Solaris 10: shminfo_shmmax

Hello,
I�m an Oracle DBA trying to understand the impacts that shmmax can have in a Solaris 10 system.
1. how does shminfo_shmmax setting affect the system performance (memory,paging,CPU) in Solaris 10?
2. Our server has 8G of RAM and runs with shmsys:shminfo_shmmax=0xffffffffffffffff . Is this setting correct and why?
3. What shmmax value do you suggest for our environment?
Thanks for your help.
Regards,
Giuseppe Centioli

Hi, My experience with Oracle from an SA's point-of-view is 1) set kernel params per Oracle's Install/Release notes and 2). increase from there as needed if you have the physical resources on the machine 3) don't exceed the system's resource capacity or not even the first instance of Oracle will be useful
Sun's doc on kernel tuning which also describes each of the parameters in question:
http://docs.sun.com/app/docs/doc/817-0404
NOTE: In Solaris10, setting kernel parameters in /etc/system is an obsolete method. Solaris 10 has a Resource Manager which is used to adjust these kernel params.
Resource Manager Admin
http://docs.sun.com/app/docs/doc/817-1592
Sun Blueprint re: Solaris/Oracle Tuning ( from 2003)
http://www.sun.com/blueprints/1003/817-3835.pdf
Solaris 10 Tuning Artticle:
http://www.unixadmintalk.com/f35/solaris-10-performance-tuning-recommendation-89335/
HTH
John

Similar Messages

ABAP IMPORT phase is slow in ECC 6.0 installation on Solaris 10

Hi gurus,
I installed ECC 6.0 on Solaris 10 operating system and Oracle 10.2.0.2. I installed all solaris patces and Oracle interim patches before the installation. Firstly I added only 2 parameters in /etc/system file. This parameters like this: "set shmsys:shminfo_shmmax=0xffffffffffffffff" ,"set rlim_fd_cur=8192". Then I started the installation. When the system check the parameters I have medium warning about set shmsys:shminfo_shmmax=0xffffffffffffffff .After that I didn't restart the solaris machine and recheck the parameters.After I removed this parameter I didn't have any warning and I started the installation. Everything was ok. I don't have any error about installation. The processes are continueing but now ABAP import is very very slow. Only 5 processes finished from 38 in ABAP IMPORT phase in 14 hours. Is it normal? Why this step is very slow? What can I do to increase this step? If I add "set shmsys:shminfo_shmmax=0xffffffffffffffff" again when processes is running, does this step accelerate? If I change the /etc/system parameters do I have to restart the solaris machine? What can I do to resolve this problem?
Best regards,

Hi Arif
<i>Only 5 processes finished from 38 in ABAP IMPORT phase in 14 hours. Is it normal?</i>
No, this is not normal. As far as i know, it is possible to set a parallel degree in sapinst. Of course this needs more hardware ressources. Did you do that?
To further diagnose your problem, you should check cpu and disk usage. It would be interesting which are the top processes running. Either your system is too slow, or some ressource is a bottleneck.
Regards
Michael

Installing oracle 10g release 2 on spark solaris!! urgent help!

hi guys!
Am installing oracle 10g r2 on spark sun solaris. I did install the software without any pb, but when installing the database using dbca, I have this errors.
ORA-27154: post/wait create failed
ORA-27300: OS system dependent operation:semget failed with status: 28
ORA-27301: OS failure message: No space left on device
ORA-27302: failure occurred at: sskgpcreates
Edited my kernel parameters and in the /etc/system file, ai've added
set noexec_user_stack=1
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=1024
set semsys:seminfo_semmsl=256
set semsys:seminfo_semvmx=32767
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
Space aint an issue and dont know how to over come this error!
Should ai just install the db using scripts and run the neccessary .sql scripts instead of going through dbca?
junior DBA in distress.

This appears to be because of low semaphores setting. Did you reboot your server after making the /etc/system parameter changes ?
-Ramesh

