Solaris Privs
I have a question about the following from the man pages on privileges:
PRIV_FILE_DAC_WRITE
Allow a process to write a file or directory whose permission bits or ACL do not allow the process write permission. All privileges are required to write files owned by UID 0 in the absence of an effective UID of 0.
I am unclear on what that means - I ask because of a truss output I collected on a process that appeared to be hung from the user's perspective but was burning up the cpu via top:
stat("M$FST000.TMP", 0xFFFFFFFF7FFFD6E8) Err#2 ENOENT
stat("M$FST000.TMP", 0xFFFFFFFF7FFFD830) Err#2 ENOENT
open("M$FST000.TMP", O_RDWR|O_CREAT, 0666) Err#13 EACCES [file_dac_write]
I did post a different question regarding this truss output - and do not intend to double post. I would simply like a description of the EACCESS (file_dac_write) error that is being returned in this example. My limited understanding of this would indicate that the process was attempting to create a file or directory with the specified permissions but failed because some element in the file ro dirs path was exclusively owned by root.
Thanks -
Mike
Yes, you are right. The EACCES indicates that the process is not allowed to open that file for writing. I guess that the [file_dac_write] indicates that the system, upon failure to write to the file, also checked if the user had the appropriate privilege (file_dac_write), which he didn't.
You could take a look at the manpage for 'open' and see what it says about EACCES.
.7/M.
Similar Messages
-
Installation Woes on Solaris 10
I have looked thru all the issues posted so far and I have worked past those, my issue occurs during the installation of the Oracle JVM:
The installation is at 36% when I hit this error:
ORA-00604: error occured at recursive SQL level 1
ORA-04031: unable to allocate 4096 bytes of shared memory.("java pool",.....)
ORA-06512: at "SYS.DBMS_JAVA", line 222
ORA-06512: at line 1
I've created a project for the oracle user and assigned a value for max shared memory:
# id -p oracle
uid=102(oracle) gid=1000(dba) projid=100(oracle)
# cat /etc/project | grep oracle
oracle:100::::project.max-shm-memory=(priv,536870912,deny)
I then activated this setting dynamically:
#prctl -n project.max-shm-memory -v 512mb -r -i project oracle
Even though I shouldn't have to reboot after this change, I went ahead an rebooted anyway.
I am trying to install 10.2.0.1 64-bit on Solaris 10 with latest recommended patch cluster as of 13-Jan-05.
Any ideas?
Thx,
CCWell, I am using scripts generated by dbca, surely those allocate enough for itself to function?
Also, under Solaris 10, memory parameters are no longer pulled from /etc/system.
-CC -
Root.sh fails while installation of 10.2.0.1 clusterware on Solaris
All
I am trying to configure RAC on solaris 10 OS running on a 64 bit SPARC machine using RAW slices for the cluster files /dev/rdsk/c1t9d0s5 (OCR)and /dev/rdsk/c1t9d0s6 (Vote Disk). Using format command, I would create these slices leaving about 10 cylinders initially I would assign 200mb for each slice.
when I try to run OUI for cluster ware 10.2.0.1 at the step when I am asked to run root.sh, my attempt to run this script continues to fail with below error message.
root@hcssun01 # /u01/app/oracle/product/10.2.0/crs_home/root.sh
WARNING: directory '/u01/app/oracle/product/10.2.0' is not owned by root
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
Checking to see if Oracle CRS stack is already configured
Checking to see if any 9i GSD is up
/u01/app/oracle/product/10.2.0/crs_home/bin/lsdb: failed to initialize interface to Cluster Manager.
Setting the permissions on OCR backup directory
Setting up NS directories
Oracle Cluster Registry configuration upgraded successfully
WARNING: directory '/u01/app/oracle/product/10.2.0' is not owned by root
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
clscfg -install -nn nodeA,nodeAnum,nodeB,nodeBnum... -o crshome
-l languageid -c clustername -q votedisk
[-t p1,p2,p3,p4] [-pn privA,privAnum,privB,privBnum...]
[-hn hostA,hostAnum,hostB,hostBnum...]
-o crshome - directory CRS is installed in
-q votedisk - path to the CSS voting disk
-c clustername - name of the cluster. 1-14 character string
-l languageid - Oracle localization language id.
e.g. AMERICAN_AMERICA.WE8ASCII37C
-nn name,num - nodename list in pairs of nodename,nodenumber
If OS clusterware is installed see vendor docs.
e.g. node1,1,node2,2,node4,4
-pn name,num - Defines private interconnect names for nodes already
specified with the -nn flag.
Defaults to the nodename if not specified.
-hn name,num - Defines hostnames for nodes specified with the -nn
flag in the same format as above.
Defaults to the nodename if not specified.
-t p1,p2,p3,p4 - Specifies TCP ports to be used by the CRS daemons
on the private interconnect.
default ports: 49895,49896,49897,49898
-force Forces overwrite of any previous configuration.
WARNING: Using this tool may corrupt your cluster configuration. Do not
use unless you positively know what you are doing.
Failed to initialize Oracle Cluster Registry for cluster
0
From the log files under the folder CRS_HOME/log/<hostname>/client does not reveal much information either... If you have any idea of whats going wrong and needs a fix/workaround could you please help me? Below are more details of the Software Stack and H/W which may be of some help.
OS: Solaris 5.10 running on SF 480R.
Cluster 10.2.0.1
I intend to install RDBMS 10.2.0.1 on top of this and later upgrade it to 11.1
I am referring to the Oracle Clusterware and Oracle Real Application Clusters Installation Guide (b14205-01) published by Oracle.
Thanks in advance.
Sarat.Thanks for the assiatance. However, I am yet to proceed beyond this root.sh.
after multiple tries.. I am coming across this Oracle UDLM software which I am sure I do not have on my cluster machines. But not in every place is it mentioned? What is this software about.. I have searched all my oracle and no where this package / patch is available.. if you know about it ( the patch number or something) can you help me?
