Passer au contenu
Technologie 7 novembre 2026 11 lecture minimale

SIP Error Codes on Outbound Calls: 403, 486, 503 and More

What each SIP failure code means on an outbound dialer, from 403 Forbidden to 486 Busy, 488 and 503, which side caused it, and the first fix to try for each.

D
7 novembre 2026

Réponse rapide

A SIP error code is the answer to a call attempt. 4xx means the request had bad syntax or could not be fulfilled at that server, 5xx means the server failed to fulfil a valid request, and 6xx means the request cannot be fulfilled at any server, typically because the callee declined. On outbound dialers, 401, 407 and 403 usually point to authentication or caller ID setup, 486, 480 and 603 mean the callee's side was reached but the person was busy, unavailable or declined, and 503 often signals a carrier capacity or rate limit.

When an outbound campaign starts failing, the first useful question is not "what is wrong with the carrier?" but "which code came back, and who sent it?" SIP, the signalling protocol behind almost every business phone call, answers each call attempt with a numeric response. The number tells you which side to look at. This guide goes through the codes an outbound dialer meets most often, what usually causes each one in a dialer setting, the first fix to try, and how each should be recorded so your retry rules do the right thing.

The code meanings come from the SIP specification itself, RFC 3261, and the registry of response codes kept by IANA. The causes and dispositions in this article are practical guidance for dialer operators, not part of the standard, and your carrier may use codes slightly differently. Treat your own call logs as the final word.

How to read a SIP response: the six classes

Every SIP response starts with a three-digit code, and the first digit is the class. RFC 3261 section 7.2 defines six of them.

ClassMeaning in the RFCWhat it looks like on a dialer
1xxProvisional: request received, continuing to process100 Trying, 180 Ringing, 183 Session Progress. Not an error.
2xxSuccess200 OK: the call was answered.
3xxRedirection: further action neededThe carrier points you to another contact. Rare on outbound trunks.
4xxClient error: bad syntax or cannot be fulfilled at this serverWrong credentials, number not allowed, person busy, number does not exist.
5xxServer error: valid request, server failed to fulfil itCarrier overloaded, gateway down, upstream timeout.
6xxGlobal failure: cannot be fulfilled at any serverThe person explicitly declined, or is busy everywhere.

Read the code together with the reason phrase and any extra headers, and note which device sent the response: a 503 from your own trunk edge and one relayed from the destination network are different problems.

4xx on outbound calls: refused or not fulfilled

Most failures an operator investigates are 4xx codes, and they split into two groups: setup problems on your side of the trunk, and outcomes about the person being called.

401 and 407: authentication challenge

RFC 3261 says 401 Unauthorized means the request requires user authentication and is issued by user agents and registrars, while 407 Proxy Authentication Required is the same idea issued by proxy servers. In normal operation a digest-authenticated trunk answers the first INVITE with a 401 or 407 and the dialer repeats the request with credentials. That single challenge is expected and is not an error.

The problem is when the challenge repeats or the call ends on it. Check the username, password and realm, and confirm which authentication mode your carrier expects (IP allow-list or credentials). If the carrier expects IP authentication but receives requests from an address that is not on its allow-list, the result can look like an authentication failure.

403 Forbidden: the one to look at first

RFC 3261 defines 403 as the server understanding the request but refusing to fulfil it, where authorisation will not help and the request should not be repeated. For an outbound dialer that makes 403 the most informative code, because the carrier is deliberately saying no. The usual causes, roughly in order of frequency:

  • Caller ID not provisioned. The number you present is not authorised on the trunk. Carriers commonly reject calls presenting a number they have not provisioned for your account. Check the number in the From and, where used, the P-Asserted-Identity header against the DIDs on your carrier account.
  • Source address not allowed. With IP authentication, the carrier does not recognise the address the request came from, for instance after a change of platform IPs.
  • Destination not permitted. The trunk is barred from the country or number range you dialled, or the account has no credit or a spending limit.

A 403 should not be repeated, so do not let a dialer retry it blindly. Pull ten failed calls and compare the presented caller ID and destination with what your carrier has provisioned. The BYOC setup guide lists the same checks from the configuration side.

404 Not Found, 484 Address Incomplete and the dialled number

RFC 3261 defines 404 as the server having definitive information that the user does not exist at the domain in the Request-URI. Inside a carrier network it is also how an unallocated number is often reported (see the RFC 3398 mapping below). Treat it as an invalid number. If it hits a whole prefix at once, suspect your number formatting (missing country code, stray leading zero) before the list. 484 means the Request-URI was incomplete, which again points at number formatting, not at the person.

408 Request Timeout

408 means the server could not produce a response in a suitable time, and the client may repeat the request later. On outbound calls it usually means no response from the far end. Retry later; if one destination keeps returning it, test another route.