10g Universal Installer kernel errors while installing on Solaris 5.10

Hi all,
I have been trying to install Oracle 10g on a Solaris 5.10 machine (32-bit) as per the official Oracle Documentation for a couple of days now. It seems that the Universal Installer fails it's pre-installation steps and throws up error at the following kernel parameters:
Checking for shmsys:shminfo_shmmax=4294967295; found no entry. Failed <<<<
Checking for shmsys:shminfo_shmmni=100; found no entry. Failed <<<<
Checking for semsys:seminfo_semmni=100; found no entry. Failed <<<<
Checking for semsys:seminfo_semmsl=256; found no entry. Failed <<<<
Check complete. The overall result of this check is: Failed <<<<
I have followed the Oracle Documentation which says to change the kernel parameters using the resource control facility. For this I already have a project "oracle" which the user "oracle" has been assigned to:
$ id -ap oracle
uid=100(oracle) gid=100(oinstall) groups=100(oinstall),101(dba) projid=100(oracle)
Following are the snapshots from the prctl command on the current state of the kernel paramters:
For Checking for shmsys:shminfo_shmmax=4294967295; found no entry. Failed <<<<
$ prctl -n project.max-shm-memory -i project oracle
project: 100: oracle
NAME PRIVILEGE VALUE FLAG ACTION RECIPIENT
project.max-shm-memory
privileged 254MB - deny -
system 16.0EB max deny -
The Oracle recommended value is 4GB - I expect Universal Installer check to fail for this, so I change the parameter using the resource control facility by logging in as root in a new terminal and keeping one oracle session active.
# prctl -n project.max-shm-memory -v 4gb -r -i project oracle
I can see that the parameter has been set correctly when I do:
$ prctl -n project.max-shm-memory -i project oracle
project: 100: oracle
NAME PRIVILEGE VALUE FLAG ACTION RECIPIENT
project.max-shm-memory
privileged 4.00GB - deny -
system 16.0EB max deny -
For Checking for semsys:seminfo_semmni=100; found no entry. Failed <<<<
$ prctl -n project.max-sem-ids -i project oracle
project: 100: oracle
NAME PRIVILEGE VALUE FLAG ACTION RECIPIENT
project.max-sem-ids
privileged 128 - deny -
system 16.8M max deny -
The Oracle recommended value is 100 - I do not expect Universal Installer check to fail at this parameter
For Checking for shmsys:shminfo_shmmni=100; found no entry. Failed <<<<
$ prctl -n project.max-shm-ids -i project oracle
project: 100: oracle
NAME PRIVILEGE VALUE FLAG ACTION RECIPIENT
project.max-shm-ids
privileged 128 - deny -
system 16.8M max deny -
The Oracle recommended value is 100 - I do not expect Universal Installer check to fail at this parameter
Also when I try to issue the command:
$ prctl -n project.max-sem-nsems -i project oracle
I get:
prctl: failed to get resource control for project.max-sem-nsems: Invalid argument
Oracle recommends I should set a value of 256.
After setting up all this I still get the following errors from Universal Installer:
Checking for shmsys:shminfo_shmmax=4294967295; found no entry. Failed <<<<
Checking for shmsys:shminfo_shmmni=100; found no entry. Failed <<<<
Checking for semsys:seminfo_semmni=100; found no entry. Failed <<<<
Checking for semsys:seminfo_semmsl=256; found no entry. Failed <<<<
Check complete. The overall result of this check is: Failed <<<<
Now when I modify the kernel settings by editing the /etc/system file without even a reboot I'm can see all my Universal Installer Kernel checks as passed.
Not sure if I'm missing out some crucial step while setting up the kernel parameter using the resource control facility?
~ Somjit

It is bug in OUI.
OUI is looking at /etc/system to check the kernel parameter settings, but this file is not used (by default) in Solaris 10
It should check the kernel parameter settings using the prctl.
Check metalink note 363436.1
workaround is to
./runInstaller -ignoreSysPrereqs
or as You already did, change /etc/system as well.

Install on Solaris 9

I am using an Ultra60, 512MB, 2*450 processors.
Nothing but Solaris9 installed.
Try to install 9.2.1.0.1 Enterprise Edition 64bit. All OK untill database creation when I get 'run out of memory' error. In Solaris 8 with Oracle 7 you had to modify /etc/system shared memory parameters. Is this still the case for Soraris9/Oracle9i.
Thanks Terry Bennett

The Oracle documentation was pretty scarce on this but there were other docs on the Net which pointed out the correct parameters to edit. I added the following lines to /etc/system:
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=1024
set semsys:seminfo_semmsl=256
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
After adding these lines, reboot your server and you can verify the values using the system command.
Hope this helps.