Also I have a question to clarify... it could be elimentary .. but request your patience... On the Sun Cluster we have defined Disk Resource Groups which are configured for cold failover along with other resource groups ... is this some thing which is not allowed/permitted.. I mean not supported...
I will need your help as I am singing this song of Oracle RAC on Sun Solaris 10 for the first time..
Regards!
Sarat -
EBS12 installation on sparc solaris 10 failing
hi !
Thanks to Hussein's great help I was able to make it work once but then I deinstalled it and now installing again from fresh but I can't get it to work , it keep failing with same error :
Note: I am not recreating the staging area from scratch but Iam using the staging area from previous installation assuming this is a static enviornment.
Log file located at /ebs/oracle/VIS/db/tech_st/11.1.0/appsutil/log/VIS_sunny/ApplyDatabase_10260217.log
| 0% completed
\ 0% completed RC-50004: Fatal: Error occurred in ApplyDatabase:
Control file creation failed
Cannot execute configure of database using RapidClone
RW-50010: Error: - script has returned an error: 1
RW-50004: Error code received when running external process. Check log file for details.
Running Database Install Driver for VIS instance
I am installing as oracle and choose oracle:dba as database owner and applmgr:dba as apps owner.
I made sure the permissions on oraInventory directory is 777 , also on the /var/tmp/.oracle . I checked every step in preinstall document for validity and they all pass.
even though its complaining about control file but i can start the database in mount state , although it crashes if i try to open.
my project settings are as follows :
default:3::::process.max-sem-nsems=(priv,256,deny);project.max-file-descriptor=(priv,65536,deny);project.max-sem-ids=(priv,100,deny);project.max-shm-ids=(priv,100,deny);project.max-shm-memory=(priv,4294967295,deny)
group.dba:100:oracle kernel parameters:oracle:dba:process.max-sem-nsems=(priv,256,deny);project.max-file-descriptor=(priv,65536,deny);project.max-sem-ids=(priv,100,deny);project.max-shm-ids=(priv,100,deny);project.max-shm-memory=(priv,4294967295,deny)
ulimit settings are :
$ ulimit -a
time(seconds) unlimited
file(blocks) unlimited
data(kbytes) unlimited
stack(kbytes) 65536
coredump(blocks) unlimited
nofiles(descriptors) 65536
memory(kbytes) unlimited
oracle group settings are:
uid=301(oracle) gid=1000(dba) groups=1000(dba),2000(applmgr)
applmgr group settings are:
uid=401(applmgr) gid=2000(applmgr) groups=2000(applmgr),1000(dba)
I can start the database in mount state meaning the control file does exist
SQL> startup mount
ORACLE instance started.
Total System Global Area 1069252608 bytes
Fixed Size 2109352 bytes
Variable Size 427823192 bytes
Database Buffers 624951296 bytes
Redo Buffers 14368768 bytes
Database mounted.
SQL> show parameter control
NAME TYPE VALUE
control_file_record_keep_time integer 7
control_files string /ebs/oracle/VIS/db/apps_st/data/cntrl01.dbf, /ebs/oracle/VIS/db/apps_st/data/cntrl02.dbf, /ebs/oracle/VIS/db/apps_st/data/cntrl03.dbf
SQL> alter database open
ERROR at line 1:
ORA-00600: internal error code, arguments: [2252], [2359], [322925629], [], [],
$ lsnrctl status VIS
LSNRCTL for Solaris: Version 11.1.0.7.0 - Production on 26-OCT-2000 11:45:27
Copyright (c) 1991, 2008, Oracle. All rights reserved.
Connecting to (DESCRIPTION=(ADDRESS=(PROTOCOL=TCP)(HOST=sunny.localdomain)(PORT=1521)))
STATUS of the LISTENER
Alias VIS
Version TNSLSNR for Solaris: Version 11.1.0.7.0 - Production
Start Date 26-OCT-2000 02:18:20
Uptime 0 days 9 hr. 27 min. 7 sec
Trace Level off
Security ON: Local OS Authentication
SNMP OFF
Listener Parameter File /ebs/oracle/VIS/db/tech_st/11.1.0/network/admin/VIS_sunny/listener.ora
Listener Log File /ebs/oracle/VIS/db/tech_st/11.1.0/log/diag/tnslsnr/sunny/vis/alert/log.xml
Listening Endpoints Summary...
(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=sunny.localdomain)(PORT=1521)))
Services Summary...
Service "VIS" has 2 instance(s).
Instance "VIS", status UNKNOWN, has 1 handler(s) for this service...
Instance "VIS", status READY, has 1 handler(s) for this service...
Service "VIS_XPT" has 1 instance(s).
Instance "VIS", status READY, has 1 handler(s) for this service...
The command completed successfully
$
any help greatly appreciated .You can use the same stage area directory as long as nothing changed on this directory.
I am installing as oracle and choose oracle:dba as database owner and applmgr:dba as apps owner.
You need to run the installer as root user.
ORA-00600: internal error code, arguments: [2252], [2359], [322925629], [], [],
ORA-600 [2252] (Doc ID 138701.1)
Please clean your previous installation, reboot the server and start the installer again as root user.
Thanks,
Hussein -
Problem with /etc/project in Solaris 10
Bonjour,
I need your help to understand and have project work correctly under Solaris 10.
What seems to be prety straight forward on paper doesn't work at all for me and can't figure why so!
Obviously I'm trying to make it work for an Oracle user.
