*** CHECKING THE SYSTEM CONFIGURATION *** I. Klug: 1) chkconfig: check or alter the configuration state 2) hinv: hardware inventory 3) ginv: graphics inventory 4) versions: software inventory 5) xdpyinfo: display information utility for X II. Cryoem: 1) xdpyinfo: display information utility for X. *** SETTING THE NETWORK ADDRESS *** I. Klug: 1) Network addresses information need to be in /etc/hosts: # local host 127.0.0.1 localhost # loghost 128.143.66.28 klug.med.Virginia.EDU klug 2) Use the nvram command to set the variable netaddr : nvram netaddr 128.143.66.28 II. Cryoem: 1) Network addresses information need to be in /etc/hosts: # local host 127.0.0.1 localhost # loghost 128.143.66.29 cryoem.med.Virginia.EDU klug 2) Edit /etc/rc.boot: Change broadcast address to the right router (128.143.66.0) under ifconfig le0 arp netmask 0xFFFFFF00 broadcast "address" *** MAKING NEW FILE SYSTEM *** I. Klug: 1) Log in as root. Shut down the system and turn it off. Install the new hard disk and configure it. 2) After the disk is installed, bring the system up to single-user mode. 3) Verify the disk is formatted and check the existing partitions with the prtvtoc (print volume header information) command. If the disk is not formatted, then use fx command to format it. 4) Use dvhtool (modify volume header information) to partition the drive: dvhtool /dev/dsk/dks0d2vh    dvhtool will respond: Command? (read, vd, pt, write, bootfile, or quit):    Type pt    for ``partitions.'' dvhtool    will respond: Command? (part nblks lstblk type, or l):    Type l to list the partition table. example drive looks like this: current contents:     part n_blks 1st_blk type 0: 32130 4590 efs 1: 81090 36720 raw 6: 2021130 117810 efs 7: 2019600 2158830 efs 8: 4590 0 volhdr 10: 4178430 0 volume Hint: n_blks: number of (512k) blocks 1st_blk: starting block. If we want to change partion 6 to 2000000 blocks, for example, then type this response to the dvhtool prompt: Command? (part nblks 1stblk type, or l) 6 2000000 117810 efs    When done, press to back to the top level menu. Then type write to commit the changes. Hint: This process destroy the existing data (if any) in the partition. 5) Use the mkfs command to make a new file system. For example: mkfs /dev/dsk/dks0d2s7 This example constructs a file system on the second disk (d2) attached to the primary SCSI controller (s0), and uses the entire portion of the disk partition s7. 6) Create the mount point for the device. For example: mkdir / extra   mount /dev/dsk/dks0d2s7 / extra   II. Cryoem: ? *** MOUNT SPECIFIC FILE SYSTEMS AT BOOT-UP *** I. Klug: 1) Edit klug:/etc/fstab: Content as in klug:/etc/fstab --- /dev/root / efs rw,raw=/dev/rroot 0 0 /dev/usr /usr efs rw,raw=/dev/rusr 0 0 /dev/dsk/dks0d2s7 /extra efs rw,noquota 0 0 /dev/dsk/dks0d2s6 /usr/people efs rw,noquota 0 0 cryoem:/home /cryoem nfs rw,soft,bg,timeo=3,retrans=5,rsize=8192, wsize=8192 0 0 Hint: The last two lines should be in one line, this mount cryoem:/home as directory /cryoem in klug. efs: IRIX extent file system. nfs: exported network file system. rw: read and write previlige. soft: return an error if the server does not respond. bg: retry in background if the first attempt fails. timeo: set the NFS timeout to n tenths of a second. retrans: the number of NFS retransmissions. rsize: set the read buffer size to n bytes. wsize: set the write buffer size to n bytes. 