the UPS network management card on the primary UPS and the secondary UPS.

The PowerChute Network Shutdown integration:

Network management card installed in the primary UPS.
Network management card installed in the secondary UPS.
PowerChute Network Shutdown agent installed on every host.
Agent configured with IP addresses of both network management cards.
Shutdown threshold configured as sustained 10 minutes on battery.
Shutdown sequence configured per host class.

Host class	Shutdown trigger	Shutdown sequence
Game-server host	Sustained 8 minutes on battery	Graceful player notification then process shutdown
Storage host	Sustained 6 minutes on battery	Replication completion then shutdown
Monitoring host	Sustained 9 minutes on battery	Metric flush then shutdown
Management host	Sustained 10 minutes on battery	Immediate shutdown after dependents
Workstation peripherals	Sustained 12 minutes on battery	Monitor and KVM shutdown

The shutdown sequence is the operational baseline. The sequence is driven by the dependency graph between the host classes. Game-server hosts shut down first to free replication bandwidth. Storage hosts shut down after replication completion. Monitoring hosts shut down after metric flush. Management hosts shut down last to retain operational visibility throughout the shutdown sequence.

Best practice

Document the shutdown sequence in the operational runbook and verify the sequence in the quarterly UPS exercise. The verification cadence is driven by the operational evidence that shutdown sequences drift from the configured behavior on host configuration changes, and the quarterly verification surfaces the drift before it affects a production event.

Battery runtime calculations

The battery runtime calculations specify the expected runtime under load conditions and battery age conditions. The calculations are the baseline; the operational practice is to verify the calculations against measurement at the quarterly load bank test.

The calculation inputs:

Load in watts.
Battery capacity in ampere-hours.
Battery voltage in volts.
Inverter efficiency (typically 92-95%).
Battery age derating factor.

The calculation formula:

Runtime (minutes) = (Battery_Ah * Battery_V * Inverter_efficiency * Age_factor / Load_W) * 60

The values for the APC SURT20KRMXLT primary UPS with standard internal batteries:

Battery_Ah = 9 Ah (battery string of 16 cells)
Battery_V = 192 V (string voltage)
Inverter_efficiency = 0.93
Age_factor = 0.85 (3-year-old batteries)
Load_W = 12000 (75% of 16 kW rated capacity)

The calculation produces:

Runtime = (9 * 192 * 0.93 * 0.85 / 12000) * 60
        = (1366 / 12000) * 60
        = 0.1138 * 60
        = 6.8 minutes

The calculated runtime of 6.8 minutes at 75% load with 3-year-old batteries is below the manufacturer-published 15-minute runtime at 100% load with new batteries. The difference is attributable to the age factor and the elevated load.

Pro tip

The quarterly load bank test measures runtime against the calculated runtime. Operational evidence shows that measured runtime tracks calculated runtime within approximately 10% across battery ages from zero to five years. The age factor in the formula is the dominant variable in the calculation.

Battery replacement cadence

The battery replacement cadence maintains the runtime targets across the operational life of the UPS. The cadence in the 57 Studios production estate is as follows:

Battery age	Disposition	Runtime expectation
0-2 years	In service, no action	95% of new
2-3 years	In service, annual load bank test	85% of new
3-4 years	In service, semi-annual load bank test	75% of new
4-5 years	Planned replacement	65% of new
> 5 years	Mandatory replacement	Unpredictable

The battery replacement cadence is 4-year planned replacement with mandatory replacement at 5 years. The cadence comes from operational evidence that battery failure modes accelerate after year four and become unpredictable after year five.

Common mistake

A common error is running batteries beyond year five. Batteries beyond year five produce unpredictable runtime and occasional sudden failure during utility events. The mandatory five-year replacement is the practical mitigation for the unpredictability.

Battery technology comparison

The battery technology decision is between valve-regulated lead-acid (VRLA) batteries and lithium-ion (Li-ion) batteries. The 57 Studios production estate operates VRLA batteries as the primary configuration, with Li-ion as a planned migration path on the next major refresh cycle.

