Looking for default (factory delivered) copies of configuration files in HPUX:
/usr/newconfig .... subdirs underneath
My random tech problems and chicken scratching for notes on what I did to solve them.
Looking for default (factory delivered) copies of configuration files in HPUX:
/usr/newconfig .... subdirs underneath
create a service type user account in AD. In the case below the username is: kermit
Run the ktpass command on a domain controller using the syntax below.
Copy the krb5.keytab file to the HPUX box under /etc/krb5.keytab
Edit the /etc/krb5.conf file, pointing to the correct domain and domain controller.
ktpass utility is included in the Windows 2000 support tools.
On domain controller:
E:\>ktpass -princ host/amisys@MYDOMAIN.COM -mapuser kermit -pass ***** -out krb5.keytab
Successfully mapped host/mybox to kermit.
Key created.
Output keytab to krb5.keytab:
Keytab version: 0x502
keysize 49 host/mybox@MYDOMAIN.COM ptype 1 (KRB5_NT_PRINCIPAL) vno 1 etype 0x1 (D
ES-CBC-CRC) keylength 8 (0x10ade54f832551ba)
Account has been set for DES-only encryption.
copied krb5.keytab to /etc/krb5.keytab and set owner: root:sys and permissions: 600
copied pam.krb5 to /etc/pam.conf
Setup Kerberos config file:
[libdefaults]
default_realm = MYDOMAIN.COM
default_tgs_enctypes = DES-CBC-CRC
default_tkt_enctypes = DES-CBC-CRC
ccache_type = 2
[realms]
MYDOMAIN.COM = {
kdc = camelot.MYDOMAIN.COM:88
kpasswd_server = camelot.MYDOMAIN.COM:464
}
[domain_realm]
.mydomain.com = MYDOMAIN.COM
[logging]
kdc = FILE:/var/log/krb5kdc.log
admin_server = FILE:/var/log/kadmin.log
default = FILE:/var/log/krb5lib.log
chmod 644 /etc/krb5.conf
Managing SAN disk array subsystem:
NOTE: must run the following variable set in your telnet session in order to display X apps on your PC: export DISPLAY=myhostname:0 where 'myhostname' is your PC's name. You also must have the X client manager running on your PC.
Command View SDM (Storage Device Management) – software used to manage the array
VAnum1 – the 'alias' name of the Virtual Array disk system
directory of binaries: /opt/sanmgr/commandview/client/sbin
launcher – X client for managing the disk and creating LUNs. On LUN Management tab, Logical LUNs, you can create LUNs.
define LUN: This term is an acronym for Logical Unit. It is used as the primary means for dividing the entire capacity of the array into smaller, individually addressable, units. When created, contains no data. Read operations directed to this logical drive will return the format pattern until writes have been performed on the same locations.
cvui – command line menu based interface for managing the array. It's nice in the fact that it displays the command being run when you select a menu item. You can also log the commands and their results to a file if you wish.
cmdviewVA VAnum1 – Same as launcher, but skips first dialog listing the VA selection screen.
Then run sam, go under "Disks and File Systems/Disk Devices" to configure the volume group, logical volume(s), file system, etc.
listing free and used space on disks:
list kbytes of disk usage; use without 's' option to display all the files
du -sk /directory/path
pretty listing of total space allocated, used and free.
df -k /filesystem_mount
df -k /u* -- shows all /u00-u12 files system stats
Startup scripts:
/sbin/init.d/oracle
/sbin/init.d/sqlnet
/sbin/init.d/agentctl*
/sbin/init.d/oemctl*
(other similiary named files are in: /etc/rc.config.d)
Connecting to database with sysdba privledges from sqlplus command:
sqlplus "/ as sysdba"
Check if database started with a pfile or spfile:
in sqlplus:
SELECT DECODE(value, NULL, 'PFILE', 'SPFILE') "Init File Type"
FROM sys.v_$parameter WHERE name = 'spfile';
Check if database is in archive log mode:
in sqlplus do: archive log list
Converting from pfile to spfile or vice-verse:
CREATE PFILE FROM SPFILE;
CREATE SPFILE FROM PFILE;
Creating password file:
orapwd file=$ORACLE_HOME/dbs/orapw{SID} password=sys entries=1
Archive tips: http://www.cryer.co.uk/brian/oracle/SYMoixcal.htm
Archive log:
Must be in init{SID}.ora or spfile for archive log mode to be useful: LOG_ARCHIVE_START = true
alter system archive log start
alter system archive log stop
alter system switch logfile
select log_mode from v$database;
ALTER DATABASE ARCHIVELOG;
ALTER DATABASE NOARCHIVELOG;
Alert logs stored in: /u02/${ORACLE_SID{/bdump/alert_${ORACLE_SID}.log
Get some info on memory:
sqlplus> select * from V$SGASTAT;
Oracle management server startup:
NOTE: if changing username/password for the OEM database, YOU MUST edit oracle's .profile to reflect this change.