I create a project called "oracle" with id 100.
here is what you can found in my /etc/project :
oracle:100::oracle::process.max-file-descriptor=(priv,1024,deny);project.max-shm-memory=(priv,4294967296,deny)
here is what you can found in my /etc/user_attr :
oracle::::project=oracle
now, once I reboot and login as oracle I do a id -p, here is the output :
uid=3000(oracle) gid=3002(oinstall) projid=0(system)
I do ulimit -a, here is the output :
time(seconds) unlimited
file(blocks) unlimited
data(kbytes) unlimited
stack(kbytes) 8192
coredump(blocks) 0
nofiles(descriptors) 256 <------------------- should be 1024 no?
vmemory(kbytes) unlimited
Shouldn't I be part of project oracle (100)???
Now, of what I understant is oracle is NOT part of the "oracle" project and thuss setting are not set up at loging.
What am I doing wrong or what am I missing?
Thanks for your timeI'd start by verifying the settings via valid OS methods, no cat/grep whenever possible
# logins -xl oracle
oracle 3000 oinstall 3002
/export/home/oracle
/bin/sh
PS 051110 -1 -1 -1
# grep project /etc/nsswitch.conf
project: files
# grep oracle /etc/user_attr
oracle::::project=oracle
# projects oracle
default oracle
# projects -l oracle
oracle
projid : 100
comment: ""
users : oracle
groups : (none)
attribs: process.max-file-descriptor=(priv,1024,deny)
project.max-shm-memory=(priv,4294967296,deny)
# ssh -l oracle 0
Password:
Sun Microsystems Inc. SunOS 5.10 Generic January 2005
$ id -p
uid=3000(oracle) gid=3002(oinstall) projid=100(oracle) -
Unable to Start OracleAgent for Cache connect on Solaris 9
Please see below for configuration details. TimesTen 6.1 installed on Solaris 9 with access control enabled. DSN is defined in sys.odb.ini file under install_dir/info directory.
unable to start cache agent either with ttAdmin or with ttIsql. Error messages is as below. Please help!!
DSN defined in sys.odbc.ini
[ATS_tt60demo1]
Driver=/u03/TimesTen/tt60demo1/lib/libtten.so
DataStore=/u03/TimesTen/tt60demo1/info/DemoDataStore/ATSdemo1data
ExclAccess=0
Logging=1
#LogPurge=1
PermSize=16
ThreadSafe=1
Isolation=1
WaitForConnect=0
Authenticate=0
#LockLevel=0
UID=scott
OracleID=atsdev
OraclePWD=tiger
PassThrough=0
environment Variables
SSH_TTY=/dev/pts/3
PATH=/u03/TimesTen/tt60demo1/bin:/u03/TimesTen/tt60demo1/demo:/u03/TimesTen/tt60demo1/demo/ttclasses:/u03/TimesTen/tt60demo1/demo/xlaPersistent:/u03/TimesTen/tt60demo1/demo/tutorial/java:/u03/TimesTen/tt60demo1/3rdparty/ant/bin:.:/usr/bin:/usr/ucb:/usr/local/bin:/tools/bin:/tools/pvcs:/export/home/timesten/bin:/usr/xpg4/bin:/bin:/usr/sbin:/etc:/sbin:/u03/app/product/10.2.0/Db_1/bin
ORACLE_BASE=/u03/app/product
EDITOR=vi
HISTFILE=/export/home/timesten/.hist4229
CLASSPATH=/u03/TimesTen/tt60demo1/lib/classes14.jar:/u03/TimesTen/tt60demo1/lib/timestenjmsxla.jar:/u03/TimesTen/tt60demo1/3rdparty/jms1.1/lib/jms.jar:/u03/TimesTen/tt60demo1/demo:/u03/TimesTen/tt60demo1/demo/tutorial/java:/u03/TimesTen/tt60demo1/demo/tutorial/java/dist/lib/TimesTenJavaDemos.jar
LOGNAME=timesten
HOSTNAME=ncnc-sun5
PVCS_BINDIR=/tools/pvcs
ORACLE_SID=atsdev
PS1=${LOGNAME}@${HOSTNAME}:$PWD>
PS2=>>
USER=timesten
XMINCLUDE=/usr/dt/include
SHELL=/bin/ksh
HOME=/export/home/timesten
XFILESEARCHPATH=/usr/openwin/lib/%T/%N/%S:/usr/openwin/lib/%T/%N/%S
XMLIB=/usr/dt/lib
ANT_HOME=/u03/TimesTen/tt60demo1/3rdparty/ant
SSH_CLIENT=138.1.113.119 1178 22
LD_LIBRARY_PATH=/u03/TimesTen/tt60demo1/lib:/usr/openwin/lib:/usr/dt/lib:/usr/lib:/lib:/tools/lib:/u03/app/product/10.2.0/Db_1/lib:/u03/app/product/10.2.0/Db_1/network/lib
TERM=xterm
ORACLE_HOME=/u03/app/product/10.2.0/Db_1
THREADS_FLAG=native
PWD=/u03/TimesTen/tt60demo1/info
TZ=US/Eastern
ENV=/export/home/timesten/.kshrc
timesten@ncnc-sun5:/u03/TimesTen/tt60demo1/info> ttisql ATS_tt60demo1
Copyright (c) 1996-2005, Oracle. All rights reserved.
Type ? or "help" for help, type "exit" to quit ttIsql.
All commands must end with a semicolon character.
connect "DSN=ATS_tt60demo1";
Enter password for 'scott':
Connection successful: DSN=ATS_tt60demo1;UID=scott;DataStore=/u03/TimesTen/tt60demo1/info/DemoDataStore/ATSdemo1data;WaitForConnect=0;DRIVER=/u03/TimesTen/tt60demo1/lib/libtten.so;OracleId=atsdev;Authenticate=0;PermSize=16;
(Default setting AutoCommit=1)
Command> call ttCacheUidPwdSet('scott','tiger');
Command> call ttCacheStart();
10024: Could not start Oracle Agent for the requested data store.. Could not initialize OCI function pointers. Error: none
The command failed.
