.CHAPTER Miscellaneous Topics .HL 1 Accounting System Accounting for user consumption of system resources is performed automatically by the system. The monthly job accounting run, 600M, cycles the accounting file at the end of the accounting month (on the 25th). .HL 1 Answering Outstanding Console Messages Messages appear at the console are of 2 types: informational and outstanding. Informational messages have no request number associated with them and do not need to be answered. When there is an outstanding message at the console, you must first log in. Then, note the request number of the message to be answered and reply using that request number. As an example, assume the following message is appearing at the console: .nf .s 2 %%%%%%%%%%% OPCOM 26-MAY-1987 21:11:06.46 %%%%%%%%%%% Request 45, from user IRVIN .no flags accept BATCH_1159, BACKUP errors - check output and restart .flags accept .s 2 .f User Irvin must have started the batch job that has entry number 1159 which generated this message. The request number is 45. After logging in at a terminal (if you aren't already), enter the following at the prompt: .s .c $ ^&REPLY/TO=45\& .s Occassionally, a program will request a message text to be included with the reply. For example, under certain conditions Backup will issue a message like the following: .nf .s 2 %%%%%%%%%%% OPCOM 26-MAY-1987 21:11:06.46 %%%%%%%%%%% Request 45, from user IRVIN %BACKUP-I-SPECIFY, specify option (QUIT, CONTINUE, OR RESTART) .s 2 .f Use the same "Reply" command but include the message text after the message number: .s .c $ ^&REPLY/TO=45_ "QUIT"\& .HL 1 Adding a Disk Drive to the System Adding a disk drive can be done without rebooting the system and need not take more than 10-15 minutes. In the following steps, assume you are adding device DUA2: to DUA1: in order to create an Oracle volume set. .ls .le;Sign in as the system and get all other users logged off. .le;If you are adding to or creating a volume set, shut down all processes that might be using the other members of the set. Do a "SHOW DEVICE" for the device you are adding to (e. g., "SH DEV DUA1:") and make sure the "Trans Count" field is 1. If not, do a "SHOW DEVICE/FILES" on that device (e. g., "SH DEV/FILES DUA1:") to find out which processes have files open on that device. Get them to release those files if possible, or else just log them out. .le;If you are working with a volume set, dismount all currently mounted volumes of that set: .s .c $ ^&DISMOUNT/NOUNLOAD_ DUA1:\& .le;Allocate the new device: .s .c $ ^&ALLOCATE_ DUA2:\& .le;Initialize the new device: .s .c $ ^&INITIALIZE_ DUA2:_ ORACLE2\& .le;Mount the device: .s .c $ ^&MOUNT/SYSTEM/BIND=ORACLE_DB_ DUA1:,DUA2:_ ORACLE1,ORACLE2\& .le;Make the appropriate modification to the SYS$MANAGER:SYSTARTUP.COM file: .s .c ^&$_ MOUNT/SYSTEM_ DUA1:,DUA2:_ ORACLE1,ORACLE2\& .le;Finally, make the appropriate modification to the MED$BACKUP:BACKUP.COM file, if necessary (for volume sets, subsequent volumes beyond the root volume are automatically backed up if the root volume is specified). .els .HL 1 Booting the 8550 System The system has several boot flags that enable auto reboot if recovering from a power failure or other traumatic event. In the event that you want to boot the system from scratch (e. g., after a protracted power failure), follow the following steps: .hl 2 Power the Equipment Up If all of the equipment has been powered off, perform this step. .LS .LE;Power up the CPU. The main power switch is located inside the front CPU panel, under the backplane, on the lower right-hand side. .LE;This should have powered up the console. If it does not, contact Digital field service. The prompt now should be ">>>". Type in .nf .s .c >>> ^&POWER_ ON\& .s .f .LE;Power up the tape drive. The switch is located inside the hood at the upper right-hand corner of the drive. .LE;Power up the disk drives. Press the "RUN/STOP" switches. .els .hl 2 Boot the Console If the console is not on, turn it on and it should boot. (If it does not, contact Digital field service.) If the P/OS Main Menu comes up, move the arrow to the "Console/PRODCL" choice and press the DO key on the keyboard. At this point, the prompt now should be ">>>". .hl 2 Connect the Console to the VAX Type in .nf .s >>> ^&EXIT\& .c $ ^&RUN_ CONTROL\& .s .f .hl 2 Boot the VAX Again, the prompt should be ">>>". Type in .s .c >>> ^&B\& .s The system will be up in about 5 minutes. .HL 1 Booting the 63xx System Antonio will write this section. .HL 1 Booting the HSC 70 Follow the procedure below to boot the HSC70: .s .LS .LE;Open the HSC front door and ensure that the HSC software diskette is in the disk drive 0 (the left one) and the drive handle is in the close position. .LE;Ensure that the power switch is ON (the power switch is located on the left side of diskette housing). .LE;Ensure that the Enable/Secure switch is in the Enable position (this switch is located on the back of the front door). .LE;Now check the front control panel. The ONLINE switch should be depressed; the INIT and the FAULT should not. The STATE indicator ligth (yellow) is OFF during initial boot, ON during self test, and blinking during load of the operating system and at all other times. The POWER indicator ligth (green) should be ON at all times. .LE;Push and release the INIT switch on the front control panel. Notice the INIT and STATE indicators show the different phases the HSC progresses through while loading the operational software. At load complition, the following message is displayed at the terminal: .nf .s HSC Version Vnnn (date time) Ma,e (System