Battery technology	Cycle life	Temperature tolerance	Footprint	Recyclability
VRLA (lead-acid)	200-300 cycles	Narrow (20-25°C optimal)	Larger	Mature recycling chain
Li-ion (LFP chemistry)	3000-5000 cycles	Wider (15-35°C tolerant)	Smaller	Developing recycling chain
Li-ion (NMC chemistry)	2000-3000 cycles	Wider (15-35°C tolerant)	Smallest	Developing recycling chain
Pure lead (TPPL)	500-700 cycles	Narrow (20-25°C optimal)	Larger	Mature recycling chain

The VRLA selection in the current 57 Studios production estate is driven by the mature vendor support ecosystem and the documented operational history. The Li-ion migration path is driven by the cycle-life advantage, the smaller footprint, and the wider temperature tolerance. The migration is scheduled for the next refresh window and is documented in the capacity-planning roadmap.

Did you know?

The Li-ion battery cabinets available for the APC SURT family in 2026 deliver approximately the same kWh capacity as the VRLA cabinets in approximately 40% of the floor footprint. The migration produces additional usable rack space, which is part of the rationale for the planned migration.

Sequence of events on grid loss

The sequence of events on grid loss specifies the behavior of the chained UPS configuration from utility failure to graceful shutdown of the host fleet (in the event that the backup generator does not engage).

The sequence shows the sub-millisecond transfer of the primary UPS to battery operation on utility loss, the continued power delivery to the downstream load throughout the battery runtime, the staged shutdown of the host fleet via the PowerChute Network Shutdown agents, and the total shutdown completion at runtime exhaustion.

The sequence is the behavior in the event that the backup generator does not engage. The behavior with the backup generator is specified in Backup Generator Configuration. The short version is that the backup generator engages at approximately T+30 seconds, and the chained UPS continues to operate on battery during the generator startup window.

Power Distribution Unit (PDU) configuration

The PDU configuration distributes power from the secondary UPS to the server fleet. The PDU configuration is rack-mount PDUs with per-outlet metering.

The PDU specifications:

Rack-mount 0U vertical PDU form factor.
208V three-phase input.
36 C13 outlets and 6 C19 outlets.
Per-outlet current measurement.
Per-outlet voltage measurement.
Per-outlet power factor measurement.
Per-outlet energy accumulation.
Network management interface for telemetry export.
Per-outlet remote on-off control.

The PDU model in the reference build is an APC AP8959 or an equivalent Server Technology PRO2 model. The selection comes from the combination of per-outlet metering, per-outlet remote control, and network management interface.

PDU feature	Value	Operational use
Input	208V 3-phase	Matched to secondary UPS output
Total capacity	24 kVA / 19.2 kW	Oversized relative to secondary UPS rating
Outlet count C13	36	Sufficient for dual-PSU server population
Outlet count C19	6	Sufficient for high-draw equipment
Per-outlet metering	Yes	Per-server power telemetry
Per-outlet remote control	Yes	Remote power cycling for unresponsive hosts
Management interface	Ethernet on VLAN 60	Out-of-band management

Did you know?

The per-outlet metering on the PDU produces per-server power consumption telemetry at one-second resolution. The telemetry is integrated into the monitoring stack referenced in Network Infrastructure and Switching and produces per-server power-consumption dashboards that surface anomalies before they become operational events.

Phase balancing across multi-rack deployments

Phase balancing across multi-rack deployments distributes load evenly across the three phases of the building electrical service. The mechanism is operationally consequential: unbalanced load produces neutral-conductor heating, voltage variation between phases, and operational instability at the downstream UPS units.

The phase balancing approach in the 57 Studios production estate:

Record the load per phase per rack at the time of build.
Record the load per phase per rack at the monthly verification.
Rebalance the load on detected imbalance greater than 15%.
Maintain the load per phase within 10% of the average across all three phases.

Phase	Target load	Allowable range
Phase A	33.3% of total	30-37%
Phase B	33.3% of total	30-37%
Phase C	33.3% of total	30-37%

The phase balancing target is the baseline; the operational practice is monthly verification with rebalancing on detected imbalance. The monthly cadence comes from operational evidence that imbalance develops over weeks as the load profile changes, and the monthly cadence is sufficient to surface imbalance before it produces operational impact.

