Shell & OS
ears890 is a CIO bus risc 1.0 … 9.0
/system/UX-CORE/index - has os level B9.041K
before reboot
/usr/local/bin/block_intellect
uv.rc stop
after bootup
uv.rc start
dialer_info
cd /usr/local/bin; ./x29pr_mstr
rebuild swap space by sam and uxgen
size <executable> - returns the size of code, data, uninitialized data (bbs) for a program. Combine the data and uninitialized data to see the amount of swap space reserved for that process
booting in single user mode
hpuv -is boot disk(scsi 6;0x0) /hp-ux
rebuilding kernel
config dfile
make -f config.mk
make /hp-ux kernel
mailx users /usr/lib/mailx/mailx.rc
xmodmap -pm XHPKeymaps /etc/newconfig/X11R5/rgb.txt.A10 - color names
default gateway
/etc/gateway
net 0.0.0.0 gateway x.x.x.x metric 1 passive
automount files
/etc/automaster /etc/auto_home
File Systems
ioscan -f # produces a list of all devices
/etc/lvmpvg - vg to disk map
lvreduce -m0 /dev/vg00/lv001 - undo mirroring
pvmove -n /dev/vg00/lvol4 /dev/dsk/c0d0s2 /dev/dsk/c7d0s1 - moves data from one disk to another
/etc/vgscan recreated the /etc/lvmtab
strict allocation - mirror is on separate physical volume
parallel scheduling - write data at same time (faster)
sequential scheduling - writes data to 1st disk, if successful then the 2nd disk
mirror consistency recovery - longer recovery time because of write buffer cache at time of crash needs to be written out.
Vgexport - vgimport
must use newfs
calculate space from vgdisplay
lvcreate -L (size) -m1 vg00 or use
lvcreate -l (logical extends) -m1 vg00
diskinfo - to get product id
newfs /dev/vg00/rlvol3 ProductId
pvcreate or pvextend to add disk to volume groups
pvcreate -f /dev/rdsk/c0d0s2
mkdir vg01
mknod /dev/vg01/group c 64 0x0100000 (the hex number represents the volume group number
use lvlnboot to update / when moving root
vgreduce to remove a physical volume from a volume group
vgreduce vg01 /dev/dsk/c2d0s2
lvremove -f /dev/vg01/lvol1
vgremove /dev/vg00
MirrorDisk/UX or DataPair
Mirror Write Cache (use -M) - keeps track of PE not written to mirror in case of crash - a hit on performance.
Mirror Consistency Recovery (use -c) - upon crash reboot, checks every LE with mirror - hit on recovery / boot up time after crash
use lvchange to remove or add these options
vgcfgbackup - updates
lvmrc
uxgen - generates all sys io devices
vgscan - rebuilds lvmtab file
mkdev, ls dev, lssf, lifls, mksf, insf
insf -H 12/36.11.0 -e - to create new devices
lssf produces no vg 20 or 21 or 26
Security
/usr/adm/inetd.sec # network services allowed by network
time zone
set in /etc/rc /etc/profile or /etc/csh-login
TZ=EST5EDT
setuid files
find / -user root -perm 4000 -print
last # shows when user logged on last
/usr/adm/sulog or messages
System Monitoring
hp support 109k per year
context switching is when a process wants resources that is taken up by a large amount of other processes. The cpu ends up being used more for swaping than processesing the program
report system swaping
sar -w
swapinfo -ta # shows all swap devices
report block devices
sar -d
dumps
look for tmpsyscore
savecore -t tapedev dumpdir
/etc/savecore / - used to dump
scancore corefile hp-ux > file where hp-ux is the kernel
display shared memory
ipcs -b display shared memory -sed
swchunk def = 2MB 2048 x def blocksize of 1024
uv = 17020 + 9068 + 444 = 26532
shared mem seg = 4096 for uv
shmseg - the number of shared memory segments attached to a process used when 1 multiple processes share data
ps listing headers
F - type of process
0 swapped
1 in core
2 system process
4 locked in core
S - state of process
0 non existent, Sleeping, Waiting, Running, Intermediate, Z terminated, T stopped, X growing
C - processor utilization for scheduling
PRI - priority, the higher number equals lowest priority
NI - nice value
ADDR - memory or disk address
SZ - size of core image of process
WCHAN - event for which process is waiting
STIME - starting time of process by fork not exec
TIME - execution time
use top monitor sar vmstat iostat glance
hp hardware problems
/var/adm/diag - scsi bus errors - use sysdiag/logtool to view info
sysdiag
logtool
status
list log=0139
tombstone log
system consol
<cntrl> b
sp
dl display tombstone log
co console
er enable remote access
sp hard reboot
Universe/Cubs
login -> ksh -> .profile -> exec /usr/local/bin/.startup ->
Universe 7.3.3.1 can’t run on hp-ux 10.x
Universe 8.3.2.2 R8 can’t run on hp-ux 9.04