Can not login to database as local connetion as not oracle user

I login as "oracle" user in unix machine. I am able to login as another user
in
sqlplus.
But when another unix user tries to use sqlplus he gets following
error:SQL*Plus: Release 8.1.7.0.0 - Production on Mon Oct 7 10:50:21 2002
(c) Copyright 2000 Oracle Corporation. All rights reserved.
Enter user-name: hra
Enter password:
ERROR:
ORA-01034: ORACLE not available
ORA-27101: shared memory realm does not exist
SVR4 Error: 2: No such file or directory
But if use /userid/passwd@service_name then it log in correctly. Also if I set

TWO_TASK sqlplus does not give error. But I am logging to directly server,
therfore without servicename I should be able to login to sqlplus.
Puzzle is that I am able to work fine when I login as unix user "ORACLE" but
get errors when try to login as another unix user. I also checked that
ORACLE_SID is in proper case. (ps -ef | grep SMON)
I want to login without using vNet8.

Thx
Vivek

Do all linux users have .profile (.login) with oracle environment variables?
ORACLE_HOME,
ORACLE_SID
...
I gess this is ur problem.

check out the permissions.
go to $ORACLE_HOME/bin
oracle executable in unix should be SHARED to different users.
Typing in 'ls -l oracle' should yield

-rwsr-xr-x oracle dba

I have set all the variables correct. I am sure thats not the problem.

Vivek

Quote:

---

Originally posted by magvivek
check out the permissions.
go to $ORACLE_HOME/bin
oracle executable in unix should be SHARED to different users.
Typing in 'ls -l oracle' should yield

-rwsr-xr-x oracle dba

---

because users have messages like :
ERROR:
ORA-01034: ORACLE not available
ORA-27101: shared memory realm does not exist
and can use db via sqlnet then they have all needed permitions
on oracle-home/bin

Another thing:
check linux kernel parameters, maybe # of os processes/semaphores are few,
and when addition user try to create new unix process (oracle dedicated server) u have problem.

When you receive this error: make sure that you have the following variables set in /etc/system. If not, copy /etc/system to /etc/system.old then make your changes to /etc/system. The variable amounts change depending on your server.

This is from Solaris 2.8, Ultra 10, with 4 G of memory.
* For ORACLE:
* 1 meg = 1048576
* 200 meg = 209715200
* 4 gig = 4294967295
set shmsys:shminfo_shmmax=209715200
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=256
set semsys:seminfo_semopm=100
set semsys:seminfo_semvmx=32767

To check the shared memory use the command (in Solaris)
ipcs -A

(in AIX) ipcs -pmb

Good luck -

You are running out of the shared memeory or the semaphores. Increase them appropriately and bounce the server. Things should be fine.

I notice that every one were talking of Linux where the problem seems to be on Unix.

Provide us the server type solaris/HP/... then we will be able to let you know where to look for the file.

Sam

I have correctly set /etc/system for sepashore.

Oracle executable has correct permission with sticky bit.

If semasphore are less then how come i am able to login using oracle account.
Vivek

This was taken from another forum on Shared Memory: Just an FYI.....

Thanks to all. I was trying to determine the current shared memory status, and
how to change their values.

Original question:
--------------------

Do you know how to determine what the current value settings are for
the following:

shmsys:shminfo_shmmni
shmsys:shminfo_shmseg
shmsys:shminfo_shmmax
shmsys:shminfo_semmni
shmsys:shminfo_semmsl
shmsys:shminfo_semmns
shmsys:shminfo_semmnu

I have checked the /etc/system file and there are no entries for the
above, I assume that means the system is runing the defaults. What are
the defaults? I used the command sysdef and some of the above are listed
but, with the value 0. Does that mean the default is 0 or 0 represents
some default? How do I find out what the actual default values are?
Any help on this would be appreciated.


Responses:
-------------------



Response #1 -- [email protected]
Hope this article helps you:-

The solaris 2.X kernel comes configured with defaults for shared memory
parameters which may not be enough for large applications, such as databases.
Many database manufacturers, such as Oracle, Sybase and Informix
require that shared memory parameters be increased to handle their
system. In general, any program which makes shared memory system
service calls will be affected.

Here are the shared memory parameters, what they are, what their default
values are, and what they represent. Please note that the maxima listed
here are based on values which will fit into their datatype, which is a
(4-byte) integer, and are not realistic. Realistic maxima are not easily
figured, as they must account for many factors, including the amount of
kernel memory required by everything else on the system, and the
amount of memory the system has.

shmmax 1048576 (1Meg) 2147482647

This is the maximum size of a shared memory segment, or the largest
value which can be requested of shmget(2). Setting this value high
does not hurt anything as the resource it controls is not
preallocated, but rather is allocated on demand.

shmmin 1 2147482647

This is the minimum size of a shared memory segment. There is no need
to change this from the default, as doing so will only break code
which allocates less than the new shmmin value allows.

shmseg 6 2147482647

This is the maximum number of shared memory segments per process. It
is used as a high limit, which is checked before another segment is
allocated. No resources are preallocated based on its value.

shmmni 100 2147482647

This is the maximum number of shared memory identifiers that can exist
systemwide. Every shared memory segment has an associated identifier,
returned from shmget(). Resources are preallocated based on this
value, (112 bytes of kernel memory per identifer) so don't set it too high.

