Asterisk documentation
Build your understanding of Asterisk one task at a time. These guides explain the configuration behind a phone system, the decisions that affect call routing, and the evidence to collect when a call fails.
Source-reviewed, 18 October 2026. Prepared from the cited documentation. The configuration examples have not been independently tested by Phone Guides. Our current reference branch is Asterisk 22 with PJSIP; individual guides state narrower conditions where needed.
Start here
| Guide | What you will learn |
|---|---|
| Asterisk architecture | How channels, dialplans, applications and bridges fit together |
| Version and installation choices | How to choose a release branch and installation owner |
| Installing Asterisk | How to follow a documented source or package installation path |
| Files, CLI and reloads | Where settings live and how to inspect the running system |
| Security baseline | How to separate calling permissions and management access |
| PJSIP configuration | How transport, endpoint, authentication and address objects relate |
| Register two phones | How to configure two example local accounts and inspect contacts |
| First internal call | How a dialplan connects those accounts and what to check |
| Contexts and patterns | How routes and includes affect what a caller can reach |
| First troubleshooting checks | How to separate registration, routing and media faults |
| SIP trunks | How registration, authentication and provider requirements fit together |
| Inbound and outbound routing | How to map incoming numbers and grant narrow calling permissions |
| NAT and RTP | How to investigate advertised addresses and actual audio paths |
| Backup and restore | What to preserve and how to plan recovery checks |
Add calling features or an integration
| Guide | What you will learn |
|---|---|
| Variables and Gosub | How to name values and organise returning subroutines |
| Voicemail | How mailbox configuration, message deposit and protected retrieval fit together |
| Auto attendant | How to offer explicit menu choices with invalid-input and timeout paths |
| Conferencing | How room identifiers, bridge profiles and participant profiles differ |
| Time and holiday routing | How timezone, weekly hours and exceptions select an explicit route |
| Transfers and pickup | How feature-code permissions differ from phone controls |
| Call parking | How lots, retrieval access and timeout destinations fit together |
| Device state and BLF | How hints, subscriptions, phone displays and button actions differ |
| Paging and intercom | How Asterisk calls, endpoint auto-answer and multicast are separate paths |
| Call recording | How to define recording scope, completion, file access and failure handling |
| Choose AGI, AMI or ARI | How to match an integration interface to its task |
| ARI and Stasis | How an application receives control and handles call-lifecycle events |
Build applications and larger deployments
| Guide | What you will learn |
|---|---|
| AGI and FastAGI | How to bound script execution, protocol handling and service failures |
| AMI actions and events | How to restrict management access and correlate requests with outcomes |
| ARI call lifecycle | How to handle late events, resource ownership, retries and cleanup |
| Database and realtime | How runtime data, configuration storage and schema ownership differ |
| Call files | How to prepare complete jobs and handle scheduling and repeated work |
| WebRTC | How browser signalling, trust, endpoint options and media traversal fit together |
| IVR service integration | How to define a bounded transaction and safe caller fallback |
| External media and speech | How to separate call control, audio processing and service results |
| AI integration architecture | How recorded-audio processing, provider-direct SIP and PBX media bridges differ |
| Live voice agents with Asterisk | How to specify a 22.8+ WebSocket audio adapter and test its call, audio and recovery paths |
| Secure AI actions and human handoff | How to authorize tool requests on the server and transfer a caller safely |
| Plan repeatable integration tests | How to combine the Asterisk Test Suite, SIPp and controlled fakes, and identify what still needs a live call |
| FreePBX AI integrations | What Scribe does with recordings and voicemail, and what to confirm about data processing |
| AI provider contract checklist | Dated model/API versions, audio and session limits, lifecycle, prices and data terms |
| Multiple sites | How to define numbering, inter-site permissions and disconnected behaviour |
| Resilience and recovery design | How service recovery, data recovery and active call survival differ |
| Performance and load-test planning | How to describe a workload and measure outcomes without unsupported sizing promises |
| Fax and T.38 | How application results differ from gateway and carrier compatibility |
| Analogue gateways and DAHDI | How port roles, signalling and exact hardware support differ |
Operate and diagnose the system
| Guide | What you will learn |
|---|---|
| Call queues | How membership, eligibility, waiting and overflow differ |
| DTMF keypad signalling | How to distinguish telephone events, inband tones and SIP INFO |
| Codecs and transcoding | How configured policy differs from negotiated formats and measured load |
| Logs, CDR, CEL and queue_log | Which evidence answers each reporting or diagnostic question |
| Upgrades and rollback | How to review changes, preserve recovery data and check the resulting service |
| TLS and SRTP | How signalling protection, media protection and certificate trust differ |
| Monitoring and alerts | How to observe service health and give each alert a useful response |
| Australian numbering and site settings | How number categories, carrier formats, regional settings and service arrangements differ |
Investigate a specific failure or migration
| Guide | What you will learn |
|---|---|
| Registration failures | How to follow contacts, outbound registrations and authentication exchanges |
| One-way or missing audio | How to compare negotiated and observed media in each direction |
| Disconnected calls | How to identify the terminating leg and preceding application or network evidence |
| DTMF failures | How to follow keypad input from the sending device to the listening application |
| Caller ID and routing failures | How peer identity, called destinations and presentation differ |
| chan_sip to PJSIP migration | How to map account behaviour into current objects and preserve reversal |
| Legacy dialplan review | How to review removed components, arguments and route permissions |
| Asterisk and FreePBX boundaries | How to identify generated configuration and supported customisation owners |
Choose the right kind of instruction
These pages address Asterisk configuration directly. A system managed by FreePBX or another product has its own configuration ownership and update process. Follow that product's supported settings and extension points before changing generated files. A raw Asterisk example is a way to understand behaviour, not permission to overwrite a managed PBX configuration.
The official Asterisk documentation remains the detailed upstream reference. Use its version-specific API pages when checking an application's arguments or a module option. Keep the installed package version and the documentation branch together; a new reference page can describe an option missing from an older installation.
Work from the task to the evidence
For configuration, identify the installed version, the relevant file or configuration database, and which component owns it. Before changing a setting, preserve the previous value and decide how to return to it. After the change, inspect the running state and follow the reader checks in the guide.
For troubleshooting, record where the failure occurs: registration, call setup, ringing, answer, audio, keypad input, transfer or hangup. A phone showing as registered answers only the registration question. It does not establish that every later stage works.
For integrations, identify whether the application needs dialplan execution, events and management, or control of channels and bridges. The phone system integrations guide links the official ARI starting point. Keep application credentials on the application server.
Scope and review method
We use historical Asterisk documentation to identify subjects worth explaining, then check present instructions against public upstream references. We write our own explanations and examples. Historical driver names and old operating-system commands are not carried forward as current recommendations.
Where the documentation leaves a device, provider or version detail unresolved, we state the limit. Future hands-on results can supplement an article without changing what its cited sources actually establish.
Related references: VoIP protocols, VoIP resources and the official Asterisk 22 reference.