April 1, 1981
Messages on Startup and Initialization In the course of VM/370 initialization or startup, in addition to the
normal expected messages that are a part of loading the system, you may
see other messages to which you must respond. These messages relate to
functions of initializat-lon: checkpoint .• warm start, and automatic
checkpoint requirements. These messages and the recommended course of
action are described in the VM/370 System Messages. Console Messages After Startup System console messages are in three categories: • System operation status messages • Virtual machine user messages • Hardware and program status messages
The operator action required depends on the type and content of the
message. SYSTEM OPERATION STATUS MESSAGES These messages inform the system operator of the logging on and logging
off of users, number of spooled files, output printer or punch
accounting data, device attention conditions, and command completion
indications. Some of these messages require operator action, while
others do not. MESSAGES FROM VIRTUAL MACHINE USERS These messages include requests and inquiries from virtual machine users
on the system; a sampling of these user messages follows: MSG FROM JONES: CAN YOU GIVE ME A TAPE MSG FROM SMITH: CAN YOU PUT UP A SCRATCH 181 MSG FROM SMITH: PLEASE TAKE RING OUT OF 181 NOW MSG FROM BROWN: PLS ATTACH 382 HARDWARE AND PROGRAM STATUS MESSAGES These messages include VM/370 error recording messages, VM/370 system
error information, and I/O error messages related to users of virtual
machines who have invoked SVC 76. SVC 76 causes VK/370 to send a
message to the system operator (for I/O error records only) and to
record the error data in VK/370 "s error recording area. The V!/310 manual describes these messages.
For more information about error detection and the recording and
editing of errors, refer to the !M/370 OL!SEP and Error The following the action that the operator must take in
response to an error message:
1. Determine the severity of the error.
24 VM/370 Operator's Guide

2. Take a storage dump if the automatic dump routine failed.
3. Try to preserve spooling and accounting data by issuing the SHUTDOWN command if it has not been done automatically.
4. Try to force error recording if it did not occur.
5. Attempt to analyze the problem by whatever means at his disposal
and if necessary, reallocate resources and priorities.
6. Call for technical assistance if necessary. Whenever there are indications that I/O errors require further
analysis or that basic integrity of a device needs to be determined,
other service programs may be invoked to help make such an assessment.
For example, to test the operational 3344 or 3350 DASD devices, the
operator may invoke the OS/VS and DOS/VS Analysis Prograa-1 (AP-1).
This utility program, if resident in the OS/VS or DOS/VS library of a
virtual machine, can be used to check the Froper operation of the
device. The use of this program along with error messages and user
action is detailed in the gnd (AP-l) In the event that technical assistance is required, it can be
obtained from the following sources: • The customer's own system programmer(s). • The service representative, who has special test equipment for
hardware problem isolation. e The IBM Program Support Bepresentative
6 who can also assist with VK/370 system problem diagnosis. You can assist with problem diagnosis by being alert for the
following error recording messages: DKKIOG550E I/O ERROR RECORDING AREA 90 PERCENT FULL; RUN CPEREP DKKIOG551E I/O ERROR RECORDING AREI FULL; RUN CPEREP You should then duap the error recording data to some other device
(usually a tape drive) and clear the error recording area to insure
enough space to reco=d additional error records. Information on how to
run CPEREP is contained in the gDg RecQrdiBg Additional information on OS/VS EREP, the program to which CPEREP passes
control is detailed in the Re£Q£ginq !ng E£in!ing publication. Other errors may not be severe enough to cause V8/370 system failure,
but might cause a specific virtual machine to abnormally terminate. In most cases, the user can IPt the virtual machine again. Should the V8/370 aachine enter an unexpected WAIT state or loop, see
the §y!de referring to the section entitled
"How To Use the VK/370 Facilities To Debug." Section 2. Initiating V8/370 Operation 25