Command>exit
Disconnecting..
Done.Solved the problem myself. Had to stop and restart cache agent using the timesten software installation account and not as root.
Documentation is not clear as to what happens when you install timesten using a "non root" account. Per documentation, we have to run setuproot script to create startup / stop scripts under init directory. With a non-root installation these scripts are created and on reboot agent process inherits the root account privs. This is a problem. For a non-root install the agent process should inherit the privs of the account that performed the installation and not root's.
May be a documentation correction is needed here. -
More than 1 Priv IPs and 1 VIP in 10gr2 RAC installations
With 10gr2 clusterware, is it possible to add more than 1 Prv IP and VIP ?
If possible, what is the procedure to add?
Thank you.Is it possible what you really want to be looking for is NIC bonding / teamed NICs (or MPIP solaris).
This is where two ethernet adapters share an ip address
Take a look at the following post Re: teamed NICs for RAC interconnect for instance. (which refers to metalink Metalink note 298891.1 ).
Having said that I cant see a need for more than one priv IP, but multpile VIP's might jsut be wanted for eg DMZ. (I dont know the answer to that offahnd though).
rgds - bigdelboy. -
Oracle RAC 10g on Solaris x86 using vmware
Guys,
I am practising 10g RAC installation on my laptop, wherein I have vmware workstation and
solaris-x86 version 10 installed on it.
I am trying to install Oracle 10g clusterware.
I have followed the steps for clusterware installation.
I am facing an errror on one of the node while running root.sh.
On the first node, the root.sh runs fine.
The following is the log.
bash-3.00# ./root.sh
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
WARNING: directory '/u01/app' is not owned by root
WARNING: directory '/u01' is not owned by root
Checking to see if Oracle CRS stack is already configured
Setting the permissions on OCR backup directory
Setting up NS directories
Oracle Cluster Registry configuration upgraded successfully
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
WARNING: directory '/u01/app' is not owned by root
WARNING: directory '/u01' is not owned by root
Successfully accumulated necessary OCR keys.
Using ports: CSS=49895 CRS=49896 EVMC=49898 and EVMR=49897.
node <nodenumber>: <nodename> <private interconnect name> <hostname>
node 1: xsan001 xsan001-priv xsan001
node 2: xsan002 xsan002-priv xsan002
Creating OCR keys for user 'root', privgrp 'root'..
Operation successful.
Now formatting voting device: /dev/rdsk/c0d0s4
Format of 1 voting devices complete.
Startup will be queued to init within 30 seconds.
Adding daemons to inittab
Expecting the CRS daemons to be up within 600 seconds.
CSS is active on these nodes.
xsan001
CSS is inactive on these nodes.
xsan002
Local node checking complete.
Run root.sh on remaining nodes to start CRS daemons.
=============================
On the second node, the root.sh gives error.
bash-3.00# ./root.sh
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
WARNING: directory '/u01/app' is not owned by root
WARNING: directory '/u01' is not owned by root
Checking to see if Oracle CRS stack is already configured
Setting the permissions on OCR backup directory
Setting up NS directories
Oracle Cluster Registry configuration upgraded successfully
WARNING: directory '/u01/app/oracle/product' is not owned by root
WARNING: directory '/u01/app/oracle' is not owned by root
WARNING: directory '/u01/app' is not owned by root
WARNING: directory '/u01' is not owned by root
Successfully accumulated necessary OCR keys.
Using ports: CSS=49895 CRS=49896 EVMC=49898 and EVMR=49897.
node <nodenumber>: <nodename> <private interconnect name> <hostname>
node 1: xsan001 xsan001-priv xsan001
node 2: xsan002 xsan002-priv xsan002
Creating OCR keys for user 'root', privgrp 'root'..
Operation successful.
Now formatting voting device: /dev/rdsk/c0d0s4
Format of 1 voting devices complete.
Startup will be queued to init within 30 seconds.
Adding daemons to inittab
Expecting the CRS daemons to be up within 600 seconds.
Failure at final check of Oracle CRS stack.
10
Can you provide some sort of clue what could be done to resolve the above error?Refer to http://www.oracleracsig.org/ Under documents section select OS=Solaris. You will find a document on Oracle RAC on Solaris10 using Vmware.
-
Need guidance on using ddi_regs_map_setup solaris system call
Hi all,
Am a beginner in using Sun Solaris.
I am working on porting our source code from linux to solaris.
In Linux we are using inb/outb to read/write to keyboard data port(0x60) .
I know that there is an equivalent system call for inb/outb in solaris. But in man page(of sun os) it is mentioned that inb/ outb are obsolete. It was mentioned that we can use ddi_get8/ddi_put8 functions respectively.
ddi_get8 call requires ddi_acc_handle_t as a parameter which we can get it from ddi_regs_map_setup system call.
Syntax for this call is:
int ddi_regs_map_setup(dev_info_t *dip, uint_t rnumber,
caddr_t *addrp, offset_t offset, offset_t len,
ddi_device_acc_attr_t accattrp, ddi_acc_handle_t handlep);
Please let me know where does the 0x60 port number fits in the above system call and what does these parameters rnumber, offset, len corresponds to ?Hi,
The Fault_PC 0x13575e8 tells you exactly where you
die. I assume you have used this in mdb on
the core file to find you are dying in the ddi_get8 call?
Are you checking return value of ddi_regs_map_setup?
Are you sure the register you are accessing is in register set 1?
Are there other places prior to the panic where
the ddi_get/ddi_put routines work correctly? A bus error
here generally means the location you are trying to access
does not exist on the device. (It could also be alignment,
but not with ddi_get8). The address
looks like 1fe.02001400, which looks like a possibly
valid address in I/O space.
