Q.850 cause 38 Network out of order

What Q.850 cause 38 means

Something in the network path is broken, and Q.850 says the condition is expected to last a while, so redialling straight away is unlikely to help. Treat cause 38 as an outage to report rather than a glitch to ride out: check whether your trunk is up, then get a fault reference from the carrier and move traffic to a backup route.

Resource unavailable class
Sent by The carrier network or an intermediate switch
Common

What Q.850 cause 38 Network out of order actually means

The switch that reported this found the network unable to carry the call and is telling you the condition is persistent rather than momentary. That single word, persistent, is the whole point of the code, because Q.850 already provides cause 41 for the short-lived version of the same problem. A gateway translating cause 38 into SIP reports it as 503 Service Unavailable under RFC 3398, which is why a SIP trace alone cannot tell you whether to retry. The Q.850 cause recorded in the call record is the part that says do not bother.

This cause indicates that the network is not functioning correctly and that the condition is likely to last a relatively long period of time; e.g., immediately re-attempting the call is not likely to be successful.
— ITU-T Q.850 section 6.2.7.2.2

The SIP equivalent of Q.850 cause 38

A gateway translating this cause into SIP reports it as SIP 503 Service Unavailable. The mapping is defined in RFC 3398 section 7.2.4.1.

What causes Q.850 cause 38 on a business phone system

A carrier or transit network outageA failed node, a cut fibre or a botched change at the carrier takes a route out of service for hours rather than seconds. Your calls to some destinations fail while others are fine, because only part of the routing is affected.
The SIP trunk registration has dropped and stayed downOn an on-premise PBX a registration that fails and does not recover leaves the system with nowhere to send calls. The symptom is total, not intermittent, and it usually starts at a specific moment you can find in the log.
A route or number translation broken by a recent changeRouting tables edited during a migration or a number port can leave a destination with no valid path. Ask what changed shortly before the first failure, because this class of fault almost never appears on its own.
The site access circuit itself has failedIf the broadband or leased line carrying your voice traffic is down, the nearest working element upstream reports the network as out of order. Check whether data services from the same site are also affected.
A legacy ISDN leg in the path has failedWhere calls still traverse an old ISDN circuit before reaching the modern network, a failure on that leg surfaces as cause 38 to everyone downstream. This is increasingly common as ageing circuits are left in place during the run-up to the PSTN switch-off.

How to fix Q.850 cause 38

Confirm whether your own trunk is up before anything elseCheck registration status or trunk state on the PBX or in the provider portal. If your side is down, nothing you do about routing will help until it is back, and you have narrowed the fault to your site in under a minute.
Test the same number from a different routeDial the failing destination from a mobile or a second provider. Success elsewhere confirms the problem sits in one carrier path rather than with the number you are calling.
Raise it with the carrier with evidence, not just a descriptionGive exact timestamps in local time, the dialled numbers, the calling number and a Call-ID if you have one. A ticket with three specific failed attempts gets diagnosed far faster than one that says calls are failing.
Move traffic while the fault is openDivert the main number to mobiles or a second provider so the business keeps taking calls. Confirm the divert works by ringing the number from outside rather than assuming the change applied.
Put a standing failover in place afterwardsA pre-configured out-of-service destination that triggers automatically turns a repeat of this outage into a minor inconvenience. Test it twice a year, because an untested failover tends to fail at the moment you need it.

Fixing the underlying problem

This page explains the code. These guides walk through the fix in detail.

Questions about Q.850 cause 38

Should I keep redialling after cause 38?

No, and that is the specific point of this cause rather than a general observation. Q.850 defines cause 38 as a condition likely to last a relatively long time, so repeated attempts mostly waste effort and hide the size of the problem. Report it and divert instead.

What is the difference between cause 38 and cause 41?

They describe the same kind of network failure and both become a SIP 503, but they carry opposite advice. Cause 38 says the fault will persist, while cause 41 says it is short-lived and trying again almost immediately may well succeed.

Does cause 38 mean the fault is at my end?

Not necessarily, but it does not rule it out either. The cause tells you the network could not carry the call, not where the break is, so start by proving your own trunk and access circuit are healthy before escalating upstream.

Related Q.850 cause values

Source. Q.850 cause 38 Network out of order is defined in ITU-T Q.850 section 6.2.7.2.2 and registered in the ITU-T Recommendation Q.850 (10/2018). The causes and fixes above are drawn from our own experience supporting UK business phone systems.

All Call hangup causes (Q.850) codes