# EDI Translator Throughput and Backlog Calculator

> Source: https://netray.co/tools/edi-translator-throughput-and-backlog-calculator
> Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems.
> Section: Free Tools

## Short answer

Size your EDI translator for peak load and estimate how long a backlog takes to drain.

## Category

erp-operations

## Audience

- EDI developer
- ERP administrator
- IT manager

## Tool

**Kind:** calculator

**Inputs:**

#### Peak documents per hour

**Key:** peak

**Type:** number

**Default:** 3000

**Min:** 0

**Step:** 100

**Unit:** docs/hr

**Help:** Busiest hour, such as month-end invoicing or a retailer order drop.

#### Normal documents per hour

**Key:** normal

**Type:** number

**Default:** 800

**Min:** 0

**Step:** 50

**Unit:** docs/hr

#### Processing time per document

**Key:** sec_doc

**Type:** number

**Default:** 0.8

**Min:** 0.01

**Step:** 0.1

**Unit:** sec

**Help:** Translate, validate, map and write to the ERP interface for one thread.

#### Parallel threads

**Key:** threads

**Type:** slider

**Default:** 4

**Min:** 1

**Max:** 16

**Step:** 1

**Unit:** threads

#### Effective availability

**Key:** avail

**Type:** slider

**Default:** 85

**Min:** 30

**Max:** 100

**Step:** 5

**Unit:** %

**Help:** Allowance for ERP import locks, retries and maintenance.

#### Burst duration

**Key:** burst

**Type:** number

**Default:** 2

**Min:** 0

**Step:** 0.5

**Unit:** hours

**Help:** How long the peak rate lasts.

**Outputs:**

#### Capacity per hour

**Key:** capacity

**Expr:** threads * 3600 / sec_doc * avail / 100

**Format:** integer

**Primary:** true

#### Utilization at peak

**Key:** util

**Expr:** peak / capacity * 100

**Format:** percent

**Primary:** true

**Description:** Above 100 percent a backlog builds during the burst.

#### Backlog at end of burst

**Key:** backlog

**Expr:** max(peak - capacity, 0) * burst

**Format:** integer

#### Spare capacity at normal load

**Key:** spare

**Expr:** max(capacity - normal, 0)

**Format:** integer

#### Hours to clear backlog

**Key:** clear

**Expr:** if(gt(spare, 0), backlog / spare, 0)

**Format:** hours

**Chart:**

#### Peak utilization by thread count

**Type:** line

**X:** threads

**From:** 1

**To:** 16

**Steps:** 16

**Series:**

- util

**Disclaimer:** Throughput rarely scales linearly with threads because of database and ERP interface limits. Validate with a load test.

## Content

**Intro:** EDI traffic arrives in bursts: order drops, month-end invoices, forecast refreshes. If your translator cannot keep up at peak, documents queue and acknowledgments miss partner deadlines. This calculator sizes capacity and backlog clearing time.

**Sections:**

#### How to use it

**Body:** Find the busiest hour in your translator logs, and the normal hourly rate. Measure the average processing time of one document from receipt to ERP write, using a representative sample of mixed sizes. Enter threads and an availability factor for retries and locks. The calculator shows capacity, peak utilization, the backlog after a burst and the time to drain it.

**Bullets:**

- Measure end to end time, including the ERP import step
- Use a mix of small and large documents
- Set availability below 100 percent to reflect real conditions

#### The model

**Body:** Capacity per hour equals threads times 3,600 divided by seconds per document, multiplied by availability. Utilization is peak load divided by capacity. If peak exceeds capacity, the backlog equals the excess per hour times burst duration. The backlog clears at the spare capacity, which is capacity minus normal load. If there is no spare capacity, the queue never clears.

#### Reading the result

**Body:** Keep peak utilization comfortably below 100 percent, typically under 70 to 80, so that retries and uneven arrival do not create queues. Customers often expect a 997 within minutes, so a growing backlog quickly becomes a compliance issue. The chart helps you decide how many threads are enough before the ERP import becomes the bottleneck.

#### Common mistakes

**Body:** Adding threads does not help if the ERP interface table, database locks or a single-threaded import job is the real limit. Also check that acknowledgments are generated before the long ERP step, so partners receive timely 997s even when the back end is slow.

**Cta:** If you want queue depth and processing time monitored with alerts from your EDI and ERP logs, we can build that on your own servers.

**Faq:**

#### How fast should a 997 be sent?

**Answer:** Partners usually expect acknowledgments within minutes to a few hours, and some set strict windows in their guides. Check each partner's rules, since late acknowledgments can trigger resends or compliance flags.

#### How do I measure time per document?

**Answer:** Use translator logs to compute average time between receipt and completion for a representative batch, or run a test file of known size through a single thread. Include the write to the ERP interface in the figure.

#### Why is availability below 100 percent?

**Answer:** Retries, lock waits, garbage collection, restarts and maintenance windows all reduce effective capacity. A factor of 70 to 90 percent is a sensible starting range until you have measured data.

#### Will more threads always fix a backlog?

**Answer:** No. If the shared resource is the ERP import or the database, extra translator threads only increase contention. Find the slowest stage in the pipeline first, then scale that stage.

## Search terms

EDI translator capacity, EDI throughput calculator, EDI backlog clear time, EDI peak load sizing, EDI processing time per document, EDI server utilization, EDI queue backlog, EDI performance sizing

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