I would print the value of wc->ioaddr to make sure
it is a valid virtual address.
Hi,
I'm in the middle of porting a device driver for a
PCI card from linux to solaris (it's wcfxo from
zaptel - part of asterisk).
>
WARNING: [AFT1] Bus Error on System Bus in privileged
mode from CPU0 Data access at TL=0, errID
0x0000000b.985ccf88
AFSR 0x00000000.84000000<PRIV,BERR> AFAR
AFAR 0x000001fe.02001400
AFSR.PSYND 0x0000(Score 05) AFSR.ETS 0x00
0x00 Fault_PC 0x13575e8
UDBH 0x0000 UDBH.ESYND 0x00 UDBL 0x0000
0000 UDBL.ESYND 0x00
panic[cpu0]/thread=2a10007dd40: [AFT1] errID
0x0000000b.985ccf88 BERR Error(s)
See previous message(s) for detailsThe actual code which seems to cause the panic is:
#define inb(a) ddi_get8(wc->devhandle, a)
ints = inb(wc->ioaddr + WC_INTSTAT);where there initialisation includes:
static struct ddi_device_acc_attr dev_attr = {
DDI_DEVICE_ATTR_V0,
DDI_STRUCTURE_LE_ACC,
DDI_STRICTORDER_ACC
ddi_regs_map_setup(wc->dip, 1, &wc->ioaddr, 0, 255,
&dev_attr,
&dev_attr, &wc->devhandle);I've tested the same card in a linux x86 box, and it
works fine, so I'm happy it's not a hardware
problem.
Any suggestions on what might be causing this?
Thanks,
Simon -
Install ASM on Solaris RAC 10g
Hello,
I installed CRS and database software on two nodes RAC 10.2.0.4 Solaris x86-64 5.10, latest updates for Solaris. I have no error with crs.
Problems description:
1) When I run DBCA to create ASM it fails to create it on the nodes, with error ORA-03135: connection lost contact.
2) I see ORA-29702 into the logs (error occurred in Cluster Group Service operation
Cause: An unexpected error occurred while performing a CGS operation.
Action: Verify that the LMON process is still active. Also, check the Oracle LMON trace files for errors.)
Question:
Do you think that the problem is that interface 10.0.0.21 node1-priv-fail2 is started? (see bellow bdump/alert_+ASM1.log).
This interface 10.0.0.21 node1-priv-fail2 is the one that frozens the prompt when I try ssh oracle@node1-priv-fail2 from node2?
Possible solution: I saw Metalink 283684.1 but don't know if/what to change in my interfaces.
Details:
I think is something with the interfaces, but I don't know what.
- One thing I noticed is that is not possible to ssh from node1 to node2-priv-fail2 (this I wad told is the private standby loopback insterface). The same is from node2 to node1-priv-fail2, it gives a frozen prompt.
- in /etc/hosts on both nodes I have:
+127.0.0.1 localhost+
+172.17.1.17 node1+
+172.17.1.18 node1-fail1+
+172.17.1.19 node1-fail2+
+172.17.1.20 node1-vip+
+172.17.1.29 node2 loghost+
+172.17.1.30 node2-fail1+
+172.17.1.31 node2-fail2+
+172.17.1.32 node2-vip+
+10.0.0.1 node1-priv+
+10.0.0.11 node1-priv-fail1+
+10.0.0.21 node1-priv-fail2+
+10.0.0.2 node2-priv+
+10.0.0.12 node2-priv-fail1+
+10.0.0.22 node2-priv-fail2+
Do you think that the problem is that interface 10.0.0.21 node1-priv-fail2 is started (see bdump/alert_+ASM1.log).
This interface 10.0.0.21 node1-priv-fail2 is the one that frozens the prompt when I try ssh oracle@node1-priv-fail2 from node2?
Starting ORACLE instance (normal)
LICENSE_MAX_SESSION = 0
LICENSE_SESSIONS_WARNING = 0
Interface type 1 e1000g2 10.0.0.0 configured from OCR for use as a cluster interconnect
Interface type 1 e1000g3 10.0.0.0 configured from OCR for use as a cluster interconnect
Interface type 1 e1000g0 172.17.0.0 configured from OCR for use as a public interface
Interface type 1 e1000g1 172.17.0.0 configured from OCR for use as a public interface
Starting up ORACLE RDBMS Version: 10.2.0.4.0.
System parameters with non-default values:
large_pool_size = 12582912
instance_type = asm
cluster_database = TRUE
instance_number = 1
remote_login_passwordfile= EXCLUSIVE
++background_dump_dest = /opt/app/oracle/db/admin/+ASM/bdump++
++user_dump_dest = /opt/app/oracle/db/admin/+ASM/udump++
++core_dump_dest = /opt/app/oracle/db/admin/+ASM/cdump++
Cluster communication is configured to use the following interface(s) for this instance
+10.0.0.1+
+10.0.0.21+
node1:oracle$ oifcfg getif
e1000g0 172.17.0.0 global public
e1000g1 172.17.0.0 global public
e1000g2 10.0.0.0 global cluster_interconnect
e1000g3 10.0.0.0 global cluster_interconnect
node1:oracle$ ifconfig -a
lo0: flags=2001000849<UP,LOOPBACK,RUNNING,MULTICAST,IPv4,VIRTUAL> mtu 8232 index 1
inet 127.0.0.1 netmask ff000000
e1000g0: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 2
inet 172.17.1.17 netmask ffff0000 broadcast 172.17.255.255
groupname orapub
e1000g0:1: flags=9040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAILOVER> mtu 1500 index 2
inet 172.17.1.18 netmask ffff0000 broadcast 172.17.255.255
e1000g0:2: flags=1040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4> mtu 1500 index 2
inet 172.17.1.20 netmask ffff0000 broadcast 172.17.255.255
e1000g1: flags=39040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAILOVER,FAILED,STANDBY> mtu 1500 index 3
inet 172.17.1.19 netmask ffff0000 broadcast 172.17.255.255
groupname orapub
e1000g2: flags=1000843<UP,BROADCAST,RUNNING,MULTICAST,IPv4> mtu 1500 index 4
inet 10.0.0.1 netmask ff000000 broadcast 10.255.255.255
groupname oracle_interconnect
e1000g2:1: flags=9040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAILOVER> mtu 1500 index 4
inet 10.0.0.11 netmask ff000000 broadcast 10.255.255.255
e1000g3: flags=39040843<UP,BROADCAST,RUNNING,MULTICAST,DEPRECATED,IPv4,NOFAILOVER,FAILED,STANDBY> mtu 1500 index 5
inet 10.0.0.21 netmask ff000000 broadcast 10.255.255.255
groupname oracle_interconnectHi,
for a 10g RAC you need:
- a host-IP for every node
- a private IP for every node
- a virtual IP for every node
Host-IP and Private IP must be assigned to both hosts and conenction between the hosts using either the host-IP or the private IP must be possible.
