LACP Bundle Capacity Calculator

Member links:

Count only same-speed member links that can forward traffic when healthy.

links

Member speed:

Use the physical line rate before hash efficiency and reserve adjustments.

Gbps

Failed members:

Model the current outage or the failure case you want the bundle to survive.

links

Planned demand:

Enter the peak aggregate demand you expect the bundle to carry.

Gbps

Hash efficiency:

Lower this when traffic has few flows or poor source/destination entropy.

Overhead and reserve:

Use this to keep capacity planning below raw line-rate math.

Largest single flow:

A single flow usually cannot consume more than one member link, even when the aggregate has spare capacity.

Gbps

Check LACP inputs

{{ error }}

Metric	Value	Details	Copy
{{ row.label }}	{{ row.value }}	{{ row.details }}

Failed links	Active links	Usable capacity	Headroom	State	Copy
{{ row.failedDisplay }}	{{ row.activeDisplay }}	{{ row.capacityDisplay }}	{{ row.headroomDisplay }}	{{ row.state }}

Check	Finding	Evidence	Action	Copy
{{ row.check }}	{{ row.finding }}	{{ row.evidence }}	{{ row.action }}

Embed:

Customize

Include current inputs

Size

Advanced

Width

Height

Aspect ratio

Max height

Collapsible embed

Allow fullscreen

Referrer policy

Sandbox tokens

A link aggregation group can make several Ethernet links behave like one logical connection, but it does not turn them into one perfect shared pipe. Link Aggregation Control Protocol (LACP) negotiates and monitors member links in many Ethernet aggregates. Capacity planning still depends on how many members are forwarding, how traffic is assigned to those members, and how much margin should be held back for overhead and bursts.

The difference matters in switch uplinks, virtualization clusters, storage networks, firewall pairs, campus trunks, and data-center leaf-spine designs. A four-member 10 Gbps bundle may be described as 40 Gbps, yet one failed member leaves 30 Gbps of raw active line rate before any reserve or uneven distribution is considered. If peak demand is 28 Gbps, the diagram looks comfortable while the operating margin may already be thin.

LACP bundle capacity terms and planning meaning
Term	Planning meaning
Member link	One physical Ethernet link that can carry traffic as part of the bundle.
Active capacity	Member speed multiplied by the links still forwarding after the modeled failures.
Hash efficiency	A planning discount for traffic that does not spread evenly across active members.
Single-flow ceiling	The practical limit for one large conversation that hashes to one member instead of the whole bundle.

Flow hashing is the usual reason aggregate capacity and single-transfer speed diverge. Keeping a conversation on one member preserves packet order, but it also means one backup job, replication stream, or storage session usually cannot consume every member link at once. More links help most when the traffic mix has many independent conversations with enough address or port diversity for the hash policy to spread them.

A useful LACP capacity estimate separates gross line rate, active line rate after failed members, reserved capacity, hash efficiency, demand headroom, and the largest single-flow risk. That separation prevents a design from looking healthy only because one weak assumption was hidden inside a single utilization number.

How to Use This Tool:

Enter the physical bundle first, then model the failure and traffic assumptions that the design must survive.

Set Member links to the number of same-speed physical links in the LACP bundle. Set Member speed to each forwarding member's line rate in Gbps.
Set Failed members to the outage or maintenance case you want to model. The active-link count and failure ladder update from that value.
Enter Planned demand as the peak aggregate load in Gbps. Use a busy-hour, backup-window, or failover-window estimate rather than a daily average.
Adjust Hash efficiency for traffic diversity. Keep it high for many well-distributed flows and lower it for repeated host pairs, tunnels, storage sessions, or backup jobs that may land unevenly.
Use Overhead and reserve for protocol overhead, burst allowance, QoS headroom, and operating margin. This percentage is removed before the hash-efficiency discount.
Enter Largest single flow when one conversation may dominate a member link. Bundle Snapshot compares it with the per-member payload ceiling.
If Check LACP inputs appears, fix the named range before trusting the result. Common fixes are keeping failed members no higher than member links, hash efficiency from 1% to 100%, and overhead and reserve from 0% to 80%.

