What I learn running NetSuite, explained plainly.

A Scheduled Script that processes everything in one run is fine with a few records. When volume grows, the trouble starts.

Scheduled Script

• A single run, capped at 10,000 governance units.
• If it runs out of units, it stops halfway.

Map/Reduce

• Works in stages: getInputData, map, reduce and summarize.
• Processes one record at a time: one error doesn’t stop the rest.
• In summarize you review what failed and why.

Before: the same payment was keyed into NetSuite, online banking, a spreadsheet and an email to the vendor. Four places, four chances for error.

After: one flow inside NetSuite

1. You select the bills to pay inside NetSuite.
2. NetSuite generates the bank file using your bank’s payment template.
3. You upload once to the bank, instead of one transfer per payment.
4. The payment is recorded against each bill.
5. The vendor gets an email notification.

A “small” change in production can break:

• A workflow that fires on another record.
• A script that depends on that field.
• An integration that expects the old format.

The rule

1. Sandbox first: every change is tested there.
2. Real data: copied from production.
3. Deploy after hours: with a way to roll back.

An external system sends order A-1042 and NetSuite creates the sales order. The response never arrives because of a network drop. The system retries… and without protection, a second order is created.

The fix is to make retries safe (idempotency). Use the external system’s identifier as the External ID: NetSuite won’t allow two records of the same type with the same External ID.

// RESTlet (SuiteScript 2.1): look up before creating
const existing = query.runSuiteQL({
  query: 'SELECT id FROM transaction WHERE externalid = ?',
  params: [payload.orderId]
}).asMappedResults()[0];

if (existing) return { id: existing.id, duplicate: true };

const so = record.create({ type: record.Type.SALES_ORDER });
so.setValue({ fieldId: 'externalid', value: payload.orderId });

BUILTIN.DF turns an ID into the text you see on screen: the customer, sales rep or location name.

SELECT
  t.tranid AS order_no,
  BUILTIN.DF(t.entity)    AS customer,
  BUILTIN.DF(t.employee)  AS sales_rep,
  BUILTIN.DF(tl.location) AS location
FROM transaction t
JOIN transactionline tl
  ON tl.transaction = t.id
WHERE t.type = 'SalesOrd'
  AND tl.mainline = 'T'
  AND t.status IN ('E', 'F')