Shared memory parameters are changed by editing the /etc/system file with
lines of the form:

set shmsys:shminfo_variable = value

For example:

set shmsys:shminfo_shmmni = 150

to change shmmni to 150.

After changing the /etc/system file, reboot the system to have it take
effect.

All InterProcess Communication (IPC) parameters, including those for shared
memory, are checked by doing the following:

$ sysdef -i

Look toward the bottom of the output, for something like:

*
* IPC Shared Memory
*
1048576 max shared memory segment size (SHMMAX)
1 min shared memory segment size (SHMMIN)
100 shared memory identifiers (SHMMNI)
6 max attached shm segments per process (SHMSEG)

The values at the left will reflect the values set in the /etc/system file.
If the values are zeros, load the shared memory module and try sysdef -i
again.
Do the following to load the shared memory module:

# modload /kernel/sys/shmsys


Response #2 -- Carlo Cosolo

Carlo suggested I call SecurID tech support, as he had problems with their
parameters and Sol 2.5.

Response #3 -- Ju-Lien Lim

man sysdef

Response #4 -- [email protected]

man ipcs

Response #5 -- Glenn Satchell - Uniq Professional Services


You can inspect the current value in th ekernel using adb:

# adb -k
physmem 3e40
shminfo_shmmni/D
symbol not found
^D
#

This means that this system is not using _any_ shared memory stuff as
the modules haven't even been loaded.

If you already have these values configured then just use the larger of
the two values (securid recommendation or current).


Response #6 -- "K.Ravi"

Hi,
In Solaris, the shared memory module (or any other driver/module for that
matter) is not loaded into memory unless some process needs it. To know the
defaults, run 'modload /kernel/sys/shmsys' as root and then run 'sysdef'. If
sysdef doesn't report all the values, try this:
# adb -k /dev/ksyms /dev/mem

What type of OS are you using?

I am usinf folwing OS :

SunOS home 5.7 Generic_106541-23 sun4u sparc SUNW,UltraSPARC-IIi-cEngine

thx
Vivek

echo $TWO_TASK

If I set TWO_TASK then it will work because then it uses Oracle sqlnet v8. I want to connect to database without giving service name.

Vivek

I should have asked this before. :) OOPS You do have an interesting problem. I hope we can help you fix.

What Oracle Version are you running on your server?

Also type the following commands and show us the output:
1) Info about Shared memory: ipcs -A
2) What is your current swap: swap -s
3) How many processors:
/usr/sbin/psrinfo -v |grep "Status of processor"|wc -l
4) Top 6 process hogs: ps -ef|sort +6| tail
5) Display Ram Size: prtconf| grep -i mem

Check your Alert log for ORA- errors. Are there any? If so what?

Check your /var/adm/messages file for unusual errors.

Just a few troubleshooting ideas. Have you tried Metalink?

Forget the last reply. It sounds like the real issue is that you do not want to use net8.

Just a question: Why would you not want to use net8?

Well it is not that I don't want to use Net8.

I just want to know why it is not working? It is a simple local connection. It works in my all other servers with same database.ipcs -A
IPC status from as of Wed Oct 9 14:56:32 PDT 2002
Message Queue facility inactive.
T ID KEY MODE OWNER GROUP CREATOR CGROUP NATTCH SEGSZ CPID LPID ATIME DTIME CTIME ISMATTCH
Shared Memory:
m 0 0x500004a4 --rw-r--r-- root root root root 1 68 210 210 16:53:25 16:53:25 16:53:25 1
m 6001 0xc020b8c0 --rw------- oracle dba oracle dba 7 375144448 6124 6200 11:25:45 11:26:01 16:28:144294967292
T ID KEY MODE OWNER GROUP CREATOR CGROUP NSEMS OTIME CTIME
Semaphores:
s 7536640 0xf16a3764 --ra------- oracle dba oracle dba 304 4:30:20 16:28:16
vivek

Can you just do

Code:

---

  tnsping servicename

---

Tell us what the result? I'm also curious to see your sqlnet.ora file settings.

Sam

Quote:

---

Originally posted by vdeo
If I set TWO_TASK then it will work because then it uses Oracle sqlnet v8. I want to connect to database without giving service name.