Is is possible to build the RAC and ASM only with the private IP without the public and the virtual IP, if yes how?The term "private" and "public" does not refer to public IPs. It refers to the fact "private" is only for communication between the nodes and public for communication between the client and database.
For a successful installation you need at least these three IPs on each system.
So for instance your public IPs reside in the network 192.168.1.0/255.255.255.0 and your private interconnect network can be 192.168.2.0/255.255.255.0. Both networks consists of private (i.e. non-routeable) IPs. -
Automatic Startup in Oracle 10.2.02 on Solaris 10 X86
I have create the /etc/init.d/oratab from the sample of
Automating Database Startup and Shutdown on Other Operating Systems
by following script:
ORACLE_HOME=/u01/app/oracle/product/10.2.0/db_1
ORACLE=oracle
PATH=${PATH}:$ORACLE_HOME/bin
export ORACLE_HOME PATH
$ORACLE_HOME/bin/dbstart $ORACLE_HOME
But I got the ERROR log
SQL> ERROR:
ORA-12546: TNS:permission denied
SQL> ORA-12546: TNS:permission denied
SQL>
from $ORACLE_HOME/startup.log
What's Wrong with the dbora script ?$ORACLE_HOME/bin/dbstart $ORACLE_HOME
$ORACLE_HOME/bin/dbstart will be sufficient, script will read the oratab file (on Solaris usual location of oratab is in /var/opt/oracle) and starts all SID that have Y flag set.
To "user52":
check ur privs on folder;
be sure the permissions are 777.This is not good approach from security point of view, really - 755 is fine and sufficient. -
Application running in Solaris 9 container generating core files. what to do?
My solaris 9 zone configuration in solaris 10 looks like:
zonecfg:sms> info
zonename: sms
zonepath: /zone/sms
brand: solaris9
autoboot: true
bootargs:
pool:
limitpriv: default,proc_priocntl,proc_clock_highres,proc_lock_memory,sys_time,priv_proc_priocntl,priv_sys_time,net_rawaccess,sys_ipc_config,priv_proc_lock_memory
scheduling-class:
ip-type: exclusive
hostid:
[max-shm-memory: 4G]
[max-shm-ids: 100]
[max-sem-ids: 100]
fs:
dir: /var
special: /dev/dsk/c1t0d0s5
raw: /dev/rdsk/c1t0d0s5
type: ufs
options: []
net:
address not specified
physical: bge0
defrouter not specified
device
match: /dev/dsk/c1t0d0s5
device
match: /dev/rdsk/c1t0d0s5
device
match: /dev/dsk/c1t0d0s6
device
match: /dev/rdsk/c1t0d0s6
device
match: /dev/dsk/c1t0d0s7
device
match: /dev/rdsk/c1t0d0s7
capped-cpu:
[ncpus: 2.00]
capped-memory:
physical: 4G
[swap: 8G]
[locked: 2G]
attr:
name: hostid
type: string
value: 84b18f64
attr:
name: machine
type: string
value: sun4u
rctl:
name: zone.max-sem-ids
value: (priv=privileged,limit=100,action=deny)
rctl:
name: zone.max-shm-ids
value: (priv=privileged,limit=100,action=deny)
rctl:
name: zone.max-shm-memory
value: (priv=privileged,limit=4294967296,action=deny)
rctl:
name: zone.max-swap
value: (priv=privileged,limit=8589934592,action=deny)
rctl:
name: zone.max-locked-memory
value: (priv=privileged,limit=2147483648,action=deny)
rctl:
name: zone.cpu-cap
value: (priv=privileged,limit=200,action=deny)
Solaris 9 zone /etc/system file looks like:
* The directive below is not applicable in the virtualized environment.
* The directive below is not applicable in the virtualized environment.
* The directive below is not applicable in the virtualized environment.
* The directive below is not applicable in the virtualized environment.
* The directive below is not applicable in the virtualized environment.
* The directive below is not applicable in the virtualized environment.
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
set rlim_fd_max=65536
set rlim_fd_cur=60000
* The directive below is not applicable in the virtualized environment.
My questions are:
1. Application running in solaris 9 container generating core files. what to do???
2. My prstat -Z for zone shows almost 95% percent cpu usage. what to do???
3. Kindly can share how to move solaris 9 into solaris 10 containers ??Based on the new questions for the same post you wrote in the other communities, some posts are removed as duplicate, here is the answer :
For the point #3, please look on table 17-1 in the following URL :
Zone Components - System Administration Guide: Oracle Solaris Containers-Resource Management and Oracle Solaris Zones
You can also customize your container /etc/system file but it cannot exceeds the global zone and the zone configuration value.
