العربية
BulkFollows
القائمة

مرحباً بك في BulkFollows

خطواتك اليومية.
كلها في مكان واحد.

سجّل الدخول لإدارة رصيدك وطلباتك والدعم.

← All guides
BulkFollows guide

SMM Panel API Timeout: Avoid Duplicate Orders

Published 2026-11-17

BulkFollows Editorial · Published 17 November 2026 · 6 min read

Your panel submits an order, the connection stalls, and no confirmation comes back. Clicking again feels like the obvious fix. It can also turn one customer request into two paid orders.

For an SMM panel API timeout, treat the outcome as unknown until you check it. A missing response tells you what your application received, not necessarily what the server did. This guide explains a cautious recovery workflow for resellers, and the records that make that recovery easier.

Guide cover: SMM Panel API Timeout: Avoid Duplicate Orders

What a timeout actually tells you

Three different events can look similar from the customer’s screen: the request never reached the provider, the provider rejected it, or the provider accepted it but the response was lost. A blank screen does not tell you which happened.

MDN’s explanation of idempotency makes the distinction useful here: a repeated request is safe only when repeating it has the same intended effect as sending it once. POST requests do not carry that guarantee by default. Waiting longer between retries does not itself prevent duplicate creation.

The BulkFollows API documentation uses POST for its actions. Its add-order example returns an order ID, while order-status lookup uses an existing ID. The published add-order parameters do not document an idempotency key. Do not assume that adding an invented reference field will make a repeat submission safe.

Record the attempt before sending

Give each customer request a reference in your own system. Keep that reference separate from the provider’s order ID: the first identifies the work you intended to submit; the second identifies the order the provider accepted.

Before sending, record the service ID, exact destination link, quantity, submission time with timezone, and the customer’s approval. Record the service conditions you checked as well. If someone later asks why a particular service was selected, a current catalog entry may not explain an earlier decision.

Use a distinct state such as “Submitting” while the attempt is underway. If a connection error leaves the result uncertain, move it to “Needs review,” not straight back to the queue. These are suggested labels for your own application, not BulkFollows status names.

Keep API keys out of that record. A useful activity log needs enough information to trace an attempt, not a copy of every secret included in the request.

Recover without a blind retry

  1. Stop additional submissions for this request. The customer should not be able to create a second attempt by refreshing, and another worker should not pick up the same job.
  2. Look for a saved confirmation. Check your application’s response record. An order ID may have been received even if the customer-facing confirmation failed.
  3. If you have the ID, check that order. Use the documented status action rather than creating another order. Display the returned status honestly; accepted does not mean completed.
  4. If you have no ID, review the account’s order history. Compare time, service, destination and quantity. A similar row is a lead, not proof when several legitimate orders share those details.
  5. Escalate an unresolved match. Ask support to confirm whether the attempt was accepted before deciding whether to submit again.

Do not use a balance change as the only evidence. Other orders, credits or funding can move the balance at the same time. The SMM reseller reconciliation guide explains how to connect a charge to the right record.

A worked reseller example

Illustrative scenario: a reseller approves one request for 2,000 units. At 10:02 UTC the application sends it, then times out. A second employee sees no confirmation and prepares to submit again.

The correct next step is to review the first attempt. If both submissions were accepted, the reseller would have requested 4,000 units across two orders, although the customer approved only 2,000. That arithmetic does not establish what either order would deliver or whether cancellation would be available.

What you findWhat to do next
A saved provider order IDCheck its status and connect it to the customer request.
A possible matching order, but no reliable matchKeep the request under review and confirm with support.
Confirmation that no order was acceptedReview current details and authorize one new attempt.

The small pause is deliberate. It protects the customer from an unexplained duplicate and gives the reseller a defensible history of what happened.

Build practical duplicate safeguards

Start with the ordinary causes: double-clicks, a slow page and two staff members handling the same request. Disable the submit control while processing, but also enforce the rule on your server. A disabled button alone cannot coordinate background workers.

Have your developer use one durable job record per approved request and prevent two workers from owning it at once. Keep an uncertain attempt out of automatic resubmission. Retain the provider response before sending the customer a success message.

Do not deduplicate solely by link and quantity forever. A customer may legitimately return later with another approved request. The identity of the approved job, plus its history, matters more than a superficial match between two forms.

Test the failure path with a mock provider or controlled local simulation first. Deliberately losing a real paid-order response is not a harmless test. Review the normal ordering workflow alongside the integration so staff know where to check existing orders.

What to send support

Make the question easy to investigate: “Was this request accepted?” Include the approximate submission time and timezone, service ID, destination link, quantity, any returned order ID, and the error your application observed. Say whether another attempt has already been made.

Use the private support route in your account. Remove API keys, passwords and unrelated customer information from screenshots or logs. Keep the case reference with the original job, then record the confirmed outcome and the next action.

Finally, close the loop with the customer in plain language. “We are checking whether the first request was accepted” is more accurate than “failed” when the result is still unknown. A reliable SMM reseller workflow makes uncertainty visible instead of hiding it behind another submit button.

Sources & further reading

DMCA.com Protection Status