# Infor OS ION API and API Flow Payload and Timeout Calculator

> Source: https://netray.co/tools/infor-os-ion-api-flow-payload-and-timeout-calculator
> Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems.
> Section: Free Tools

## Short answer

Find a records-per-call batch size that stays inside ION API Gateway time-outs and API Flow payload limits, and estimate total run time.

## Category

developer

## Erp

Infor ION

## Audience

- Integration developer
- Infor OS architect
- ERP developer

## Tool

**Kind:** calculator

**Inputs:**

#### Total records to move

**Key:** total

**Type:** number

**Default:** 50000

**Min:** 1

**Step:** 1000

**Unit:** records

#### Size of one record in JSON

**Key:** bytes

**Type:** number

**Default:** 1500

**Min:** 1

**Step:** 100

**Unit:** bytes

**Help:** Measure a real response and divide by its record count

#### Records per call

**Key:** rpc

**Type:** slider

**Default:** 500

**Min:** 10

**Max:** 5000

**Step:** 10

**Unit:** records

#### Payload limit per call

**Key:** limit_mb

**Type:** number

**Default:** 3

**Min:** 0.5

**Step:** 0.5

**Unit:** MB

**Help:** Infor pages document 3 or 5 MB per API Flow task and 5 or 10 MB per flow; enter the value that applies to you

#### Gateway target time-out

**Key:** timeout

**Type:** number

**Default:** 60

**Min:** 1

**Step:** 5

**Unit:** seconds

**Help:** Target timeout is configured between 60 and 300 seconds; API Flows are capped by the 5-minute global gateway time-out

#### Measured server time per record

**Key:** ms

**Type:** number

**Default:** 20

**Min:** 0.1

**Step:** 1

**Unit:** ms

#### Fixed overhead per call

**Key:** overhead

**Type:** number

**Default:** 400

**Min:** 0

**Step:** 50

**Unit:** ms

**Help:** Token check, network round trip, connection setup

#### Your allowed calls per minute

**Key:** cpm

**Type:** number

**Default:** 100

**Min:** 1

**Step:** 10

**Unit:** calls/min

**Help:** Your own throttle or agreed limit

**Outputs:**

#### Payload per call (MB)

**Key:** payload_mb

**Expr:** rpc * bytes / 1048576

**Format:** decimal

#### Max records per call by payload limit

**Key:** max_by_size

**Expr:** floor(limit_mb * 1048576 / bytes)

**Format:** integer

#### Call duration (seconds)

**Key:** duration

**Expr:** (overhead + rpc * ms) / 1000

**Format:** decimal

#### Call fits time-out and payload limit (1 = yes, 0 = no)

**Key:** fits

**Expr:** if(and(lte(duration, timeout), lte(payload_mb, limit_mb)), 1, 0)

**Format:** integer

#### Calls needed

**Key:** calls

**Expr:** ceil(total / rpc)

**Format:** integer

#### Total run time

**Key:** total_min

**Expr:** max(calls * duration / 60, calls / cpm)

**Format:** minutes

**Primary:** true

**Description:** The larger of sequential call time and the time your calls-per-minute limit allows

#### Safe records per call (30% headroom)

**Key:** safe

**Expr:** max(0, min(max_by_size, floor((timeout * 0.7 * 1000 - overhead) / ms)))

**Format:** integer

**Primary:** true

**Description:** Smallest of the payload cap and the count that finishes within 70% of the time-out

**Chart:**

#### Call duration (s) and payload (MB) vs records per call

**Type:** line

**X:** rpc

**From:** 100

**To:** 5000

**Steps:** 50

**Series:**

- duration
- payload_mb

**Disclaimer:** Limits are as documented by Infor at time of writing and can change by release and component. Server time per record must be measured on your tenant; this is a planning estimate, not a guarantee.

## Content

**Intro:** This calculator sizes batches for calls through the Infor ION API Gateway and ION API Flows so each call stays inside the documented time-out and payload limits. Enter record size, server time per record and your limits to get a safe records-per-call figure and total run time.