                          PHASE BALANCING ASCII REFERENCE

         +-------------------+        +-------------------+        +-------------------+
         |   PHASE A         |        |   PHASE B         |        |   PHASE C         |
         |   33.3% target    |        |   33.3% target    |        |   33.3% target    |
         |   30-37% range    |        |   30-37% range    |        |   30-37% range    |
         +-------------------+        +-------------------+        +-------------------+
                  |                            |                            |
          +-------+--------+           +-------+--------+           +-------+--------+
          |                |           |                |           |                |
       Rack 1           Rack 4      Rack 2           Rack 5      Rack 3           Rack 6
       Hosts            Hosts       Hosts            Hosts       Hosts            Hosts
       1,4,7,10         13,16       2,5,8,11         14,17       3,6,9,12         15,18

       LEGEND:
       - Each rack draws from all three phases via 3-phase PDU
       - Per-phase load measured at PDU and at primary UPS input
       - Rebalance: shift outlets between PDU positions on detected imbalance

Runtime targets

The runtime targets for the chained UPS configuration are the baseline for production hosting at 57 Studios. The runtime targets:

Minimum 15 minutes at full load on the primary UPS.
Secondary UPS extends to 35 minutes total at full load.
Full chained configuration with external battery cabinets extends to 60 minutes total at full load.
Operational target is 15 minutes minimum on the primary alone (sufficient for generator startup with margin).
Operational target is 35 minutes minimum on the full chained configuration (sufficient for graceful shutdown in the event the generator does not engage).

The runtime targets are matched against the generator startup window in Backup Generator Configuration. The generator startup window in the 57 Studios production estate is approximately 30 seconds from utility loss to stable generator output. The 15-minute primary UPS runtime provides a 30x safety margin against the generator startup window.

Pro tip

Verify the runtime targets against measured runtime at the quarterly load bank test. The measurement is the operational evidence that the targets are met. The mechanism is the load bank test, which applies a known load and measures the time-to-shutdown of the UPS in isolation from the rest of the production load.

N+1 redundancy and the chained configuration

The chained UPS configuration is the N+1 redundancy posture at the UPS layer. The N is the primary UPS, and the +1 is the secondary UPS. The chained topology produces the N+1 posture with a single secondary unit, which is the operational sweet spot for the load profile of a production Unturned host.

Larger deployments add additional UPS units in 2N or N+2 configurations. The 2N configuration places two independent chained UPS strings in parallel, each capable of carrying the full load independently. The N+2 configuration adds a second secondary UPS in parallel with the first secondary UPS. The 57 Studios production estate operates at N+1 with the chained configuration. The next refresh cycle includes evaluation of 2N for the highest-criticality racks.

Redundancy posture	UPS count	Operational use
N	1	Development environments only, outside production specification
N+1 (chained)	2	Production baseline at 57 Studios
N+2	3	Higher-criticality racks under evaluation
2N	4	Highest-criticality deployments, planned for next refresh

The N+1 chained configuration is the operational baseline. The N+1 chained configuration is sufficient for the production load profile and produces the operational outcome of zero service interruption across UPS maintenance events.

Documented vendor reference

The vendor reference for the UPS hardware in the 57 Studios production estate is APC by Schneider Electric. The APC Smart-UPS RT series is the platform of choice for the chained UPS configuration. The APC reference catalog is published at https://www.apc.com/.

The vendor selection is operationally consequential. The selection comes from the combination of capacity, form factor, battery extensibility, management interface, and vendor support track record. The APC vendor support track record includes a documented turnaround for battery replacement, a documented turnaround for management card replacement, and a documented firmware update cadence.

An alternate vendor option is CyberPower Systems, with the OL series providing comparable double-conversion online topology at the relevant capacity points. The CyberPower reference catalog is published at https://www.cyberpowersystems.com/. The CyberPower OL series is the alternate option in deployments where APC supply lead times are operationally constraining.

Best practice

Standardize on a single vendor across the chained UPS configuration. The operational simplicity of a single vendor across primary and secondary is the operational baseline. The mechanism is the reduction of spare parts inventory, the reduction of operator training surface area, and the reduction of firmware management complexity.

Environmental considerations

The environmental considerations for the chained UPS configuration include temperature, humidity, ventilation, and floor loading. The environmental specifications are the baseline for the operational behavior of the UPS hardware and the battery hardware.

