Every MQ AMQ Message

When your application prints Reason 2035, the queue manager often wrote an AMQ message seconds earlier explaining which rule blocked access. AMQ is IBM MQ's operator language: structured IDs, severity, and sentences in AMQERR logs that survive restarts and feed SIEM pipelines. Beginners grep for words like unauthorized and miss AMQ9777E naming the CHLAUTH record. This encyclopedia page catalogs the AMQ families you will see in production—how numbering works, where logs live, how AMQ maps to MQRC, and detailed entries for the messages that dominate incident tickets. IBM publishes thousands of AMQ texts per release; you learn families and lookup discipline, not memorization.

AMQ Message Anatomy

A typical line includes timestamp, process name, message ID, severity, and text. Example pattern: AMQ8077E: IBM MQ channel 'PARIS.TO.LONDON' is blocked by channel authentication record. The E suffix is error. I suffix on AMQ8006I means informational confirmation after successful DEFINE. Warnings (W) deserve attention during migrations—disk space, certificate expiry windows, and retry storms often start as W. Fatals (F) indicate the queue manager or component may stop or require immediate intervention. Always copy the full AMQ ID into tickets; IBM Knowledge Center articles key off the ID, not your summary.

AMQ Numbering Families

IBM groups message numbers by subsystem rough ranges. AMQ4xxx and AMQ5xxx appear in older contexts; modern 9.x logs heavily use AMQ6xxx for commands and objects, AMQ7xxx for authority and security, AMQ8xxx for queue manager lifecycle and MQSC, AMQ9xxx for channels, listeners, SSL, and distributed components. The number is not a severity indicator by itself—AMQ9638E is SSL; AMQ8006I is benign confirmation. When building log alerts, filter on ID patterns and severity together, not on 9 alone meaning disaster.

AMQ6xxx and AMQ8xxx: Commands and Objects

These messages appear when you run runmqsc, when the command server validates DEFINE and ALTER, and when internal components create or delete objects. AMQ8006I-style text confirms an object was created or displayed. Errors in this family include syntax problems, object already exists without REPLACE, and path errors during queue manager creation. Operators running scripted MQSC in CI should fail the pipeline when stderr contains AMQ8xxxE, not only when the shell exit code is non-zero—runmqsc can return success with embedded errors depending on script handling.

AMQ7xxx and AMQ8077: Authority and CHLAUTH

Authority failures surface as AMQ8077E and related AMQ9777 family messages when channel authentication blocks a partner, when CONNAUTH rejects a user, or when OAM denies an operation. The text usually names the channel, user, or object. MQRC 2035 in the application log often pairs with these AMQ lines. Fix order: read the named CHLAUTH rule, determine if block is intentional, adjust rule or partner certificate and MCAUSER mapping, then grant setmqaut if the failure moved past CHLAUTH to OAM. Never disable all CHLAUTH as a default troubleshooting step in production without risk acceptance.

1
2
3
4
# Find authority-related AMQ in the last hour (Linux example)
grep -E 'AMQ8077|AMQ9777|AMQ9209|2035' /var/mqm/qmgrs/QM1/errors/AMQERR01.log | tail -50
DISPLAY CHLAUTH('*') TYPE(BLOCKUSER) ALL
dspmqaut -m QM1 -n SYSTEM.DEFAULT.MODEL.QUEUE -t q -p mqmapp

AMQ9xxx: Channels, Listeners, and SSL

Channel failures generate AMQ9208 and related texts when partners cannot connect, when sequence numbers disagree, when XMITQ processing fails, or when the channel enters retry. SSL handshake problems produce AMQ9638, AMQ9616, and similar messages referencing GSKit, cipher specs, or certificate labels. Correlation requires DISPLAY CHSTATUS on both queue managers with the same timestamp. A sender in RETRY with repeating AMQ9208 every SHORTTMR points to network or partner listener; repeating AMQ9638 points to TLS before MQ channel protocol runs.

Linking AMQ to MQRC

Many AMQ lines end with (Reason xxxx) or embed MQRC_NOT_AUTHORIZED. Treat the numeric Reason as the application-facing code and the AMQ ID as the queue manager narrative. If only AMQ appears, search the line for Reason or CompCode. If only MQRC appears in the app, search AMQERR within a five-second window of the application timestamp—clock skew between servers can widen the window in global deployments. Document both in the ticket so the next engineer does not repeat the search.

Reading and Rotating AMQERR

AMQERR01 is the active file; when size limits hit, MQ renames to AMQERR02 and opens a fresh AMQERR01. Long incidents may span both files. Use tail -f during a controlled test connect. For compliance, ship logs to centralized storage before rotation deletes local history. fmtmqfnc and FFST relate to severe failures—pair with AMQ F messages when IBM support asks for diagnostics.

Explainer: AMQ as the Building Intercom

Applications talk to MQ with numbers (MQRC). The queue manager announces problems on the building intercom (AMQ) so operators in the machine room hear what failed even if no application logged yet. Same incident, two speakers—you need both when debugging end-to-end.

Explain Like I'm Five: AMQ Messages

When the mail robot has a problem, it writes a note in a special notebook with a code like AMQ9208. Grown-ups look up that code in IBM's dictionary to see if the door was locked, the key was wrong, or the road was blocked.

Practice Exercises

Exercise 1

Provoke CHLAUTH block in lab; capture full AMQ9777 or AMQ8077 line and matching MQRC.

Exercise 2

Build a grep alias for your top five AMQ IDs used in your environment.

Exercise 3

From one SSL failure, list AMQ, CHSTATUS LASTCHLERR, and client error in a table.

Frequently Asked Questions

Test Your Knowledge