Interpreting Results:

Start with Usable planning capacity and Planned demand. Positive headroom means the modeled demand fits after failed members, reserve, and hash efficiency. Negative headroom means the bundle is short by the displayed Gbps amount.

The status label is a capacity cue, not a measurement from a switch. Capacity short appears when demand exceeds usable capacity. Near ceiling starts at 90% utilization, and Watch demand starts at 70%. Capacity ok still needs review when Single-flow pinning warns that the largest flow exceeds the per-member payload ceiling.

Member Failure Ladder and Failure Capacity Curve help with maintenance and N+ planning. The ladder shows each failed-link count from zero through the full bundle count and marks the first step where demand no longer fits. The curve makes the drop in usable capacity visible as active members disappear.

Before committing a design or maintenance plan, compare the estimate with port-channel member counters, per-member utilization, real flow mix, and the configured hash policy on both ends of the bundle.

Technical Details:

IEEE link aggregation lets multiple full-duplex point-to-point Ethernet links act as one logical link for the MAC client. LACP adds negotiation and monitoring, but forwarding is still performed over physical member links. Most platforms use a distribution algorithm based on packet fields, so a conversation maps to a member instead of being byte-striped across every link.

The model uses same-speed members because one member speed is entered for the whole bundle. Mixed-rate aggregates, chassis-specific weighting, minimum-link rules, standby members, and vendor-specific hashing behavior need platform documentation or live counters. The calculator's planning value is intentionally conservative because it discounts active capacity for reserve and imperfect flow spread.

Formula Core:

The governing equation subtracts failed members, applies the reserve percentage, then applies hash efficiency. Headroom is the remaining usable capacity after planned demand.

\begin{array}{lll} N_{active} & = & \max (0, N_{total} - N_{failed}) \\ payload factor & = & 1 - \frac{overhead and reserve percent}{100} \\ hash factor & = & \frac{hash efficiency percent}{100} \\ usable capacity & = & N_{active} \times member speed \times payload factor \times hash factor \\ headroom & = & usable capacity - planned demand \end{array}

With four 10 Gbps members and one failed member, three active members remain. A 3% reserve gives a payload factor of 0.97, and 78% hash efficiency gives a hash factor of 0.78. The usable planning capacity is 3 x 10 x 0.97 x 0.78 = 22.698 Gbps, so 18 Gbps of planned demand leaves about +4.70 Gbps of headroom.

LACP bundle input ranges and meanings
Input	Meaning	Accepted range
`Member links`	Total physical links configured in the bundle.	1 to 64 links
`Member speed`	Line rate for each same-speed member.	0.1 to 800 Gbps
`Failed members`	Members removed from forwarding in the modeled case.	0 to member count
`Planned demand`	Peak aggregate demand compared with usable capacity.	0 Gbps or higher
`Hash efficiency`	Discount for uneven flow distribution across active members.	1% to 100%
`Overhead and reserve`	Capacity held back before the hash factor is applied.	0% to 80%
`Largest single flow`	Largest conversation that may hash to one member.	0 Gbps or higher

The single-flow check uses a different ceiling from the aggregate capacity. The per-member payload ceiling is Member speed x payload factor. A 10 Gbps member with 3% reserve has a 9.70 Gbps per-member payload ceiling, so a 9.8 Gbps flow is flagged even if aggregate headroom remains positive.

LACP bundle status thresholds
Status	Rule	Interpretation
`No active members`	Active members equal `0` and demand is above `0 Gbps`.	No forwarding capacity remains in the modeled case.
`Capacity short`	Headroom is below `0 Gbps`.	Demand is larger than usable planning capacity.
`Near ceiling`	Demand utilization is at least `90%`.	The demand fits, but margin is thin.
`Watch demand`	Demand utilization is at least `70%` and below `90%`.	The demand fits with visible pressure.
`Capacity ok`	Headroom is non-negative and utilization is below `70%`.	The modeled demand fits with wider margin.