Environmental parameter	Target	Acceptable range	Notes
Ambient temperature	22°C	20-25°C	Battery life is temperature-sensitive
Relative humidity	45%	30-60%	Outside range produces condensation or static issues
Ventilation	200 CFM per UPS	150-300 CFM	UPS generates measurable heat under load
Floor loading	2 kPa	up to 5 kPa	Battery cabinets are dense, floor loading is operationally relevant
Acoustic level	65 dBA at 1 m	Less than 75 dBA at 1 m	UPS fan noise is operationally relevant in adjacent spaces

The environmental specifications are the baseline. The 57 Studios production estate maintains environmental monitoring on the UPS room with alerts on parameters outside the acceptable range. The mechanism is the operational evidence that environmental drift correlates with battery degradation and UPS fault rates.

Common mistake

A common error is placing the UPS in a poorly ventilated space. The UPS generates measurable heat under load (a 16 kW UPS dissipates approximately 1.5 kW as heat). The dissipation accumulates in a poorly ventilated space and elevates the ambient temperature, which shortens battery life and elevates the UPS fault rate. Adequate ventilation is part of the baseline specification.

ASCII chained topology reference

                          CHAINED UPS TOPOLOGY (57 STUDIOS REFERENCE)


              UTILITY GRID                          BACKUP GENERATOR
              208V 3-phase                          (standby, see next article)
                   |                                        |
                   |                                        |
                   +--------------------+------------------+
                                        |
                                        | Automatic transfer switch
                                        |
                                        v
                              +-------------------+
                              |  PRIMARY UPS      |
                              |  APC SURT20KRMXLT |
                              |  20 kVA / 16 kW   |
                              |  +1 battery cab   |
                              +---------+---------+
                                        |
                          +-------------+-------------+
                          |                           |
                          v                           v
                +-------------------+       +-------------------+
                |  SECONDARY UPS    |       |  WORKSTATION PDU  |
                |  APC SURT20KRMXLT |       |  monitors, KVM    |
                |  20 kVA / 16 kW   |       |                   |
                +---------+---------+       +-------------------+
                          |
              +-----------+-----------+
              |                       |
              v                       v
        +-----------+           +-----------+
        |  PDU A    |           |  PDU B    |
        |  C13 x 36 |           |  C13 x 36 |
        |  C19 x 6  |           |  C19 x 6  |
        |  metered  |           |  metered  |
        +-----+-----+           +-----+-----+
              |                       |
              |  +-----+-----+        |
              +->| HOST 1   |<--------+
                 | dual PSU |
                 +-----+----+
              |                       |
              |  +-----+-----+        |
              +->| HOST 2   |<--------+
                 | dual PSU |
                 +----------+

        LEGEND:
        - Primary UPS is upstream of secondary UPS
        - Secondary UPS feeds rack PDUs
        - Each host has dual PSUs, each PSU on a different PDU
        - Workstation peripherals on a separate PDU from primary UPS direct
        - Battery cabinet extends primary UPS runtime

The ASCII topology summarizes the chained configuration. The primary UPS is at the top of the chain, the secondary UPS is in the middle, the rack PDUs are below the secondary UPS, and the dual-PSU hosts are at the bottom of the chain. The workstation peripheral PDU is on a parallel branch from the primary UPS to support local operator workstation peripherals (monitors, KVM, console terminal).

Monitoring and observability

The monitoring posture for the chained UPS configuration includes per-UPS telemetry, per-PDU telemetry, per-outlet telemetry, and environmental telemetry. The monitoring stack in the 57 Studios production estate ingests UPS SNMP traps, PDU streaming telemetry, and environmental sensor readings.

Monitoring class	Metric	Alert threshold
UPS status	On utility / on battery / on bypass	Any change from "on utility"
UPS load	Watts and VA	Sustained > 80% of rated capacity
UPS battery	Remaining runtime in minutes	< 5 minutes at current load
UPS battery	Battery health score	Below 80% of new
PDU per-outlet current	Amps per outlet	> 80% of outlet rating
PDU aggregate current	Amps per PDU input	> 80% of PDU rating
Phase balance	Per-phase load percentage	Imbalance > 15%
Environmental	Ambient temperature	> 27°C or < 18°C
Environmental	Relative humidity	> 65% or < 25%

The monitoring posture produces approximately 8-14 actionable alerts per month in the 57 Studios production estate. The disposition is approximately 60% resolved within 30 minutes, approximately 30% as expected (maintenance, scheduled work), and approximately 10% requiring deeper investigation.