first start repository database:
export ORACLE_SID=OEMREP
sqlplus /nolog
connect / as sysdba
startup
exit
Then start the OMS server:
oemctl start oms
oemctl ping oms - shows status without having to issue username/password
oemctl stop oms sysman/{sysman's_password}
'oemctl start oms' output should be similar to:
mybox Oracle preprod1> oemctl start oms
OEMCTL for HP-UX: Version 9.2.0.1.0 Production
Copyright (c) 1998, 2002, Oracle Corporation. All rights reserved.
Starting the Oracle Management Server....
The OMS is now started and ready.
How does one start the Oracle Intelligent Agent?
One needs to start an OIA (Oracle Intelligent Agent) process on all machines that will to be managed via OEM. The agent needs to be stopped/started when a new database is created on that node. Databases that are removed will have to be deleted manually from the OEM console.
For OEM 9i and above:
agentctl start agent
agentctl stop agent
agentctl restart -- This doesn't seem to work.....
$ agentctl
Usage:
agentctl start|stop|status|restart [agent]
agentctl start|stop|status blackout [<target>]
[-d/uration <timefmt>] [-s/ubsystem <subsystems>]
The following are valid options for blackouts
<target> name of the target. Defaults to node target.
<timefmt> is specified as [days] hh:mm
<subsystem> is specified as [jobs events collections]
defaults to all subsystems
dbca - database configuration assistant X-gui app.
or
Create an init file and proper directory structures for values in init file. Make sure remote_login_password=NONE
Change /etc/tnsnames.ora and /etc/listener.ora to contain new database values. Reload the listener: lsnrctl, the reload.
set ORACLE_HOME and ORACLE_SID to proper values for new database.
run sqlplus and the following (/home/oracle/createdb.sql:
create database soup
logfile group 1 ('/ora/soup/redo01.log') SIZE 100M,
group 2 ('/ora/soup/redo02.log') SIZE 100M,
group 3 ('/ora/soup/redo03.log') size 100M
datafile '/ora/soup/system.dbf' SIZE 575M
character set WE8ISO8859P1
national character set utf8
extent management local
undo tablespace UNDOTBS1
datafile '/ora/soup/undotbs01.dbf'
SIZE 2000M
default temporary tablespace temp
tempfile '/ora/soup/temp01.dbf'
size 4000M autoextend on next 50M maxsize 7000M;
Starting EM (enterprise manager) from unix (x-gui):
oemapp console
OMS navigator needs to have databases deleted from it when the database is phyisically deleted. Refreshing the node won't remove the database from the navigator. In fact, it's best to delete the node from the navigator and re-add it again.
Stop Apache web server:
$ORACLE_HOME/Apache/Apache/bin/apachectl stop
SQL> select * from v$version;
BANNER
----------------------------------------------------------------
Oracle9i Enterprise Edition Release 9.2.0.1.0 - 64bit Production
PL/SQL Release 9.2.0.1.0 - Production
CORE 9.2.0.1.0 Production
TNS for HPUX: Version 9.2.0.1.0 - Production
NLSRTL Version 9.2.0.1.0 - Production
Removing Oracle from a test server (sun box):
rm -r $ORACLE_HOME
rm -r /var/opt/oracle
Dropping a tempfile.