480 Temporarily Unavailable

RFC 3261 describes 480 as the callee's end system being contacted successfully but the callee currently unavailable, for example not logged in or with do-not-disturb on, and the response may suggest a better time to call in a Retry-After header. Carriers commonly return it for a phone that is switched off, out of coverage or not answering; RFC 3398 maps the network causes "no answer from the user" and "subscriber absent" to 480. For a dialer this is a "reached the network, not the person" outcome: record it as no answer or unavailable and retry later according to your rules.

486 Busy Here

486 means the callee's end system was contacted but is currently not willing or able to take additional calls at this end system. It is the classic busy signal. Record it as busy and retry sooner than a no-answer, because busy lines tend to free up quickly. Note that RFC 3261 says the person could still be available elsewhere, so it is a per-device signal.

487 Request Terminated

487 is returned when the original INVITE is terminated by a CANCEL or BYE. Your dialer will see plenty of these and they are normal: a predictive or power dialer cancels attempts that are still ringing when it no longer needs them, or that have timed out. Do not count 487 as a carrier failure. In reports, separate "cancelled by us" from real failures, otherwise a healthy trunk looks broken.

488 Not Acceptable Here: a media problem

RFC 3261 gives 488 the same meaning as 606 (Not Acceptable) but only for the specific resource addressed, so the request may succeed elsewhere. In practice on trunks it almost always means the offer in the call setup, the codecs or media formats, was not acceptable to the other side. Check that your codec list shares at least one codec with the carrier for the destination, and review any encryption setting such as SRTP that the far end does not support. The setup guide linked above covers codec preferences, and the guide to SIP 488 and codec mismatch goes through the causes in detail.

5xx: the server side

500 Server Internal Error and 502 Bad Gateway

500 means the server hit an unexpected condition, and the client may retry after several seconds. 502 means a gateway or proxy received an invalid response from the downstream server. Both point at the carrier or a network element beyond it. A few now and then are normal; a stream of them is a ticket for your carrier with timestamps and call identifiers.

503 Service Unavailable

RFC 3261 defines 503 as the server being temporarily unable to process the request because of temporary overloading or maintenance. The server may include a Retry-After header, and a client receiving a 503 should try an alternate server and should not send further requests to that server for the Retry-After period if one is given. If no Retry-After is present, the client must treat it as a 500.

For outbound dialers, 503 is the most common sign of a rate or capacity limit. The pattern to recognise: calls succeed normally, then a burst of 503s appears at campaign start, after a break, or when many agents log in together, and it clears when the dialing rate falls. That points at a calls-per-second ceiling, not at bad numbers. The companion article on SIP trunk CPS limits for predictive dialers shows how to work out the rate a campaign needs. Other causes of 503 are channel limits (all licensed concurrent calls in use), carrier maintenance and a failed route to the destination network.

504 Server Time-out means the server did not receive a timely response from an external server it asked: treat it as a network failure, retry, and ask the carrier to route differently if it concentrates on one destination.

6xx: global failures

600 Busy Everywhere means the callee is busy and does not wish to take the call, and is returned only if the client knows no other endpoint, such as voicemail, will answer. 603 Decline means the callee's machine was reached but the user explicitly does not wish to or cannot participate; again it is for when no other endpoint will answer. Both may carry a Retry-After.

A 603 on an outbound call is often a person pressing reject. Record it as a decline and avoid hammering the number: one more attempt at another time of day is reasonable, and repeated declines are a signal to rest the number.

Why the same failure shows different codes per carrier

Many calls cross telephone networks that signal with ISDN User Part (ISUP) cause values, and a gateway translates them into SIP responses. RFC 3398 recommends the mapping, which explains why the same event can surface differently. Some recommended mappings:

ISUP causeSignificationRecommended SIP response
1Unallocated number404
17User busy486
18No user responding408
19No answer from the user480
21Call rejected403 (603 may be given instead if the cause location is "user")
28Address incomplete484
34, 38, 41, 42, 47No circuit, network out of order, temporary failure, congestion, resource unavailable503
65Bearer capability not implemented488
102Recovery on timer expiry504

Three consequences follow. Cause 21 (call rejected) maps to 403 or 603, so a 403 does not always mean your own setup is wrong. Many different internal failures collapse into 503, so 503 needs the context of timing and pattern. And carriers may choose their own mapping, so the same event through a different route can return a different code. Ask your carrier for the original Q.850 cause when they investigate; RFC 3398 notes it can travel in a SIP Reason header.

Which codes should count as busy, no answer, invalid or stop

A dialer needs each failure turned into a disposition, because the disposition decides whether and when the number is called again. The table below is a suggested starting point, not a standard. Tune it with your own data and with the rules that apply to your campaign.