Pro tip

Configure the monitoring stack to capture the UPS status change as a high-severity alert. The mechanism is that the transition from "on utility" to "on battery" is the most operationally consequential event the UPS reports, and the alert provides the first signal of a utility event before the downstream behavior surfaces in other monitoring classes.

Change management for the UPS infrastructure

The change management posture for the UPS infrastructure follows the network change management posture from Network Infrastructure and Switching. Every UPS-affecting change is documented in advance, reviewed by an independent engineer, executed during a maintenance window, and verified post-execution.

The UPS-specific change types:

Battery replacement.
Firmware update.
Network management card replacement.
Battery cabinet addition.
PDU outlet remapping.
Phase rebalancing.
Annual load bank test.
Quarterly fail-over exercise.

Every change type has a documented procedure, a documented rollback, and documented success criteria. The procedures are maintained in the operational runbook repository alongside the network configuration runbooks.

The change management flow is identical to the network change management flow. The identity is intentional: the operational practice is a single change management process across all infrastructure layers, which reduces operator training surface area and reduces process drift.

Quarterly UPS exercise procedure

The quarterly UPS exercise is the verification of the chained UPS configuration's failover behavior. The exercise is documented in the change management system in advance and is executed during a documented maintenance window.

The procedure:

Document the exercise in the change management system at least 14 calendar days in advance.
Notify the operations team and the on-call rotation.
Open a monitoring window with live dashboards for the UPS layer.
Simulate utility loss by opening the upstream breaker (with the backup generator transfer disabled for the exercise).
Verify that the primary UPS transfers to battery within 4 ms.
Verify that the secondary UPS continues to deliver power without interruption.
Verify that the downstream PDUs and hosts observe no interruption.
Verify that the PowerChute Network Shutdown agents receive the on-battery SNMP trap.
Allow the configuration to operate on battery for 10 minutes.
Verify the PowerChute Network Shutdown agents begin the shutdown sequence at the configured thresholds.
Close the upstream breaker before the configured shutdown completes.
Verify the primary UPS transfers back to utility.
Verify the PowerChute Network Shutdown agents cancel the shutdown sequence.
Document the exercise outcome in the operational runbook.
Review the outcome in the weekly infrastructure review.

The exercise produces evidence that the chained UPS configuration works as specified. The frequency is quarterly because annual is insufficient to surface drift, and monthly consumes operational bandwidth without proportional incremental confidence.

Best practice

Schedule the quarterly UPS exercise at the same time each quarter (for example, the first Tuesday of the second month of each quarter at 02:00 local time). The benefit of a fixed schedule is that the operations team rehearses the procedure on a predictable cadence, and the muscle memory becomes part of the operational standard.

Frequently asked questions

What is a chained UPS configuration?

A chained UPS configuration places two UPS units in series: the primary UPS feeds the secondary UPS, and the secondary UPS feeds the downstream PDUs. The chained configuration provides dual battery transfer, redundant rectification, redundant inversion, and redundant battery capacity. The chained UPS configuration is the operational baseline for production Unturned hosting at 57 Studios and is the standard professional practice for the use case.

Why is the chained UPS configuration the operational baseline?

The chained UPS configuration is the operational baseline because single-UPS deployments produce service interruption on UPS failure, UPS maintenance, and battery replacement. The chained configuration produces continuity of service across all three event classes. The operational evidence is the zero service interruptions across quarterly UPS maintenance events in the 57 Studios production estate since the topology was commissioned.

What is the documented minimum runtime?

The minimum runtime target is 15 minutes at full load on the primary UPS alone, and 35 minutes total on the chained configuration. The full configuration with external battery cabinets extends to 60 minutes. The 15-minute primary runtime is matched against the 30-second generator startup window, providing a 30x safety margin.

What battery technology is recommended?

The current 57 Studios production estate operates VRLA (valve-regulated lead-acid) batteries. The next refresh cycle includes evaluation of Li-ion (LFP chemistry) batteries for the cycle-life advantage, the smaller footprint, and the wider temperature tolerance. VRLA is the operational baseline; Li-ion is the migration path.