> sqlplus /nolog
sql> connect as sysdba
sql> alter database tempfile '/u08/mydb1/tempdave.dbf' drop;
-database altered
exit
drop tablespace logmnrts including contents;
CREATE
TEMPORARY TABLESPACE "TEMPNEW" TEMPFILE '/u09/mydb1/
TEMPNEW1.dbf' SIZE 5000M AUTOEXTEND
ON NEXT 640K MAXSIZE 7600M EXTENT MANAGEMENT LOCAL UNIFORM
SIZE 1024K;
ALTER DATABASE DEFAULT TEMPORARY TABLESPACE "TEMPNEW"
THEN
CREATE
TEMPORARY TABLESPACE "TEMP" TEMPFILE '/u07/mydb1/
TEMP01.dbf' SIZE 2500M REUSE AUTOEXTEND
ON NEXT 640K MAXSIZE 7600M EXTENT MANAGEMENT LOCAL UNIFORM
SIZE 1024K;
ALTER DATABASE DEFAULT TEMPORARY TABLESPACE "TEMP"
Recovery Manager User's Guide = a96566.pdf
Quick Start to Recovery Manager: page 75
Duplicating a Database with Recovery Manager: page 365
Recovery Manager Reference = A96565.pdf
Duplicating DB / random RMAN thoughts:
get a good RMAN backup of Production (takes about 10 MINUTES):
. rman target /
r shutdown immediate;
r startup mount;
r backup database;
r alter database open;
r quit
Then duplicate this backup to the auxiliary instance:
Target database must be open; auxiliary database must be started, in nomount mode.
When RMAN duplicate is finished the auxiliary database is opened.
Duplicate takes about 15 minutes.
Before running duplicate command, get the tempdata file name:
$ find /u??/mydb1/ -name "temp*"
. sqlplus "sys/null@mydb1 as sysdba"
s shutdown normal;
s startup nomount;
s quit
. rman target / auxiliary sys/null@mydb1
r duplicate target database to mydb1;
r quit
Re-add any datafiles that belong to the temp tablespace, and finish up other tasks:
. sqlplus "sys/null@mydb1 as sysdba"
s ALTER TABLESPACE TEMP
s ADD TEMPFILE '/ora/soup/temp01.dbf' SIZE 4000M REUSE;
in the auxilary init{sid}.ora and it's spfile:
DB_FILE_NAME_CONVERT=('/u00/proddb1/','/u00/mydb1/',
'/u01/proddb1/','/u01/mydb1/',
'/u03/proddb1/system01.dbf','/u05/mydb1/system01.dbf',
'/u09/proddb1/undotbs01.dbf','/u07/mydb1/undotbs01.dbf',
'/u07/proddb1/myapp_data01.dbf','/u05/mydb1/myapp_data01.dbf',
'/u08/proddb1/indx01.dbf','/u06/mydb1/indx01.dbf',
'/u03/proddb1/myapp_data02.dbf','/u02/mydb1/myapp_data02.dbf',
'/u09/proddb1/indx02.dbf','/u10/mydb1/indx02.dbf',
'/u12/proddb1/core_repository.dbf','/u12/mydb1/core_repository.dbf',
'/u06/proddb1/myapp_data02.dbf','/u05/mydb1/myapp_data02.dbf',
'/u12/proddb1/myapp_data02.dbf','/u11/mydb1/myapp_data02.dbf')
LOG_FILE_NAME_CONVERT=('/u00/proddb1/','/u00/mydb1/',
'/u01/proddb1/','/u01/mydb1/',
'/u02/proddb1/','/u02/mydb1/',
'/u03/proddb1/','/u03/mydb1/',
'/u04/proddb1/','/u04/mydb1/',
'/u05/proddb1/','/u05/mydb1/',
'/u06/proddb1/','/u06/mydb1/',
'/u07/proddb1/','/u07/mydb1/',
'/u08/proddb1/','/u08/mydb1/',
'/u09/proddb1/','/u09/mydb1/',
'/u10/proddb1/','/u10/mydb1/',
'/u11/proddb1/','/u11/mydb1/',
'/u12/proddb1/','/u12/mydb1/')
CREATE SPFILE FROM PFILE='/u01/app/oracle/admin/DUP/pfile/init.ora';
recreate tempfile in temp tablespace.
delete tempfile in OS first.
/u02/mydb1/temp01.dbf
ALTER TABLESPACE "TEMP"
ADD TEMPFILE '/u02/mydb1/temp01.dbf' SIZE 4000M REUSE;
MeasureWare (aka HP OpenView Performance Agent (OVPA)) – MWA (agent)
(NOTE The name MeasureWare Agent (MWA) has been replaced with HP
OpenView Performance Agent (OV Performance Agent or OVPA) for
UNIX and the name PerfView for UNIX has been replaced with OV
Performance Manager for UNIX throughout this documentation.
However, the process names and software components operationally
remain MeasureWare Agent (MWA) and PerfView.)