Install oracle 8.1.7 in Solaris Problem

I have completed to install Oracle in solaris and
created database, when I use sqlplus login to new created database
The following error is displayed:
"ora-07217 : sltln : environment variable cannot be evaluated"
The following is the environment variable for oracle user
HOME=/oracle_data1/oracle/product/8.1.7
HZ=100
IFS=
LD_LIBRARY=/oracle_data1/oracle/product/8.1.7/lib:/usr/openwin/lib:/usr/db/lib
LD_LIBRARY_PATH=/oracle_data1/oracle/product/8.1.7/lib:/usr/openwin/lib:/usr/db/lib
LOGNAME=eprc817
MAILCHECK=600
OPTIND=1
ORACLE_BASE=/oracle_data1/oracle
ORACLE_HOME=/oracle_data1/oracle/product/8.1.7
ORACLE_SID=eprc
PATH=/usr/bin:/usr/ucb:/etc:/oracle_data1/oracle/product/8.1.7/bin:/usr/ccs/bin:/usr/openwin/bin:.
PS1=$
PS2=>
SHELL=/bin/sh
TERM=ansi
TZ=Hongkong
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=400
set shmsys:shminfo_shmseg=50
set shmsys:seminfo_semmni=100
set semsys:seminfo_semmsl=1024
set semsys:seminfo_semmns=1024
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767

You may try to add the environment variable
ORACLE_TERM=vt100 which is not include in your envornment setting.

Install Errors for 9i on Solaris 9

I am installing Oracle 9i Solaris 9 and the install went well until the Oracle Database Configuration Assistant. It failed with the following errors:
Oracle Database Configuration Assistant Errors for install
ORA-27303
ORA-27146
ORA-27300
ORA-27301
ORA-27302
Can someone enlighten me as to what these mean?
Message was edited by:
d072330

For all of those newbies like me I forgot to set some system settings. Here are the system settings I needed to put into /etc/system:
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=3000
set semsys:seminfo_semmsl= 256
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=200
set shmsys:shminfo_shmseg=20

Installing Oracle 8i on Solaris 9

I have tried to install Solaris 8i on solaris9 with no success.
The error massages I got were:
- Oracle not available
- Shared memory realsm does not exit
The kernel parameter set is:
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
set semsys:seminfo_semmns=2000
set semsys:seminfo_semmsl=1000
set semsys:seminfo_semmni=100
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767
unset TWO_TASK
Plus all the enrironment variables set up correcty.
With the same configuation, i successfully install 8i on Solaris 8.
Could anyone help, please?
Thanks

have a look in your $ORACLE_HOME/dbs directory. If the init<ORACLE_SID>.ora file has remote_login_password=exclusive then you should also have a file called orapw<ORACLE_SID>. If you haven't, then create one by running "orapwd file=orapw<ORACLE_SID> password=internal entries=5".
This has worked for me in the past, hope it helps.
Stef

Support for rfc1514 on Solaris 9

All,
Does anyone have experience configuring the snmp management agent on a Solaris 9 system?
I need to use the rfc1514 standard (host-resource mib) to support a large server environment for monitoring a variety of health stats that we are currently unable to capture, and I have not had a great deal of success finding the technical docs that outlines and addresses the procedures on how to do this.
If anyone has done this for a Solaris 9 build I would appreciate your assistance.
Sincerely,
Matt S.

For all of those newbies like me I forgot to set some system settings. Here are the system settings I needed to put into /etc/system:
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=3000
set semsys:seminfo_semmsl= 256
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=200
set shmsys:shminfo_shmseg=20

NIS install on solaris 9

I am trying to install NIS on solaris 9 but I don't have a Makefile in /var/yp. I tried copying the Makefile from one of my solaris 8 boxes but then I ended up missing a bunch of files from /usr/lib/netsvc/yp. I think I must be missing some packages but I can't find any listing of packages necessary to install NIS. I installed the developer bundle on the solaris 9 box that appears to be missing packages. Does anyone know what pkg(s) I am missing?
Some of the files I am missing from /usr/lib/netsvc/yp are mutli, ypxfer udpublickey and many more. Any help would be greatly appreciated.
Thanks,
Daniel

The Oracle documentation was pretty scarce on this but there were other docs on the Net which pointed out the correct parameters to edit. I added the following lines to /etc/system:
set semsys:seminfo_semmni=100
set semsys:seminfo_semmns=1024
set semsys:seminfo_semmsl=256
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
After adding these lines, reboot your server and you can verify the values using the system command.
Hope this helps.

