IBM Support

MustGather: Data to collect for a Hung Job - SERVICEDOCS

Troubleshooting


Problem

This document details the steps for a user to collect SERVICEDOCS.

Resolving The Problem

NOTE:  Before running SERVICEDOCS or SEIZEINFO macros, please make sure the following PTF for APAR MA49489 is applied to the system:
V7R5:  MF69519
V7R4:  MF69518
V7R3:  MF69517
V7R2:  MF69516
NOTE:  Before running SERVICEDOCS  macro, please make sure the following PTF for APAR MA50379 is applied to the system:
V7R5:  MF71435
V7R4:  MF71434
V7R3:  MF71436
 
This document details the steps for a user to collect SERVICEDOCS.
NOTE:  It is recommended that SERVICEDOCS be collected while the job is hanging if root cause analysis is needed.  Sometimes SERVICEDOCS does not contain enough information to fully determine root cause and other data may be needed.  A main store dump is the best data to collect to assist with job and system hangs.

Instructions to collect SERVICEDOCS follow:

If you are collecting SERVICEDOCS for a hung job situation, please send in the job number(s) of the hung job(s) to IBM. Without the job number(s), it might not be possible to debug the SERVICEDOCS.
Important Considerations for a hung job situation:
  • Do not attempt to end the subsystem. This makes the entire subsystem unusable and will not help to end the hung job.
  • Do not issue ENDJOBABN without a basic understanding of the condition of the job. In some cases, this can actually cause the job to take longer to end. Using the ENDJOBABN on a job may change the original call stacks for the job impeding troubleshooting efforts.
  • Do not IPL without any analysis. In many cases an IPL is not necessary. If you must IPL, at the very least collect SERVICEDOCS before the IPL.
This video explains more about issues you may run into when using the ENDJOBABN command:
Before starting SST, please verify that your jobs spooled files go to an output queue that does not have an active writer assigned to it. 
 
Please take the following steps to dump SERVICEDOCS:
  1. On the operating system command line, type STRSST and press the Enter key.
  2. Select Option 1, Start a service tool, and press the Enter key,
  3. Select Option 4, Display/Alter/Dump, and press the Enter key,
  4. Select Option 2, Dump to printer, and press the Enter key,
  5. Select Option 2, Licensed Internal Code (LIC) data, and press the Enter key,
  6. Select Option 14, Advanced analysis, and press the Enter key,
  7. Select SERVICEDOCS from the list, and press the Enter key,
  8. Leave OPTIONS blank, and press the Enter key,
    Important Note: By default, the SERVICEDOCS macro no longer collects the LOCKINFO information.
    This change was made because including the LOCKINFO macro can cause the spooled file to be large, this macro is not necessary unless you have jobs stuck in LCKW status.
    If you need to collect the lock information also, enter one of the following OPTIONS:
    Option Purpose
    -lock Use this option if you can see the locks are on non-database objects (IFS files, profiles, queues, etc). Includes a call to macro LOCKINFO with option TABLE
    -lockd Use this option if the locks are database file and records in files. Includes a call to macro LOCKINFO with option TABLE = D
  9. The next screen asks for a Dump Title. You must give it a title that does not contain the name and number of the hung job (for search reasons). Increase the Through Page to 9999999. After the title and Through Page are set, press the Enter key.
  10. After pressing the Enter key, you will immediately see the Dump to printer successfully submitted message at the bottom of the screen.

    Important Note: This does not mean that the dump is done; exiting SST before it is done will cause it to be incomplete. To determine when the dump is done, you should do the following:

    a. Press F12 back to the Display/Alter/Dump Output Device screen (press F12 four times).
    b. Select Option 7. Display dump status.
    c. Do not exit the service tools until the Dump requests not complete is 0.
  11. The spooled file is named QPCSMPRT.
Transferring Files to IBM
Method 1  QMGTOOLS - FTP Spoolfile to IBM:
If the QMGTOOLS library is loaded on the system and you are able to connect to Internet directly from the System, perform the steps as per the following document:
Method 2 Downloading the Spool file to your PC using ACS: 


To retrieve the spooled file that you own, do the following:

1.
Open the IBM i Access Client Solutions Printer Output: 
image 2539
2. Log in with your user ID (preferably QSECOFR or equivalent).
3.
Find the spool file on the list and select the option to Download:
image 2540
4.
Save the spool file to your PC:
image 2541
Do not use the PDF format.  
5. Once the Spool file has been created use the following methods to send the data to IBM.
For transferring files to IBM that are smaller than 2GB each (QPCSMPRT files can exceed 2GB in total)
If you have an existing Case
  1. Upload the file at the following URL: https://www.secure.ecurep.ibm.com/app/upload_sf
  2. Enter you Case number and click Continue.
  3. Click to Select Files and find them in your local PC.
  4. Select Upload Files to transfer your file to the IBM i Global Support Center.
For transferring files to IBM that exceed 2GB each
Refer to technote doc - MustGather: Instructions for Sending Data to IBM i Support:
http://www.ibm.com/support/docview.wss?uid=nas8N1019224

[{"Type":"MASTER","Line of Business":{"code":"LOB68","label":"Power HW"},"Business Unit":{"code":"BU070","label":"IBM Infrastructure"},"Product":{"code":"SWG60","label":"IBM i"},"ARM Category":[{"code":"a8m0z0000000CHAAA2","label":"Operating System"}],"ARM Case Number":"","Platform":[{"code":"PF012","label":"IBM i"}],"Version":"7.1.0;7.2.0;7.3.0;7.4.0;7.5.0"}]

Historical Number

491199748

Document Information

Modified date:
28 February 2025

UID

nas8N1013593

Page Feedback