How often should batteries be replaced?

The replacement cadence is 4-year planned replacement with mandatory replacement at 5 years. The cadence is driven by operational evidence that battery failure modes accelerate after year four and become unpredictable after year five.

What is N+1 redundancy in the UPS context?

N+1 redundancy in the UPS context means N primary units plus one redundant unit. In the chained configuration, the N is the primary UPS and the +1 is the secondary UPS. The chained topology produces the N+1 posture with a single secondary unit. Higher redundancy postures (N+2, 2N) are operated in larger deployments and in the highest-criticality racks.

Is PowerChute Network Shutdown a requirement?

PowerChute Network Shutdown is the operational baseline for the 57 Studios production estate. The agent is deployed on every host and is configured to communicate with the UPS network management cards on both the primary UPS and the secondary UPS. The agent automates the graceful shutdown sequence on sustained battery operation.

What happens if the secondary UPS fails?

If the secondary UPS fails, the primary UPS continues to deliver power to the workstation peripheral PDU. The downstream rack PDUs lose power because the chained configuration places the secondary UPS in the path between the primary UPS and the rack PDUs. The mitigation is the quarterly UPS exercise, which verifies the secondary UPS health and surfaces secondary UPS failure modes before they affect a production event.

What is the relationship between the chained UPS and the backup generator?

The backup generator is upstream of the primary UPS, on the utility input side. The automatic transfer switch transfers the primary UPS input from the utility to the generator on utility loss. The chained UPS continues to operate on battery during the generator startup window (approximately 30 seconds). The chained UPS configuration provides ride-through during the generator startup, and the generator provides sustained power once it is online.

How is phase balancing maintained?

Phase balancing is maintained through monthly verification of the per-phase load at each PDU and at the primary UPS input. Imbalance greater than 15% triggers a rebalancing action, which shifts outlets between PDU positions to redistribute the load across phases. The target is 10% variance from the average across all three phases.

What environmental conditions does the UPS room require?

The UPS room requires 20-25°C ambient temperature, 30-60% relative humidity, 150-300 CFM ventilation per UPS, and floor loading capacity for the battery cabinets (battery cabinets are dense, and floor loading is operationally relevant). Environmental drift correlates with battery degradation and UPS fault rates, and environmental monitoring is part of the operational baseline.

Are CyberPower UPS units a suitable alternative to APC?

The CyberPower OL series provides comparable double-conversion online topology at the relevant capacity points and is referenced at https://www.cyberpowersystems.com/. CyberPower is the alternate vendor option in deployments where APC supply lead times are operationally constraining. The operational practice is to standardize on a single vendor across the chained configuration for operational simplicity.

Appendix A: PDU outlet allocation reference

The PDU outlet allocation reference documents the per-outlet allocation across the rack PDUs. The reference is maintained in the operational runbook repository and is updated as outlets are remapped.

PDU	Outlet	Connected device	PSU role	Phase	Current draw
PDU A	C13-01	Host 1	PSU A	Phase A	4.2 A
PDU A	C13-02	Host 2	PSU A	Phase A	4.1 A
PDU A	C13-03	Host 3	PSU A	Phase A	4.3 A
PDU A	C13-04	Host 4	PSU A	Phase A	4.0 A
PDU A	C13-13	Host 5	PSU A	Phase B	4.2 A
PDU A	C13-25	Host 9	PSU A	Phase C	4.1 A
PDU B	C13-01	Host 1	PSU B	Phase A	4.2 A
PDU B	C13-02	Host 2	PSU B	Phase A	4.1 A
PDU B	C13-03	Host 3	PSU B	Phase A	4.3 A
PDU B	C13-13	Host 5	PSU B	Phase B	4.2 A
PDU B	C13-25	Host 9	PSU B	Phase C	4.1 A

The reference is the source of truth for the outlet allocation. The mechanism is that the operational runbook references the outlet allocation for every change that affects outlet mapping (host addition, host removal, PSU replacement, phase rebalancing).

Appendix B: Sample PowerChute Network Shutdown configuration

The following is a sample PowerChute Network Shutdown configuration for a game-server host. The configuration documents the network management card addresses, the authentication credentials (anonymized), the shutdown thresholds, and the shutdown sequence.