2) Edit cryoem:/etc/exports: Content as in cryoem:/etc/exports --- /home -rw=klug.med.virginia.edu,root=klug Hint: root=klug: give root access to the root users from klug. root users are always considered "unknown" by the NFS server, unless they are included in the "root" option. II. Cryoem: 1) Edit cryoem:/etc/fstab the same as above except that all file systems have to be nfs. 2) Edit klug:/etc/exports: Content as in klug:/etc/exports --- /usr/people -rw,nohide,root=cryoem,anon=root cryoem.med.Virginia.EDU /extra -rw,nohide,root=cryoem,anon=root cryoem.med.Virginia.EDU *** TRUSTED REMOTE HOSTS AND USERS *** I. Both: Edit file /etc/host.equiv, put in names of the machines. *** GIVE SUPER USER PRIVILAGE TO OTHER COMPUTERS' "ROOT" USERS *** I. Both: 1) Create file .rhosts in root directory. 2) Put in machine names followed by "root". (e.g. crick.med.Virginia.EDU root) *** PRINTER SETUP *** I. Klug: 1) Run System Manager, and then choose printer type "Generic Poscript" under "Printers --- Add" option. or, Copy /usr/spool/lp/model/gpsinterface to /usr/spool/lp/interface/"NAME". 2) Edit /usr/spool/lp/interface/"NAME" to prevent banner page printing: Change BANNER=1 to BANNER=0 3) Edit /usr/spool/lp/interface/"NAME" to prevent backward printing: Under case "$magic" in *script*|*ascii*|*command*|*English*|*ext*|*iso8859*) Change $send $file $opts 2>/dev/null to $send -u $file $opts 2>/dev/null II. Cryoem: ? *** LPR FROM CRYOEM TO KLUG *** 1) Edit cryoem:/etc/printcap and add printer information: 0|NEC|lp|NEC Silentwriter on SGI Indigo2:\ :lp=:rm=klug.med.virginia.edu:rp=NEC:\ :sd=/var/spool/print/klug: 2) Create the spool directory cryoem:/usr/spool/print/klug for the new printer (e.g. NEC), chmod 775, chown, chgrp lp. Reboot or restart /usr/etc/lpd to get it working. 3) Create two files, "NEC" (e.g.) and "lprtops" in klug:/usr/spool/lp/etc, then chgrp and chown to lp lp. The contents of these two files are identical: #! /bin/sh lp -s -dNEC Hints: -s: supress messages -d: destine 4) Edit klug:/etc/printcap: 0|NEC|lp|NEC SilentWriter 95:\ :lp=/dev/null:of=/usr/spool/lp/etc/NECP:sd=/usr/spool/lpd:sf:sh: Hints: sd: spool directory sf: supress form feed sh: supress printing of burst page header. 5) Edit klug:/etc/hosts.lpd and add the names of all the machines allowed to print to this printer. *** MODEM SETUP *** I. Klug (for dial-in only): 1) Edit /etc/inittab: t2:23:respawn:/usr/lib/uucp/uugetty -Nt60 -ihayes14in, conn ttyf2 dx_38400 # Modem Hints: The above should be in one line. t2: port 2 -hayes14in: script setting in /usr/lib/uucp/Dialers ttyf2: hardware flow control dx_38400: 38400 baud 2) Reset inittab: Execute telinit q 3) Set modem parameters: Execute /usr/lib/uucp/fix-hayes -i -m ACCURA -s 38400 2 Hints: -i: dial-in -m ACCURA: model V.32 -s : speed in baud rate 2: port 2 2) + 3) are put in script "/usr/local/bin/initmodem". This script can also be used to reset the ttyf2 if the line is somewhat trapped. *** TIMESLAVE SETUP *** I. Klug: 1) Kill 'timed' if it's running and turn it off with "/etc/chkconfig timed off". 2) Create file /etc/config/timeslave.options and add the following line: -H galen.med.Virginia.EDU -P /usr/tmp/.timetrim 3) Turn on timeslave with "chkconfig timeslave on" 4) Recycle the network with: /etc/init.d/network stop /etc/init.d/network start II. Cryoem: ? *** REBUILD THE KERNEL *** I. Klug: Use lboot to rebuid the kernel after adding configurable software subsystems or adding software drivers for new hardware devices, or removing software drivers for hardware devices that no longer exist. 