**Sections:**

#### Documented limits behind the inputs

**Body:** The ION Desk guide (Infor OS 2026.x) says an API Flow's total duration must be under the global gateway time-out of 5 minutes, that asynchronous requests can run up to 2 hours, and that each individual task must finish in under 5 minutes. Payload limits are stated differently on two pages: 'Creating and using API flows' gives under 10 MB per flow and 5 MB per task, while 'Error handling' gives under 5 MB per flow and 3 MB per task. That is why the payload limit is an input defaulting to the lower 3 MB. The API Gateway documentation sets the configurable target time-out for an API between 60,000 and 300,000 ms. API Flows accept JSON request and response bodies only.

**Bullets:**

- API Flow: under 5 minutes end to end (synchronous)
- Asynchronous requests: up to 2 hours
- Per task: under 5 minutes; payload 3 to 5 MB depending on page
- Target time-out: 60 to 300 seconds

#### Formulas

**Body:** Payload per call in MB is records per call times bytes per record divided by 1,048,576. The payload cap is the floor of the limit in bytes divided by bytes per record. Call duration in seconds is fixed overhead plus records per call times milliseconds per record, divided by 1,000. Calls needed is the total records divided by records per call, rounded up. Total run time is the larger of calls times duration (calls made one after another) and calls divided by calls allowed per minute. The safe batch is the smaller of the payload cap and the number of records that finish within 70% of the time-out after overhead. The 30% headroom is a rule of thumb for slow periods, not an Infor figure.

#### Worked example

**Body:** With 50,000 records of 1,500 bytes, 20 ms per record, 400 ms overhead, a 60-second time-out and a 3 MB limit: 500 records per call is 0.72 MB and takes 10.4 seconds, so it fits. The payload cap is 2,097 records; the time cap is floor((42,000 - 400) / 20) = 2,080 records, so the safe batch is 2,080. At 500 per call you need 100 calls, about 17.3 minutes sequentially. Moving to 2,000 per call cuts it to 25 calls of 40.4 seconds each, about 16.8 minutes, with fewer requests to throttle. The chart shows where duration crosses the time-out and payload crosses the limit.

#### Common mistakes

**Body:** Measuring server time on a quiet test tenant and applying it to month-end production. Forgetting that JSON grows when characters are escaped or fields are added in a later release. Counting the time-out per API step but forgetting that an API Flow's total duration must also be under 5 minutes, so five 70-second steps fail. Retrying failed calls immediately and multiplying load. Ignoring that requests need a valid token; refresh it before it expires during long runs. Not checking the Infor OS service limits documentation for throttling that applies to your tenant.

**Cta:** If you need to move large volumes between an on-premises ERP and Infor OS, or let a private AI assistant call ION APIs safely, we can design the batching and error handling with you.

**Faq:**

#### What are the Infor OS service limits for API Flows?

**Answer:** Infor documents a 5-minute global gateway time-out for a synchronous flow, up to 2 hours for asynchronous requests, under 5 minutes per task, and payload limits given as 10 MB per flow and 5 MB per task on one page and 5 MB and 3 MB on another. Check the current release notes.

#### What is the ION API Gateway time-out?

**Answer:** The target time-out element in the API Gateway is configurable between 60,000 and 300,000 milliseconds per the Infor documentation. API Flows are also bound by the 5-minute global gateway time-out, so a flow of several slow steps can fail even when each step is under its own limit.

#### What are API Flows in Infor ION?

**Answer:** API Flows are modelled in ION Desk to chain REST API steps, with parallel and decision variants, filters, JQ transforms and Do While loops, and published as one endpoint. They are synchronous end to end and work with JSON only.

#### Why include a calls-per-minute input?

**Answer:** Many teams throttle their own integrations, and tenants can have usage limits. Total run time is the larger of sequential call time and the throttle-limited time, so a high rate limit does not help when each call is slow, and the reverse.

## Search terms

infor os service limits, infor os limits, infor ion api flows, api flows infor ion, ion api gateway timeout, ion api payload limit, infor ion api batch size

---
Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json