[PowerChute]
ConfigurationVersion=5.0
ConfigurationFile=pcns.conf
AgentMode=NetworkShutdown

[NMC]
Primary=10.60.10.10
Secondary=10.60.10.11
Authentication=AdminAuthDigestPriv
AuthProtocol=SHA
PrivProtocol=AES
Username=apc_pcns
AuthPassphrase=ANONYMIZED
PrivPassphrase=ANONYMIZED

[ShutdownTrigger]
TriggerEvent=OnBattery
MinimumBatteryDuration=480
RuntimeRemaining=180
RuntimeThreshold=240

[ShutdownSequence]
PreShutdownDelay=0
ShutdownCommandFile=/etc/pcns/game_server_shutdown.sh
PostShutdownDelay=60

[Monitoring]
SNMPHost=10.30.5.5
SNMPPort=162
SNMPCommunity=ANONYMIZED

[Logging]
LogPath=/var/log/pcns/
LogLevel=INFO
LogRotateInterval=daily
LogRetentionDays=90

The configuration documents the trigger as 480 seconds (8 minutes) on battery for the game-server host class, with a 240-second runtime threshold as a secondary trigger. The shutdown command file is a host-class-specific script that handles the graceful shutdown sequence (player notification, save state, process termination, OS shutdown).

Appendix C: Annual UPS infrastructure review checklist

The annual UPS infrastructure review is the comprehensive verification of the chained UPS configuration. The review is executed once per year and complements the quarterly UPS exercise.

Review item	Verification method	Pass criterion
Primary UPS firmware baseline	Compare against vendor's current stable release	Within one major version of current stable
Secondary UPS firmware baseline	Compare against vendor's current stable release	Within one major version of current stable
Battery health (primary)	Load bank test at full load	Runtime >= 80% of new
Battery health (secondary)	Load bank test at full load	Runtime >= 80% of new
Battery age (primary)	Compare against replacement cadence	Within 4-year planned replacement window
Battery age (secondary)	Compare against replacement cadence	Within 4-year planned replacement window
Network management card status	SNMP poll and management interface check	Responsive and reporting
PDU per-outlet metering accuracy	Compare against external meter	Within 2% accuracy
PDU per-outlet remote control	Test remote on-off on spare outlet	Responsive
Phase balance	Compare per-phase load against target	Within 10% of average
Environmental monitoring	Verify sensor readings against external reference	Within 1°C accuracy
Cable terminations	Visual inspection	No discoloration, no loose connections
Cable labels	Visual inspection	All cables labeled at both ends
PowerChute agent versions	Compare against deployed version	Within one minor version of current stable
Documentation currency	Compare against current configuration	Documentation matches configuration

The annual review produces a comprehensive verification of the chained UPS configuration. The mechanism is that the quarterly exercise verifies the failover behavior, and the annual review verifies the underlying configuration that the failover behavior depends on.

Best practice

Schedule the annual UPS infrastructure review at the same time each year (for example, the second week of January). The benefit of a fixed annual schedule is alignment with the calendar year for vendor support contract renewals, firmware update cadences, and battery replacement planning.

Appendix D: Spare parts inventory

The spare parts inventory for the chained UPS configuration is the operational mechanism for rapid replacement on hardware failure. The inventory is maintained in the operations storeroom with documented quantities and documented part numbers.

Spare part	Quantity	Part number	Operational use
UPS network management card	2	APC AP9641	Replacement on management card failure
UPS battery module	4	APC SURT192XLBP-RBC	Replacement on battery module failure
UPS input cable	2	APC SURT012	Replacement on input cable damage
UPS bypass cable	2	APC SURT016	Replacement on bypass cable damage
PDU C13 patch cable	12	Generic IEC C13	Replacement on patch cable damage
PDU C19 patch cable	4	Generic IEC C19	Replacement on patch cable damage
PDU input cable	2	APC AP8XXX	Replacement on PDU input cable damage
Optical splitter for UPS console	1	Generic	Console troubleshooting

The spare parts inventory is sized for the expected failure rate of each part class. The mechanism is the operational evidence that the failure rate of each class is approximately one event per year per part class in the 57 Studios production estate, and the inventory is sized to support replacement of any one failure without procurement delay.