1) cd /usr/sysgen, run lboot specifying the new kernel as "unix.install": lboot -u /unix.install    2) Save the original unix kernel (in case): ln /unix /unix.save    3) Reboot --- the system overwrites the current kernel, /unix, with the newly created kernel, /unix.install 4) Recovering from an unbootable kernel --- boot from unix.save: Press the button. Bring out the Command Monitor prompt and enter boot dksc(0,1,0)unix.save Once the system is running, replace the saved kernel as /unix. II. Cryoem: ? 1) Choose a name for your configuration of the system; for example, CRYOEM. 2) Create the config file by making a copy of GENERIC (or other protocol): cp GENERIC CRYOEM; chmod +w CRYOEM 3) Edit CRYOEM to reflect the system configuration. 4) Run config: /etc/config CRYOEM Hint: directory ../CRYOEM will be made and a "make depend" will be done. 5) Make the new system: cd ../CRYOEM make 6) Typically the running kernel should be "/vmunix" because programs like 'ps' and 'w' expect "/vmunix" to be the running kernel. Save the original kernel, install the new one in /vmunix, and try it out: mv /vmunix /vmunix.old cp vmunix /vmunix /etc/halt b vmunix 7) If the system does not appear to work, boot and restore the original kernel, then fix the new kernel: /etc/halt b vmunix.old -s mv /vmunix.old /vmunix ^D [Brings the system up multi-user] *** INSUFFICIENT SPACE ON THE ROOT FILE SYSTEM *** I. Both: The main reason for running out of space on the root partion is application programs creating many, sometimes very large, files in /tmp. The fixes: Make /tmp a symbolic link to /usr/tmp. or, Identify applications that cause the most problems, and configure them to use /usr/tmp instead of /tmp for temporary files. Most applications recognize the TMPDIR environment variable, which specifies the directory to use instead of the default. So, setenv TMPDIR /usr/tmp. *** CHANGE SCREEN GAMMA *** I. Klug: Run gamma "number"; gamma without argument shows current gamma. II. Cryoem: N/A *** KILL HUNG-UP CONSOLE *** I. Klug: 1) Run /etc/killall Xsgi. or, Press left Shift, left Ctrl, / (on keypad), and F12 simutaneously. II. Cryoem: N/A *** CHANGE USER LOGIN ICON *** I. Klug: 1) Make icon size close to 100 x 100 pixels in IRIS "RGB" format. 2) Name the icon login.icon and put in directory .icons (under user's root directory). Hints: Edit /etc/passwd (rearrange the order of users) if you like to re- arrange the login icon line-ups in login screen. II. Cryoem: N/A *** IMPORTANT SYSTEM FILES *** I. Klug (same for cryoem unless indicated): 1) /etc/cshrc: containing the default environment for /bin/csh. Cryoem: N/A 2) /etc/fstab: containing the file systems to be mounted at boot time. 3) /etc/exports: containing the list of file systems exported to NFS clients. 4) /etc/gettydefs: containing information used by getty to set the speed and terminal settings for a line. Cryoem: /etc/gettytab 5) /etc/group: describing each group to the system. 6) /etc/hosts: containing information about the known hosts on the network. 7) /etc/hosts.equiv: containing a list of hosts trusted for rlogin. 8) /etc/inittab: containing the instructions to define the processes created or terminated by init for each initialization state. Cryoem: N/A 9) /etc/passwd: identifying each user to the system. 10) /etc/profile: containing the default environment for /bin/sh   users. Cryoem: N/A 11) /etc/rc0: executed by shutdown that executes shell scripts in /etc/rc0.d to bring the system to run-level 0 (power-down state). Cryoem: /etc/rc.local 12) /etc/rc2: executed by init that executes shell scripts in /etc/rc2.d and /etc/rc.d on transitions to run-level 2 (multiuser mode). Cryoem: /etc/rc.boot 13) /etc/sys_id: containing the system name. Cryoem: N/A 14) /usr/adm/SYSLOG: containing all system and daemon error messages. Cryoem: /usr/adm/messages