For the other point, #2, this can be complicated without a complete image of what the complete system do.
Try fist to check if you have not a busy process in your zone, then you can check if a bottleneck exists in the I/O side. You use maybe wrong parameters, a wrong configuration or your system configuration is insufficient in term of resources.
What I can see in the outputs that you provided is that the S9 zone uses the half of the swap space. This can impact your zone performance and I/O activity, and can have in this case a side effect on some processes. Check why your zone uses the swap and how you can remedy this. -
Solaris 8 rebooting with very little info as to why.
I have a Sun E-250 with 2 400MHz CPUs and 1GB of memory. Lately my box will reboot without warning. I get a panic message in the logfile /var/adm/messages with very little other information. Those messages are as follows:
Feb 22 10:48:01 storrs SUNW,UltraSPARC-II: [ID 727572 kern.warning] WARNING: [AFT1] EDP event on CPU0 Instruction access at TL=0, errID 0x00000262.3179c556
Feb 22 10:48:01 storrs AFSR 0x00000000.80400008<PRIV,EDP> AFAR 0xffffffff.ffffffff
Feb 22 10:48:01 storrs AFSR.PSYND 0x0008(Score 95) AFSR.ETS 0x00 Fault_PC 0x100db82c
Feb 22 10:48:01 storrs UDBH 0x0000 UDBH.ESYND 0x00 UDBL 0x0000 UDBL.ESYND 0x00
Feb 22 10:48:01 storrs SUNW,UltraSPARC-II: [ID 868619 kern.info] [AFT2] errID 0x00000262.3179c556 No error found in ecache (No fault PA available)
Feb 22 10:48:01 storrs unix: [ID 836849 kern.notice]
Feb 22 10:48:01 storrs ^Mpanic[cpu0]/thread=30001fa8f00:
Feb 22 10:48:01 storrs unix: [ID 249780 kern.notice] [AFT1] errID 0x00000262.3179c556 EDP Error(s)
Feb 22 10:48:01 storrs See previous message(s) for details
Has anyone ever seen this before? Anybody have a clue as what to look into?
Also, this box seems to be very slow in performing relatively simple commands. It gets hung up on "format" and "pkgadd". So it appears that anything I do which may affect or use the kernal hangs for a long period of time and then most times the system reboots.
Please help if you can.I'm not quite sure what the error is about, but check the version of Solaris your running on the server. If it's less than Solaris 8, check that the server was built using the "Special Operating Environment CD".
This CD has to be used when installing solaris 2.5, 2.6 or 2.7 ( I think ) on the new hardware like E220, E250, E420 etc. -
P2v migration,brandzone solaris10 on solaris 11-Exiting with exit code 254
hi...
doing p2v migration on vmware : solaris 10 as branded zone on solaris 11
im moving my physical machine solaris 10 to brand zone as in soialris 11
some how im facing the issues..while zone installation
it fails at the end
follwing are the logs::
root@solaris11:/var/log/zones# tail -f zoneadm.20120304T231155Z.s10zone.install
[Monday, March 5, 2012 04:42:03 AM IST] | install_flar
[Monday, March 5, 2012 04:42:07 AM IST] setting ZFS property zoned=on on zpool1/zones/rpool
[Monday, March 5, 2012 04:42:10 AM IST] setting ZFS property canmount=on on zpool1/zones/rpool
[Monday, March 5, 2012 04:42:10 AM IST] setting ZFS property mountpoint=/rpool on zpool1/zones/rpool
[Monday, March 5, 2012 04:42:14 AM IST] setting ZFS property canmount=noauto on zpool1/zones/rpool/ROOT
[Monday, March 5, 2012 04:42:14 AM IST] setting ZFS property mountpoint=legacy on zpool1/zones/rpool/ROOT
[Monday, March 5, 2012 04:42:15 AM IST] Creating active_ds zpool1/zones/rpool/ROOT/zbe-0
[Monday, March 5, 2012 04:42:16 AM IST] Mounting boot environment in zpool1/zones/rpool/ROOT/zbe-0 at /zpool1/zones/root (including child datasets)
[Monday, March 5, 2012 04:42:17 AM IST] Preparing to mount zpool1/zones/rpool/ROOT/zbe-0 at /zpool1/zones/root
[Monday, March 5, 2012 04:42:17 AM IST] Mounting zpool1/zones/rpool/ROOT/zbe-0 at /zpool1/zones/root/ with ZFS temporary mount
[Monday, March 5, 2012 05:23:37 AM IST] ERROR: Unpacking the archive failed
[Monday, March 5, 2012 05:23:37 AM IST] Destroying datasets under zpool1/zones that are not pinned
[Monday, March 5, 2012 05:24:57 AM IST] Destroyed unpinned dataset zpool1/zones/rpool/ROOT/zbe-0
[Monday, March 5, 2012 05:24:57 AM IST] Unpinning datasets under zpool1/zones
[Monday, March 5, 2012 05:24:58 AM IST] Unpinning zpool1/zones
[Monday, March 5, 2012 05:24:58 AM IST] Unpinning zpool1/zones/rpool
[Monday, March 5, 2012 05:24:58 AM IST] Unpinning zpool1/zones/rpool/ROOT
[Monday, March 5, 2012 05:24:58 AM IST] Unpinning zpool1/zones/rpool/export
[Monday, March 5, 2012 05:24:58 AM IST] Unpinning zpool1/zones/rpool/export/home
[Monday, March 5, 2012 05:24:58 AM IST] Exiting with exit code 254
[Monday, March 5, 2012 05:24:58 AM IST] ==== Completed: ====
zone configuration :
root@solaris11:/var/log/zones# zonecfg -z s10zone
zonecfg:s10zone> info
zonename: s10zone
zonepath: /zpool1/zones
brand: solaris10
autoboot: false
bootargs:
file-mac-profile:
pool:
limitpriv:
scheduling-class:
ip-type: exclusive
hostid: 0005b1bc
fs-allowed:
[max-lwps: 32500]
[max-processes: 16250]
attr:
name: zonep2vchk-info
type: string
value: "p2v of host test"
attr:
name: num-cpus
type: string
value: "original system had 2 cpus"
rctl:
name: zone.max-processes
value: (priv=privileged,limit=16250,action=deny)
rctl:
name: zone.max-lwps
value: (priv=privileged,limit=32500,action=deny)
will provide the truss o/p if needed..
