addresses;   and   adding   new   application   software. Changes   to   memory,   buffer   size,   and   hardware addresses require changes to the operating system. In addition,   you   must   follow   the   manufacturer’s instructions. Memory and buffer size changes are done when   system   performance   has   dropped   because   of increased use or changes to peripherals. By increasing memory  and  buffer  sizes,  the  throughput  time  of  the system   is   improved   and   the   user   access   time   is shortened.  Hardware  addresses  are  changed  when  there is a change to hardware status, such as the addition or loss  of  a  piece  of  equipment.  The  addition  of  new application   software   will   cause   the   system   to   be unavailable to users during the implementation. When the  software  arrives,  it  will  include  the  installation instructions that are to be followed. Powering Up the Computer The CPU is normally powered up first; then one by one, each of the I/O devices is turned on. This is done to prevent a sudden surge of power that might overload the circuits.  With  the  computer  system  powered  up,  you can then proceed to IPL the system. Booting the System Anytime the system loses power, is shut down, or becomes disabled because of a system crash, it usually needs  to  be  rebooted  or  initial  program  loaded  (IPLed). A system crash is the result of a hardware, software, or operation   problem: a   malfunction   in   the   CPU,   a programming  error  from  which  the  operating  system could  not  recover,  or  an  operator  error  caused  by  an incorrect response to a message. Booting most systems resets  all  status  indicators  and  reloads  the  supervisor (the   executive-system   program   along   with   other resident routines) into the CPU memory. The manner in which  the  system  is  booted  depends  upon  the  computer system used and the software included in its operating system. Many of the larger mainframe computers store their operating systems on disk, and this disk is referred to as the SYStem RESident (SYSRES) pack. Once the disk unit with the SYSRES pack is in a ready status, you can then boot the system. Some systems are so simple to boot that all you need do is depress the start (or load) button on the CPU (or master console) and enter the date and time on the console keyboard. Some of the more complex  systems  may  require  you  to  take  additional steps—assigning   various   I/O   devices,   partitioning (sectioning  off)  memory,  and  so  on.  It  is  because  of these   differences   that   boot   procedures   are   well documented with each step explained to the point that anyone,  even  someone  unfamiliar  with  the  system, could follow the step-by-step directions. Computer Diagnostics Computer diagnostics are of two types: those that are run when the computer is booted (called  internal diagnostics),  and those that can be run from a peripheral device (called external diagnostics). INTERNAL   DIAGNOSTICS.—   Internal diagnostics  are  performed  when  the  computer  is  booted and are designed to test the computer hardware. They return  a  pass/fail  result  to  the  operator.  Some  of  the internal diagnostics are designed to test all or selected sections  of  the  computer.  An  example  of  an  internal diagnostic  test  is  the  power  on-self  test  (POST),  that checks the components upon power on. EXTERNAL   DIAGNOSTICS.—   External diagnostics  can  be  loaded  into  the  computer  from  a peripheral  (disk  or  magnetic  tape  unit)  but  initiated  and controlled by the computer. These diagnostics are very thorough and offer the option of testing all or specific sections of a computer. They are more comprehensive than the internal diagnostics. RESPONDING  TO  MESSAGES Unless   you   respond   correctly   to   a   system-   or program-generated message, CPU time can be lost, the wrong program options can be executed, and the system can even become disabled. As a console operator, you are the most important part of a team effort that makes computerized procedures successful. It is true that while complex  operating  systems  make  it  easier  for  you  to perform  various  tasks,  correct  responses  to  messages are  just  as  critical. Never  second-guess  a  response. Always  refer  to  the  proper  documentation  to  determine the   response   needed. Manuals   provided   by   the manufacturer,  operator’s  manuals,  and  error  message manuals  should  be  located  as  near  to  the  console  as possible for your convenience. Basically, you will be concerned with two types of messages:  system-generated  messages  and  program- generated messages. System-Generated Messages System  messages  are  generated  by  the  operating system  software. To   be   more   specific,   it   is   the executive-program (the supervisor) that is responsible for  monitoring  and  allocating  the  resources  of  the 2-4