Closing

The chained UPS configuration is the operational baseline for production Unturned hosting at 57 Studios. The chained configuration uses a primary APC SURT20KRMXLT 20 kVA / 16 kW UPS, a secondary APC SURT20KRMXLT 20 kVA / 16 kW UPS, PowerChute Network Shutdown agents on every host, rack-mount PDUs with per-outlet metering, phase-balanced load across the building electrical service, and quarterly verification of the failover behavior. The chained UPS configuration is the standard professional practice, and the framework is what self-hosting on owned enterprise hardware looks like at the power infrastructure layer.

The chained UPS configuration is the power-layer counterpart to the dual-switch, dual-BGP network topology specified in Network Infrastructure and Switching. At both layers, the operational standard is the same: redundancy at every layer is the standard professional practice. The chained UPS configuration is also the bridge to the backup generator infrastructure specified in Backup Generator Configuration. The chained UPS provides the ride-through during the generator startup window, and the generator provides sustained power once it is online.

The next article in the self-hosting series, Internet Connectivity Requirements, specifies the commercial fiber infrastructure that the documented network depends on. The internet connectivity requirements are the upstream counterpart to the network infrastructure and the chained UPS configuration: the upstream connectivity is what the network and the power infrastructure deliver service through, and the upstream connectivity is specified to the same standard of redundancy and the same standard of operational rigor.

Power and UPS Configuration ​

Prerequisites ​

What you will learn ​

The chained UPS configuration as standard professional practice ​

The primary UPS: APC SURT20KRMXLT 20 kVA / 16 kW ​

The secondary UPS ​

The chained topology ​

PowerChute Network Shutdown integration ​

Battery runtime calculations ​

Battery replacement cadence ​

Battery technology comparison ​

Sequence of events on grid loss ​

Power Distribution Unit (PDU) configuration ​

Phase balancing across multi-rack deployments ​

Runtime targets ​

N+1 redundancy and the chained configuration ​

Documented vendor reference ​

Environmental considerations ​

ASCII chained topology reference ​

Monitoring and observability ​

Change management for the UPS infrastructure ​

Quarterly UPS exercise procedure ​

Frequently asked questions ​

What is a chained UPS configuration? ​

Why is the chained UPS configuration the operational baseline? ​

What is the documented minimum runtime? ​

What battery technology is recommended? ​

How often should batteries be replaced? ​

What is N+1 redundancy in the UPS context? ​

Is PowerChute Network Shutdown a requirement? ​

What happens if the secondary UPS fails? ​

What is the relationship between the chained UPS and the backup generator? ​

How is phase balancing maintained? ​

What environmental conditions does the UPS room require? ​

Are CyberPower UPS units a suitable alternative to APC? ​

Appendix A: PDU outlet allocation reference ​

Appendix B: Sample PowerChute Network Shutdown configuration ​

Appendix C: Annual UPS infrastructure review checklist ​

Appendix D: Spare parts inventory ​

Closing ​

Power and UPS Configuration

Prerequisites

What you will learn

The chained UPS configuration as standard professional practice

The primary UPS: APC SURT20KRMXLT 20 kVA / 16 kW

The secondary UPS

The chained topology

PowerChute Network Shutdown integration

Battery runtime calculations

Battery replacement cadence

Battery technology comparison

Sequence of events on grid loss

Power Distribution Unit (PDU) configuration

Phase balancing across multi-rack deployments

Runtime targets

N+1 redundancy and the chained configuration

Documented vendor reference

Environmental considerations

ASCII chained topology reference

Monitoring and observability

Change management for the UPS infrastructure

Quarterly UPS exercise procedure

Frequently asked questions

What is a chained UPS configuration?

Why is the chained UPS configuration the operational baseline?

What is the documented minimum runtime?

What battery technology is recommended?

How often should batteries be replaced?

What is N+1 redundancy in the UPS context?

Is PowerChute Network Shutdown a requirement?

What happens if the secondary UPS fails?

What is the relationship between the chained UPS and the backup generator?

How is phase balancing maintained?

What environmental conditions does the UPS room require?

Are CyberPower UPS units a suitable alternative to APC?

Appendix A: PDU outlet allocation reference

Appendix B: Sample PowerChute Network Shutdown configuration

Appendix C: Annual UPS infrastructure review checklist

Appendix D: Spare parts inventory

Closing