Codedisposition suggéréeRetry guidance
486, 600BusyRetry after a short interval within your attempt limit.
480, 408, no final answerNo answer / unavailableRetry at a different time of day.
603DeclinedRetry sparingly; rest the number after repeated declines.
404, 484Invalid numberCheck formatting first; if the format is right, stop dialing that number.
401, 407 (repeated), 403System error: trunk or accountDo not count against the contact. Fix configuration, then requeue the affected records.
488System error: mediaFix codecs, then requeue.
500, 502, 503, 504System error: carrierRetry soon, honour Retry-After, fail over to another route, and throttle if the pattern is rate related.
487Cancelled by the dialerNot a failure. Requeue per normal rules.

The key rule is to separate contact outcomes from system errors. A 503 burst recorded as "no answer" burns attempts against contacts who were never reached. A "do not call" request is never inferred from a SIP code: it comes from the person.

Session establishment metrics that use these codes

RFC 6076 defines standard measures of session establishment. The Session Establishment Ratio (SER) is the number of new INVITE requests that result in a 200 OK, divided by the total attempted INVITEs less those that resulted in a 3xx. The Session Establishment Effectiveness Ratio (SEER) is similar but also counts INVITEs ending in 480, 486, 600 or 603 in the numerator, because those responses show the network did its job and the outcome was the user's.

A drop in SER alone may be nothing more than more people being busy or declining. A drop in SEER points at the network or your setup: 403s, 503s, timeouts. Track both by carrier and destination, and set thresholds from your own baseline; the RFC defines the measures, not a target.

Comment DialerBee s'intègre

On a bring-your-own-carrier trunk the SIP responses described here are exactly what your carrier sends back, so the troubleshooting starts at the gateway. The BYOC feature et le step-by-step setup guide cover authentication mode, caller ID, codecs and the CPS and channel limits on the Routing tab, which are the settings behind most of the codes above. Outcomes are recorded as dispositions: the glossary entry on call dispositions explains how system outcomes such as busy, no answer and invalid number can be coded by rule, and how those codes drive retry behaviour. A numéroteur prédictif adds pacing, so a rate-related 503 can be addressed from both the gateway limit and the dialing intensity; see the tuning guide. For the basics of the trunk itself, start with what a SIP trunk is.

Foire aux questions

What does SIP 403 Forbidden mean on an outbound call?

The server understood the request but refuses to fulfil it, and RFC 3261 says the request should not be repeated. On outbound trunks the usual causes are a caller ID that is not provisioned on the account, a source address that is not on the carrier's allow-list, or a destination the trunk is not permitted to call.

What is the difference between SIP 486 Busy Here and 600 Busy Everywhere?

486 says the callee is busy at this end system, and the person may be reachable elsewhere. 600 says the callee is busy everywhere and is returned only if the client knows no other endpoint will answer. A dialer can record both as busy and retry later.

Why do I get SIP 503 at the start of a campaign?

The most common reason is a calls-per-second or capacity limit at the carrier. A dialer can launch many calls in the first second of a campaign. The 503 clears when the launch rate falls below the carrier limit. Check the Retry-After header, the channel count and the CPS limit configured on both sides.

Is SIP 480 Temporarily Unavailable the same as no answer?

Not strictly. The network reached the callee's end system, but the callee was unavailable, for instance switched off or in do-not-disturb. For disposition purposes most dialers treat it like no answer and retry at another time. RFC 3398 maps ISUP cause 19 (no answer) and cause 20 (subscriber absent) to 480.

What is SIP 603 Decline?

The callee's device was reached and the user explicitly does not wish to or cannot participate. RFC 3261 says it is returned only when no other endpoint would answer. Record it as a decline and avoid repeated retries.

How do Q.850 cause codes relate to SIP codes?

Q.850 cause values are used inside traditional telephone networks. When a call crosses between those networks and SIP, a gateway translates the cause into a SIP code, using a mapping such as the one recommended in RFC 3398. Different carriers may map the same cause differently, so ask for the original cause when investigating.

Sources

Note: Causes, suggested dispositions and retry guidance in this article are practical guidance for dialer operators, not part of the standards cited. Carriers may use codes differently. Confirm behaviour with your own carrier and call logs.

Lectures complémentaires

À lire aussi : SIP Trunk Registration Failed: How to Fix It

Prêt à voir DialerBee en action ?

Demandez une démo en direct de 15 minutes, ou démarrez un essai gratuit de 14 jours — sans diaporama, sans engagement. Les appels démarrent dès que votre propre trunk SIP ou PBX est connecté et approuvé.

Essai gratuit de 14 jours · crédit IA gratuit · sans carte · 11 langues · BYOC · contrôles qui contribuent à la conformité

Voir le site complet en anglais →