Oracle 8.1.7 on solaris 7 install

Hi all
This is about a half solved problem that I thought someone can shed some light on. I tried to set up oracle 8.1.7 on a solaris 7 unix box and the install failed, just at the point the db assistant was connecting to the oracle db to create the first database instance with the error ORA - 03114: not connect to ORACLE
I called in some help, and together we managed to get the database working, but I'm still not fully understanding why.
There were two thing we discovered - using the debugger, adb -kw, that the shared memory :
shminfo_shmmax=4294967295
shminfo_shmmin=1
were not being set, even though they were in the etc/system file. These are recommended values in the oracle installation guide. We set this manually, and tried the install again. That failed the same way as before. So we tried an install of oracle 8.1.6, and that worked a treat, went on straight away. So investigating, we found that the shared memory pool on dbassist setup was smaller on the 8.1.6 setup than the default values for 8.1.7 and this caused the database assistant not to install the oracle instance. We adjusted the shared memory pool down, and it worked ok.
#extract from init file that worked
open_cursors = 100
max_enabled_roles = 30
db_block_buffers = 15300
shared_pool_size = 20889600
large_pool_size = 614400
java_pool_size = 0
#end extract
So it seems the shmem values don't have any affect on the oracle database, it was only the default values in the db assistant that was causing a problem.
So my question is does the shmem values really matter, and if they do, whats happened to them? Why doesn't setting them manually help?
Also why does the shared memory value in the db assistant have a effect on whether the database works or not?
Thanks
Regards
\Andrew
null

I ran into same problem with /etc/system data and it worked okay after I rebooted the OS.
Try a reboot and see if the implements the settings.

Installation in solaris Problem

I have completed to install Oracle in solaris and
created database, when I use sqlplus login to new created database
The following error is displayed:
"ora-07217 : sltln : environment variable cannot be evaluated"
The following is the environment variable for oracle user
HOME=/oracle_data1/oracle/product/8.1.7
HZ=100
IFS=
LD_LIBRARY=/oracle_data1/oracle/product/8.1.7/lib:/usr/openwin/lib:/usr/db/lib
LD_LIBRARY_PATH=/oracle_data1/oracle/product/8.1.7/lib:/usr/openwin/lib:/usr/db/lib
LOGNAME=eprc817
MAILCHECK=600
OPTIND=1
ORACLE_BASE=/oracle_data1/oracle
ORACLE_HOME=/oracle_data1/oracle/product/8.1.7
ORACLE_SID=eprc
PATH=/usr/bin:/usr/ucb:/etc:/oracle_data1/oracle/product/8.1.7/bin:/usr/ccs/bin:/usr/openwin/bin:.
PS1=$
PS2=>
SHELL=/bin/sh
TERM=ansi
TZ=Hongkong
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=400
set shmsys:shminfo_shmseg=50
set shmsys:seminfo_semmni=100
set semsys:seminfo_semmsl=1024
set semsys:seminfo_semmns=1024
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767

You might try setting ORACLE_SID=EPRC.

Memory leak in solaris 9

Hi ,
I have Oracle9i database running on Solaris 9. After every 6 days my solaris box went out of space although its a test box and there is no activity on the system. Why the hell it takes all the ram and swap space when there is no activity on the system except Oracle Instance is running with no user connection. What process are running in the background that takes all the memory and i have to restart this box every week. IS THERE ANT PATCH OF MEMORY LEAK FOR SOLARIS9?. I am not a solaris administrator and i have to fix this problem before moving my system to production. I would really appreciate your help.
Thanks.

Hi ,
I have Oracle9i database running on Solaris 9. AfterI have had for some months been running oracle 9i (and 8i) under solaris 9. There have been no problems and the server has been stable.
It sounds like you may have misconfigured the solaris kernel. It also sounds like your SGA is too large to fit in memory.
So I guess the first step would be to check the SGA and then the /etc/system file.
Here is what mine looks like for example:
SQL> show sga
Total System Global Area 219644872 bytes
Fixed Size 730056 bytes
Variable Size 167772160 bytes
Database Buffers 50331648 bytes
Redo Buffers 811008 bytes
SQL>
bash-2.05$ tail /etc/system
set shmsys:shminfo_shmmax=4294967296
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=100
set shmsys:shminfo_shmseg=10
set semsys:seminfo_semmni=100
set semsys:seminfo_semmsl=256
set semsys:seminfo_semmns=1024
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767
bash-2.05$
bash-2.05$ vmstat
kthr memory page disk faults cpu
r b w swap free re mf pi po fr de sr dd dd f0 s0 in sy cs us sy id
0 0 0 2076240 1302024 0 1 1 0 0 0 0 0 4 0 0 411 92 135 0 1 99
bash-2.05$
You need to gather more system information to help you troubleshoot.