Failure tolerance is found by recalculating usable capacity for each failed-member count from zero through the total member count. A row remains safe when usable capacity is greater than or equal to planned demand. Additional failures safe is the distance between the current failed-member count and the last failed-member count that still fits demand.

Limitations:

The result is a planning estimate based on the values entered in the browser. It does not inspect live switch counters, packet captures, interface configuration, LACP state, minimum-link settings, or vendor-specific hash buckets.

Mixed-speed aggregates need platform-specific weighting. This model uses one member speed for every link.
Actual traffic spread can change with source and destination diversity, tunnel headers, VLANs, transport ports, hash policy, and member failures.
Some platforms support standby or backup member behavior that is not the same as all members actively forwarding.
Reserve is a planning allowance. Align it with real overhead, burst tolerance, monitoring confidence, and operational policy.

Worked Examples:

Routine uplink review. Use 4 member links, 10 Gbps member speed, 1 failed member, 18 Gbps planned demand, 78% hash efficiency, 3% overhead and reserve, and a 6 Gbps largest single flow. The active raw capacity is 30 Gbps, usable planning capacity is about 22.70 Gbps, and headroom is about +4.70 Gbps.

Near-ceiling demand. Keep the same bundle and reserve settings, but raise planned demand to 21 Gbps. Usable capacity remains about 22.70 Gbps, utilization crosses 90%, and the status becomes Near ceiling. That should trigger a member-utilization and hash-distribution check before a drain or failover decision.

Two failed members. With two failed links in the four-member bundle, usable capacity drops to about 15.13 Gbps under the same reserve and hash assumptions. An 18 Gbps demand becomes Capacity short, so the action is to restore a member, reduce demand, increase member speed, add members, or improve traffic spread.

Single-flow edge. Keep aggregate demand at 18 Gbps and raise the largest single flow to 9.8 Gbps with a 3% reserve. The per-member payload ceiling is 9.70 Gbps, so aggregate capacity can pass while single-flow pinning still warns.

FAQ:

Why is usable capacity lower than member count times speed?

Failed members are removed first. The remaining active capacity is then reduced by overhead and reserve, and reduced again by the hash-efficiency assumption.

Can one large transfer use the whole bundle?

Usually no. Flow hashing commonly keeps one conversation on one member link, so the largest-flow check compares that conversation with the per-member payload ceiling.

What should I use for hash efficiency?

Use a higher value when many flows are likely to spread across members. Lower it when traffic is dominated by a few host pairs, tunnels, storage sessions, backups, or a hash policy that ignores useful fields.

How do I model an N+1 maintenance case?

Set failed members to the number of links you expect to lose, then read the failure tolerance and scan the member failure ladder for the first short row.

Why did the input warning appear?

The warning names a range problem, such as failed members above member links, hash efficiency outside 1% to 100%, overhead and reserve outside 0% to 80%, or a negative demand or flow value.

Glossary:

LACP: Link Aggregation Control Protocol, a protocol used to negotiate and monitor Ethernet link aggregation.
LAG: Link aggregation group, the logical link formed from multiple physical member links.
Member link: One physical Ethernet link participating in the aggregate.
Hash efficiency: The planning percentage used to discount active capacity for uneven flow spread.
Headroom: Usable planning capacity minus planned demand.
Single-flow pinning: The condition where one large conversation is limited by one member link instead of the aggregate.

References:

802.1AX Link Aggregation Revision, IEEE 802.1 Working Group.
Understand EtherChannel Load Balance and Redundancy on Catalyst Switches, Cisco.
Load Balancing on Aggregated Ethernet Interfaces, Juniper Networks.
Aggregated Ethernet Interfaces Overview, Juniper Networks.

LACP Bundle Capacity Calculator

Introduction:

How to Use This Tool:

Interpreting Results:

Technical Details:

Formula Core:

Limitations:

Worked Examples:

FAQ:

Glossary:

References: