Printable View
Hey Guys,
i have a small problem. I want to start our db on a sun solaris ultar80 with solaris2.7.
I connect me as internal at sqlplus. Then i make a startup and after that the error occurs:
ORA-27102 out of memory
SVR4 Error:22:Invalid argument
Then i look after the error ORA-27102
and i should consult the trace file for details:
i looked after all the trace-files but there are no. So i controlled our init.ora if the all path are right. All is right.
Whats now.
Thanks for help
Thomas Schmidt
You either need to adjust your kernal parameters or reduce your memory parameters in the init.ora.
But way i thake round about 50MB from the ram and with top i can see that there are 831MB RAM are free. Why memory problems
Thomas Schmidt
[Edited by tomate on 08-21-2001 at 10:06 AM]
hello
try this export ORACLE_SID=my oracle sid
then try svrmgrl>>
regards
hrishy
hey hrishy,
the oracle_sid is set:
I think the error from thre next line says more but i don't know what this mean:
Here the line again:
SVR4 Error:22:Invalid argument
Thanks for help
On SQL/Plus after connecting as internal try to start the db in steps to narrow down the problem --
SQL> startup nomount;
SQL> alter database mount;
SQL> alter database open;
and then post the exact error message and what stage it comes at.
- Rajeev
Hey rsuri,
this have i make as one of my first steps:
The error is the same after i connect to sqlplus or svrmgrl as internal and i make startup mount the error occurs:
ORA-27102 out of memory
SVR4 Error:22:Invalid argument
This is a very good joke from oracle, that there is no more information.
Do you know more
Thanks
As Jeff mentioned, you need to lower your memory parameters, could you post the following parameters from the init.ora file --
db_block_buffers
shared_pool_size
db_block_size
- Rajeev
db_block_buffers = 5000 and this is reduced from me from 9746
shared_pool_size = 26613760
db_block_size =8192
But i think the parameters are all right or??
Thanks very much for the help
It's not an Oracle error, it's an OS error. This is why Oracle can't give you any more information...Quote:
Originally posted by tomate
This is a very good joke from oracle, that there is no more information.
thanks marist89 and all other. The Problem is in the OS.
All people said me thats all right and so i must search the error. I' describe in some hours here the error and what i make.
So long
Thomas Schmidt
Here is what Oracle says about that:Quote:
Originally posted by tomate
Hey Guys,
i have a small problem. I want to start our db on a sun solaris ultar80 with solaris2.7.
I connect me as internal at sqlplus. Then i make a startup and after that the error occurs:
ORA-27102 out of memory
SVR4 Error:22:Invalid argument
Then i look after the error ORA-27102
and i should consult the trace file for details:
i looked after all the trace-files but there are no. So i controlled our init.ora if the all path are right. All is right.
Whats now.
Thanks for help
Thomas Schmidt
Starting with kernel 2.3.51 shared memory is programmed to act like a virtual
filesystem. This is done to make Linux more POSIX compatible.
The change was discussed in December 1999 on the Linux Kernel Mailinglist and
was fast adapted.
No other changes or relinking is necessary for Oracle to work with 2.3.51 or
higher. The 2.3.51 is the last release before the Pre-2.4 kernel series starts.
This means that Oracle will work without any problems under the 2.4 kernel
series which is expected to be released later this year.
Here is what you should do:
1. Create the directory, as root, /var/shm if it does not exist
2. Enter the following line in /etc/fstab:
none /var/shm shm defaults
3. Mount the shared memory as follows:
mount shm
4. Start Oracle as usual
Cheers,
Julian
his OS is not Linux, it's Solaris.
I bet it comes from kernel parameters.
Thomas, can you tell us which values you have in /etc/system ??
Yes, you are right, he is on Solaris. Here are some sample values for Sun Solaris system values (8.1.7):Quote:
Originally posted by pipo
his OS is not Linux, it's Solaris.
I bet it comes from kernel parameters.
Thomas, can you tell us which values you have in /etc/system ??
* Oracle 8.1.7 settings:
set shmsys:shminfo_shmmax=4294967295
set shmsys:shminfo_shmmin=1
set shmsys:shminfo_shmmni=300
set shmsys:shminfo_shmseg=50
set semsys:seminfo_semmni=400
set semsys:seminfo_semmsl=600
set semsys:seminfo_semmns=400
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767
forceload: drv/sd
forceload: drv/rdnexus
forceload: drv/rdriver
forceload: drv/vxio
forceload: drv/vxspec
forceload: drv/scsi
forceload: drv/pci
forceload: drv/sf
forceload: drv/pln
forceload: drv/ses
forceload: drv/soc
forceload: drv/socal
rootdev:/pseudo/vxio@0:0
set vxio:vol_rootdev_is_volume=1
thanks to all for the very good help.
The Problem was the kernel parameters. The Server was installed by an other person not from me and he told me that all server was instaalled as same. So on this mashine the parameters was not set . Now i controlled all and i set the kernel parameters write and now the database is open.
Thanks for the success.
The only time I can remember my Database not starting was when I clumsily tried to impement multiple buffer pools. Can't remember if I did get an error message pointing me in that direction or I figured it out for myself.
Recycle, keep pool taken in total must not exceed db_block_buffers or the database will not start.
The Drain Surgeon
Keep pool is very useful, be careful with how many lru_latches you assign to it. Do not need recycle pool if you really do not need it. Estimate its size carefuly (rather too big than too small).Quote:
Originally posted by drain surgeon
The only time I can remember my Database not starting was when I clumsily tried to impement multiple buffer pools. Can't remember if I did get an error message pointing me in that direction or I figured it out for myself.
Recycle, keep pool taken in total must not exceed db_block_buffers or the database will not start.
The Drain Surgeon
hello Thomas,
i got the same problem than you about the kernel parameter, i would like to know which parameter has you changed to correct the problem on your /etc/system file..?
thanks a lot
David.
dude this thread is 5 years old