Name) .s .f .LE;The system software is now running. Place the Secure/Enable switch in the Secure position. This prevents accidental operation of the INIT wsitch. .LE;Close the front door. .hl 1 CPME System There are 3 system jobs that should be started during boots and are stopped and started during backups. The job names are MONITOR, DISKMON, and MONEXTR. The first 2 collect data about the cpu, disks, and memory; the last one extracts the data and write them into a "mailbox" file. This mailbox is checked when 538J runs. This procedure is started at the end of backup and writes an output tape that is read by 538J on the Sperry. .s If problems occur with any of these jobs, please contact Tech Services. .HL 1 Fiching Production Output All output to be fiched resides in the directory MED$PRD. The previous period's extract data also resides there. When job 185Q runs, it first deletes the previous period's data. It then writes the current output to tape and flags these fiched files as already extracted. .s If problems occur with this job, please contact Tech Services. .HL 1 Halting the 8550 System .hl 2 Warn the users Aside from the Tech Services and Operator signons, no one is authorized to be on the system during normal backup hours. (Normal backup hours are 9 p.m. - midnight, Sunday through Thursday, and 9 p.m. - 7 a.m., Friday/Saturday and Saturday/Sunday.) It is still a good idea to make sure no one is logged on when the system comes down. First, prevent new users from logging on with .s .c $ ^&SET_ LOGINS/INTERACTIVE=0\& .s Make sure those currently logged on get warned. For example, .s .c $ ^&SH_ USERS\& .s displays all of the current users on the system. You may want to warn them of what is about to happen with .s .c $ ^&REPLY/ALL/BELL_ "The_ VAX_ is_ coming_ down_ now."\& .s Stubborn users may be removed by stopping them. (See the "Killing Processes" section.) You should also check all of the queues to make sure that nothing is printing. If something is printing and you cannot wait for it to complete, requeue it. (See "Requeueing a Print Job" in the "Batch and Print Topics" chapter.) .hl 2 Shutdown running processes Once you are comfortable with the current state of the machine, type .s .c $ ^&SHUTDOWN\& .s Answer the appropriate queries, taking all defaults (i.e., keep hitting the carriage return). .hl 2 Halt the CPU When the following message appears: .s .c SYSTEM SHUTDOWN COMPLETE - USE CONSOLE TO HALT SYSTEM .s enter .nf .s .c ^&\& >>> &H .s .f The CPU is now halted. .hl 2 Power the Equipment Down If you want all of the equipment to be powered off, perform this step. .LS .LE;Power off the tape drive. The switch is located inside the hood at the upper right-hand corner of the drive. .LE;Power off the disk drives. Press the "RUN/STOP" switches. .LE;At the console, type in .nf .s .c >>> ^&POWER_ OFF\& .s .f This will turn power off to the CPU. If you choose to turn power off to the CPU manually, the main power switch is located inside the front CPU panel, under the backplane, on the lower right-hand side. .le;Turn off the console by pressing the on/off switch located by the diskette drives. .els .HL 1 Halting the 63xx System Antonio will write this section. .HL 1 Halting the HSC 70 Antonio will write this section. .hl 1 Hardware Troubleshooting If you suspect hardware problems, the Digital Diagnostic Center (DDC) will want to call the system and perform tests. Use this procedure to contact them: .list .le;Check the console for possible problems (it may say "machine check", "bug check", or "system hung"). The system may not display an error. .le;Call 1-800-525-6570. .le;DDC might ask for the CPU serial number (87040300H), our CPU model (8530), the current level of VMS (5.0-2), and the number to call back ((612) 574-4648). Describe the symptoms as best as you can. .le;DDC will use the "FIELD" account. This account is normally disabled. It will need to be enabled by using the "MAINT" utility (select the modifying a user's password option). Communicate the password to DDC. .le;Turn the remote services console (which is next to the console display) to "Remote". DDC will now walk you through the rest of the procedure. .le;When DDC is done, a message at the console will tell you about a "remote connection broken". Turn the remote services console back to "Lock out". .le;Disable the "FIELD" account as follows: .nf .s [RUCKER] $ ^&AUTH\& .c UAF> ^&MODIFY_ FIELD/FLAG=DISUSER\& UAF> ^&EXIT\& .f .els .HL 1 Killing Processes This is done when you wish to log a user out, for example, when the system will be shutting down. First, locate the process identifier (PID) by entering a "SHOW SYSTEM" command. (To log users out, you may enter "SHOW USERS" instead.) Once the PID is known, give your process WORLD privilege. Then enter "STOP/ID=", where the leading zeroes of the PID may be omitted. Finally, give up WORLD privilege. .hl 2 Example Assume that GANZB's session is hung and the user has requested that the process be killed. .nf .s [RUCKER] $ ^&SHOW_ USERS\& VAX/VMS Interactive Users 29-DEC-1988 08:28:00.60 Total number of interactive users = 10 Username Process Name PID Terminal GANZB GANZB 0000163D LTA347: GERMSV GERMSV 00001C18 LTA341: HOLTEK HOLTEK 00001B17 LTA340: . . . [RUCKER] $ ^&SET_ PROCESS/PRIV=WORLD\& [RUCKER] $ ^&STOP/ID=163D\& [RUCKER] $ ^&SET_ PROCESS/PRIV=NOWORLD\& .f .HL 1 Setting the System Date and Time If the system has the wrong date and/or time, this may be changed without the necessity of rebooting the system. To set it, use the following example as a guide: .s .c $ ^&SET_ TIME=25-MAR-1988:10:08:00\&