Oracle 8i on Solaris 8

I am trying to get oracle 8i enterprise server to work on solaris 8
I have tuned my kernel using /etc/system using this lines
START oracle mods
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=600
set shmsys:shminfo_shmseg=100
set semsys:seminfo_semmap=100
set semsys:seminfo_semmni=1024
set semsys:seminfo_semmns=14000
set semsys:seminfo_semmnu=400
set semsys:seminfo_semume=100
set semsys:seminfo_semmsl=1024
set semsys:seminfo_semopm=100
* Message Queue
set msgsys:msginfo_msgmap=1024
set msgsys:msginfo_msgmax=8192
set msgsys:msginfo_msgmnb=262140
set msgsys:msginfo_msgmni=1024
set msgsys:msginfo_msgssz=2048
set msgsys:msginfo_msgtql=1024
set msgsys:msginfo_msgseg=4096
* END oracle mods
I have also dedicated a port for the listener in the /etc/services
but I am still haveing the following problems when I try to tnsping my DB it gives me the folling error
TNS Ping Utility for Solaris: Version 8.1.7.0.0 - Production on 8:08
(c) Copyright 1997 Oracle Corporation. All rights reserved.
TNS-03505: Failed to resolve name
when I do an lsnrctl status it gives me the following print out
LSNRCTL for Solaris: Version 8.1.7.0.0 - Production on (c) Copyright 1998 Oracle Corporation. All rights reserved.
Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=tomiwa)(PORT=1521)))
STATUS of the LISTENER
Alias LISTENER
Version TNSLSNR for Solaris: Version 8.1.7.0.0 - Production
Start Date 21-MAR-
Uptime 0 days 3 hr. 13 min. 48 sec
Trace Level off
Security OFF
SNMP OFF
Listener Parameter File /u01/app/oracle/product/8.1.7/network/admin/listener.o
ra
Listener Log File /u01/app/oracle/product/8.1.7/network/log/listener.log
The listener supports no services
The command completed successfully
which says at the bottom the listener suports no services.
when I try using sqlpus internal i get the following
SQL*Plus: Release 8.1.7.0.0 - Production on Sun
(c) Copyright 2000 Oracle Corporation. All rights reserved.
Enter user-name: internal
ld.so.1: oracle: fatal: libskgxp8.so: open failed: No such file or directory
ERROR:
ORA-24314: service handle not initialized
can anybody help it is driving me crazy
null

Hi,
from your previous post, are you sure there is no
error during the installation process? Can you check the
log file make.log (and other log files also) to see if any
error there? because from what
you have said, it seems the installation is not complete(have
some error). --- especially check if there is errors such
as "can't find '-lordsdo'".
Another thing is make sure the LD_LIBRARY_PATH is set correctly.

Installing Oracle 9i on Solaris 10 SPARC, /etc/system vs. projects

Can anyone verify that the kernel tuning parameters that used to be set in /etc/system on Solaris 8 have any effect in Solaris 10? All the documentation I can find suggests that most (if not all) of these settings have been marked obsolete or removed in Solaris 10. It is my understanding that these settings (such as shared memory) need to be set by means of a project. If we are running with no project setup in Solaris 10 are we pretty much running "untuned" ? Thanks
Some links...
New or change parameters in Solaris 10
http://docs.sun.com/app/docs/doc/817-0404/appendixb-16?a=view
Solaris Tunable Parameters manual
http://dlc.sun.com/pdf/817-0404/817-0404.pdf

We have carried over the following parameters from /etc/system on our old Solaris 8 DB servers. We just want to know if these settings are having any effect. We are currently seeing some performance issues on T5220's with 32GB RAM (old DB servers were SunFire 4800's).
shmsys:shminfo_shmmax obsolete as of Solaris 10
shmsys:shminfo_shmseg removed in Solaris 10
shmsys:shminfo_shmmin removed in Solaris 10
shmsys:shminfo_shmmni obsolete as of Solaris 10
semsys:seminfo_semmns removed in Solaris 10
semsys:seminfo_semmni obsolete as of Solaris 10
semsys:seminfo_semmsl obsolete as of Solaris 10
semsys:seminfo_semopm obsolete as of Solaris 10
semsys:seminfo_semvmx removed in Solaris 10
rlim_fd_max=8096 In the Solaris 8 version, the default is 1024. In later Solaris releases, the default is 65,536
rlim_fd_cur=1024 Still supported
noexec_user_stack Still supported
Actual /etc/system settings...
set shmsys:shminfo_shmmax=31457280000
set shmsys:shminfo_shmseg=16
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=128
set semsys:seminfo_semmns=11000
set semsys:seminfo_semmni=0x402
set semsys:seminfo_semmsl=4096
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767
set rlim_fd_max=8096
set rlim_fd_cur=1024
set noexec_user_stack=1
Edited by: [email protected] on Mar 6, 2009 12:25 PM

Solaris 10: shminfo_shmmax

Similar Messages

Maybe you are looking for