my solaris 11 having the system/zones/brand/brand-solaris10
please guide me to clear the issue...any doc will really helpful....
thanks again
muvvas
Edited by: user13152599 on Mar 5, 2012 8:17 AMsource machine:::
bash-3.2# cat /etc/release
Solaris Express Community Edition snv_101a X86
Copyright 2008 Sun Microsystems, Inc. All Rights Reserved.
Use is subject to license terms.
Assembled 31 October 2008
and its ufs
flarcreate -S -n s10-system /nfs/s10-system.flar
on target solaris 11::i have executed the following cmds::;
zonecfg -z s10-zone -f zone.config
zoneadm -z s10-zone install -u -a /export/home/muvvas/s10-system.flar
rpool 20G 39K 9.3G 1% /rpool
zpool1 47G 32K 40G 1% /zpool1
zpool1/zones 47G 6.5G 40G 14% /zpool1/zones -
Misconfigured rcap on Solaris 10 with zones - SMF fails to start rcap
Hi,
I have just installed Solaris 10 on a brand new T5220 server (Niagara processor, 16GB of RAM). After creating first non-global zone with capped memory and cpu, SMF fails to start rcap service:
+root@bs10 [~]$ svcs -xv+
svc:/system/rcap:default (resource capping daemon)
State: maintenance since Mon Nov 03 10:28:53 2008
Reason: Start method exited with $SMF_EXIT_ERR_CONFIG.
See: http://sun.com/msg/SMF-8000-KS
See: man -M /usr/share/man -s 1M rcapd
See: man -M /usr/share/man -s 1M rcapstat
See: man -M /usr/share/man -s 1M rcapadm
See: /var/svc/log/system-rcap:default.log
Impact: This service is not running.
Content of service's log:
+root@bs10 [~]$ tail /var/svc/log/system-rcap:default.log+
+[ Nov 3 09:36:21 Method "start" exited with status 96 ]+
+[ Nov 3 10:25:37 Leaving maintenance because clear requested. ]+
+[ Nov 3 10:25:37 Enabled. ]+
+[ Nov 3 10:25:37 Executing start method ("/usr/lib/rcap/rcapd") ]+
+[ Nov 3 10:25:37 Method "start" exited with status 0 ]+
+[ Nov 3 10:26:59 Stopping because service disabled. ]+
+[ Nov 3 10:27:00 Executing stop method (:kill) ]+
+[ Nov 3 10:28:53 Executing start method ("/usr/lib/rcap/rcapd") ]+
svc.startd could not set context for method: chdir: No such file or directory
+[ Nov 3 10:28:53 Method "start" exited with status 96 ]+
Clearing service status helps, but this error is reported after every system restart. FSS is set as the default system's scheduling class.
Zone configuration:
zonecfg:bs10z1> info
zonename: bs10z1
zonepath: /zones/bs10z1
brand: native
autoboot: true
bootargs:
pool:
limitpriv:
scheduling-class: FSS
ip-type: exclusive
inherit-pkg-dir:
dir: /platform
inherit-pkg-dir:
dir: /install
fs:
dir: /opt/mirror1/oracle
special: /dev/md/dsk/d105
raw: /dev/md/rdsk/d105
type: ufs
+options: [rw]+
net:
address not specified
physical: e1000g1
capped-cpu:
+[ncpus: 10.00]+
capped-memory:
physical: 6G
+[swap: 6G]+
+[locked: 4G]+
rctl:
name: zone.cpu-cap
value: (priv=privileged,limit=1000,action=deny)
rctl:
name: zone.max-swap
value: (priv=privileged,limit=6442450944,action=deny)
rctl:
name: zone.max-locked-memory
value: (priv=privileged,limit=4294967296,action=deny)
What have I missed?
Edited by: Maciejos on Nov 3, 2008 2:29 AMThe reason of this behave: the root's home directory was located on the filesystem different than root filesystem. After moving it everything works fine.
Maybe you are looking for
-
Hi, Recently my MacBook Pro's 17" late 2007 A1229 graphics card died. I bought a refurbished one on eBay. The guy that solds the logic boards has tested it and it works (he got a whole refurbisched Mac parts company). So with my old logic board, when
-
Itouch not sending audio to car speakers
I recently added an AUX cable to my Alpine car stereo, but my ipod touch will not send audio to the car speakers. When its plugged in to the cable it just plays through the built in speaker on the ipod. The cable is the older 12v type, so I underst
-
Free Programmed Search Help Question
I want to use the free programmed search help interface to make my own search help. But the problem is, I want to embed SAP's search help into my own search help.How can I do that? Thank you.
-
Exception in thread "main" java.lang.OutOfMemoryError: Java heap space
i have a huge XML file of size 9 MB. I need to parse it and store it in a MYSQL database. When i execute the program, it throws me out of memory error and only 12668 rows have been inserted into the database. I use SAX parser to parse the XML. i have
-
Cfthread of coldfusion on different systems-windows and RedHat
Recently I use cfthread writing a multithread samples. Each thread ask a http request.When I run this cfc file with coldfusion9 on windows, it runs well. Then I run the same cfc file on coldfusion9 on linux(RedHat and centOS), unfortunately, it seems