Vivek

---

I am not asking you to use TWO_TASK, I was asking what's your TWO_TASK value

Look at the permissions on your output:
__________________________________________________________
I just want to know why it is not working? It is a simple local connection. It works in my all other servers with same database.ipcs -A
IPC status from as of Wed Oct 9 14:56:32 PDT 2002
Message Queue facility inactive.
T ID KEY MODE OWNER GROUP CREATOR CGROUP NATTCH SEGSZ CPID LPID ATIME DTIME CTIME ISMATTCH
Shared Memory:
m 0 0x500004a4 --rw-r--r-- root root root root 1 68 210 210 16:53:25 16:53:25 16:53:25 1
m 6001 0xc020b8c0 --rw------- oracle dba oracle dba 7 375144448 6124 6200 11:25:45 11:26:01 16:28:144294967292
T ID KEY MODE OWNER GROUP CREATOR CGROUP NSEMS OTIME CTIME
Semaphores:
s 7536640 0xf16a3764 --ra------- oracle dba oracle dba 304 4:30:20 16:28:16
____________________________________________________________

The Oracle file should give rw to oracle and the group dba, but your output shows that it does not. Also Semaphores gives ra to oracle only, it should be oracle and dba.

Looks like a permission problem. My output looks like this:
/export/home/>ipcs -A
IPC status from as of Thu Oct 10 09:06:23 MDT 2002
T ID KEY MODE OWNER GROUP CREATOR CGROUP CBYTES QNUM QBYTES LSPID LRPID STIME RTIME CTIME
Message Queues:
T ID KEY MODE OWNER GROUP CREATOR CGROUP NATTCH SEGSZ CPID LPID ATIME DTIME CTIME I
SMATTCH
Shared Memory:
m 0 0x50005f4f --rw-r--r-- root root root root 1 4 259 259 10:52:44 10:52:44 10:52:44
0
m 101 0xa53ab7ac --rw-rw---- oracle dba oracle dba 10 160915456 461 4965 8:12:27 8:12:27 15:15:39
10

T ID KEY MODE OWNER GROUP CREATOR CGROUP NSEMS OTIME CTIME
Semaphores:
s 458752 0xe035d35c --ra-ra---- oracle dba oracle dba 104 8:19:32 15:15:40
s 720897 0x42aab180 --ra-ra---- oracle dba

I know you said you checked the file permissions, but could you check again? Check $ORACLE_HOME/dba/init.ora, ....
Actually I would go through all of $ORACLE_HOME and check the permissions on all the files. Check $ORACLE_BASE/admin/sid files also.

Good Luck -

Hi I checked my other servers and there also I have same sepasphore permossion. I don't have any problem there.

vivek

The issue here could be the load on this system might be different to the others. The applications that run on this system might be different.

If you want to solve the issue, then you got to try first and find out whether you can solve. If successful then spend your time on thinking and contemplating. Its not worth the time of us to provide you any suggestions when you are not ready to test them out!

This is my personal opinion on reading the entire thread once again. Don't think that we are not willing to help. If we need to help you for no cost, then you have to provide us with some test feedbacks. Then only we can eliminate to possibilities.

GOOD LUCK!

Sam

I think I provided all the test result. Please let me know which test results you eant.

Thanks
Vivek

Can you answer the following:

1. Are you trying to connect to a remote server or is
the db on the same server?

2. IF the db on the same server, did you try increasing the
semaphore limits, reboot the server and then restart the
db and then try?

3. What was the result of PANDO's suggestion?

4. Can you tnsping servicename?

5*. Last call the oraenv and set the environment variable
and try again? In my opinion, its the environment
variable that causeing all the pains. Recheck with
the same server and set. Don't compare with the other
servers as the path might be different from this
installation.


-Sam

[Edited by sambavan on 10-10-2002 at 03:44 PM]

1. Db is on the same server.
2. increase semesphoe limit. I didn't tried but how much I should increase? if this value is less then how come i am able to login using oracle unix account.

3.TWO_TASK value is blank.
4.Yes I can tnsping service name.
5. I tried oraenv. Still no luck.

Although I have something new to add:
1. I upgraded the oracle to 8.1.7.4 from 8.1.7.0.
2. My oracle software is installed on /opt/oracle directory.
Which is linked to oracle.
lrwxrwxrwx 1 root other 11 Sep 30 10:42 oracle -> /opt/oracle

cd /opt
========
$drwxr-xr-x 5 oracle dba 512 Oct 9 20:39 oracle


df -k
======
/dev/md/dsk/d70 11207576 9766135 1329366 89% /opt

Does anythingh look suspicious ? Somehow I feel that another unix user does not have permission to create sepashpore for local connection.

Thanks
Vivek

Thank GOD Problem is solved.

I was setting my ORACLE_HOME as
/opt/oracle/product/8.1.7 instead of /oracle/product/8.1.7
(In unix when I say cd /oracle it goes to /opt/oracle becasue there is a link>>>

lrwxrwxrwx 1 root other 11 Sep 30 10:42 oracle -> /opt/oracle


But with old oracleHOME I am able to connect to database with servicename attached.

Vivek

Wasn't the very first suggestion (Shestakov's) you've got in this thread about ORACLE_HOME and ORACLE_SID? To which you've responded with:

Quote:

---

Originally posted by vdeo
I have set all the variables correct. I am sure thats not the problem.

---

So, never be sure of anything untill you try/check/recheck it. :o

But Still I have my old oracle_home in listener.ora and oratab files.

With Old ORACLE_HOME how I was able to connect to database with service name ?

Vivek

As long as you have configured your ORACLE_HOME to the right directory path where it can find the executables, it will work.

Sam