/opt/perf/examples/ - example config files
/opt/perf/bin/perfstat – checks status of agent processes
/var/opt/perf/datafiles – raw binary data files kept here
/var/opt/perf - config files kept here
alarmdef – alarm definition config file
parm - overall parameters config file
/opt/perf/bin/utility – outputs data from logfiles to stdout or files
/opt/perf/bin/extract – makes copies of raw binary data files for archival purposes or exports data from log files to standard format (csv, etc.)
mwa stop servers - stop mwa servers: alarm server, scopeux, etc. -- needed if you see a defunct process continually coming up: possible alarmdef file setting screwed up or corrupt.
Usage = mwa <action> <subsys> <parms>
<action> = status, start, stop, restart, demo, version
<subsys> = scope, server, alarm, db, desktop, all
<parms> = -midaemon <miparms>
See man midaemon for <miparms>
<action> and <subsys> may be entered in any order,
but if <parms> is supplied, then it must be entered last
Defaults: <action>=start, <subsys>=all, <parms>=""
note: mwa stop alarm is not valid.
mwa stop server will stop alarms and servers.
GlancePlus – glance or (x) gpm
top – display top processes
uptime - system load, users and time since last boot
PRM – Process Resource Management: assigns levels of CPU/MEM/DISK/Application resources to particular users or processes. Sounds kind-of cool
Add/remove printers by using SAM.
Optionally you may add jetdirect connected printers via: /opt/hpnpl/admin/hppi
Directories:
/var/spool/lp/request/{printername} = spool area for print jobs
disable printername = disable lp subsystem from sending data to the printer
enable printername = enable lp subsystem to send data to the printer
(the above commands are like using the "pause printing" function in windows)
Start/stop users from sending print jobs to the queue.
accept printername = allow the lp spooler to accept requests for the printer
reject printername = do NOT allow the lp spooler to accept requests for the printer
(the above commands are similar to turning on/off sharing on a printer in windows)
lpstat –t = display all printers status and output jobs (full report)
lpstat -pprintername = displays the specified printer's status
lpstat -s = print summary status for all printers
lpstat -o = list all output jobs on all printers
lpstat -oprintername = list all output jobs on the specified printers
lpstat -d Lists the system default printer
lpstat (no options) Lists the user's outstanding printer requests
man net_ljx000 = displays man page for net_ljx000 model script which display options the printer can accept.
lp –d printername filename = prints filename to the printer named: printername
lp –d printername –o options filename = prints filename to the printer named: printername using the printer specific options (such as landscape, nb, etc.)
list of common –o options:
c = compressed text
letter = 8.5 X 11
legal = 8.5 X 14
fp# = font pitch; Set the font pitch to <#> characters per inch.
lpi<#> = Print specified lines per inch, where <#> is 1, 2, 3, 4, 5, 6, 12, 16, 24, or 48.
duplex = prints on both sides of page
landscape = landscape orientation
portrait = portrait orientation (standard, default)
Either hit the cancel button on the laserjet or use the cancel command:
commonly used:
cancel printer12-123 - Cancels the printer12-123 job request
cancel printer12 - Cancels the currently serviced job on printer: printer12
cancel -e printer12 - Cancels all jobs (all users) on printer: printer12
cancel -udavid printer12 -Cancels all jobs submitted by user: david on printer: printer12
cancel Man page:
cancel Command
The cancel command cancels requests that were made with the lp
command, even if they are currently printing. At least one id or
printer must be specified.
The cancellation of a request that is currently printing frees the
printer to print its next available request.
id ... Specifies one or more requests. id is a request ID
returned by lp or lpalt.
printer ... Specifies one or more printers. printer is the name of
a printer, not a class. Either cancel the request that
is currently printing on each printer, or, if an -a, -
e, or -u option is specified, specify the printer on
which to perform the corresponding operation.
-a Remove all requests the user owns on each printer. The
owner is determined by the user's login name and the
host name of the machine where the lp command was
invoked.
-e Empty the spool queue of all requests for each printer.
Only users with appropriate privileges can use this
option.
-i Cancel only local requests.
-uuser Remove any requests belonging to user. You can repeat
the -u option to specify more users. Only users with
appropriate privileges can use this option.
Sometimes users will receive the following:
From: lp@mybox.mydomain.com
error code 2 in request printer12-555 on printer printer12
They receive the message because of either hitting cancel on the printer's control panel (most likely), or there's a problem with the Jetdirect timing.