Solaris System Admin tips

Solaris Operating System for x86 Installation Check Tool

2007-09-21T12:45:00.000+01:00

Solaris Operating System for x86 Installation Check Tool checks the system on which you try to install the operating system and report if the system can run the Sun solaris OS properly.

The download is a simple ISO file which then needs to burnt onto a CD to create a boot CD. Once created, start the X86 system with the CD. The ISO can be downloaded from here.

For Sun Solaris Express Edition, click here

The tool checks and reports whether your x86 system can install the Solaris 10 8/07 OS. The Installation Check Tool loads the Solaris 10 8/07 kernel and looks for a hard disk on your system. Reports whether Solaris drivers are available for the devices the tool detects on your x86 system. This information helps you to determine whether the Solaris 10 8/07 OS is likely to run on your system.

Installation Check Tool also produces a table that shows whether a Solaris driver is available for each device the tool detects. The table tells you whether the driver is built in to the Solaris OS or whether a third-party driver is available.

Installation Check Tool first loads the Solaris 10 8/07 kernel. Then the tool generates a report that shows the major PCI components in the system and whether Solaris drivers exist for these devices.

This report package is written to a tar file that is compressed with gzip. You have the option to copy this report file to a floppy disk or to USB storage. The report package includes the following files:

A text file that shows the same information as the screen output.

An HTML report file that shows the full content of the report. No abbreviations are used.

The output of prtconf -pv, prtdiag, and prtconf -v.

The following is an example of the text information that is displayed on your monitor:

If the system includes two devices that require a third-party driver, the driver for the second such device is labeled “T-2” in the driver column of the table and in the Third Party Driver Legend.
If a report such as the example shown above is displayed on your screen, and if you do not receive any warning message that no disk is found, then the Solaris 10 8/07 OS can be installed on this system.
The next step is to check whether drivers exist for other on-board devices, such as network devices, video devices, and audio devices.

NOTE: Sun does not warrant the behavior of any third-party driver.

Installation Check Tool gives you the option to copy this driver report to a floppy disk or to USB storage.

Solaris Operating System for x86 Installation Check Tool displays a message if your system is suitable for installing the Solaris 10 8/07 OS.

Installation Check Tool is updated periodically to include knowledge of new drivers in the driver database that the tool uses. Drivers that are integrated into the Solaris OS and third-party drivers that are known to work with the Solaris OS are included in these updates. Check this web page periodically to make sure you are using the most current release of Installation Check Tool. The tool also displays a message to remind you to check for a new release.

Sun Device Detection Tool 1.2 for X86 systems

2007-09-21T12:23:00.000+01:00

The Sun Device Detection Tool can help you to determine if the Devices on the system is detected by the Sun Solaris Operating System.

If your device or system is already on the Hardware Compatibility List (HCL) then this may not be necessary. Click here for the HCL in Sun Solaris.

However, there is no harm in running the tool. The tool also allows you to submit the device to HCL under the Reported to Work Function (RWF).

The tool can test the X86 system for either Sun Solaris OS or for the Sun Solaris Express Edition.

The Sun Device Detection Tool 1.2 can be downloaded from here and can run on X86 systems running Windows XP, Linux (with PCI tools installed) with Sun Java Runtime Environment (JRE) installed.

Acccept the License Agreement and download the tool.

1. Once downloaded, click Run to start the tool.

2. Select the Operating System (Solaris OS or Solaris Express Edition) and click Start. This starts the tool and checks all the detected tool in the system.

3. Once complete, the tool will provide a list of device detected and the status of the drivers available.

A Green indicates the device is compatible with the Sun Solaris OS

A Red indicates that the device is not compatible with Sun Solaris or in other words there is no driver for the device

A Yellow indicates tha the device has limited support and may or may work properly with Sun Solaris Operating System.

Harware Compatibility List (HCL) for Solaris 10 & Solaris Express Edition

2007-09-21T11:02:00.000+01:00

Every Operating System has a Hardware Compatibility List (HCL). The Hardware Compatibility List (HCL) identifies hardware that is compatible with the Operating System and provides information about available software support.

Hardware Compatibility List (HCL) for Solaris 10 OS:

http://www.sun.com/bigadmin/hcl/data/sol/

There are totally 906 system which includes OEM & System Vendor Products, VMWare,Servers, Desktops, Laptops andMotherboards and 1568 system components

Hardware Compatibility List (HCL) for Solaris Express Developer Edition:

http://www.sun.com/bigadmin/hcl/data/sx/

This includes 975 Systems and 1602 system components

How to Reset/Recover the Root password in Sun Solaris (x86)

2007-03-30T16:07:00.000+01:00

Click here for procedure on SPARC platform

There is every little chance that one loses or rather forgets the root password of his Sun Solaris servers.

In the event, this happens, there is a way out of it. Well the way is to reset the password as there is no way to recover it. Ofcourse, it is recommeded that the security for the physical access to the server is restricted so as to ensure that there is no unauthorized access and anyone who follows this routine is an authorized personnel.

Boot the server with a Sun Solaris Operating System CD (I'm using a Solaris 10 CD but doesn't matter really) or a network boot with a JumpStart server. Change the Boot order accordingly in your x86 system and start the server and launch a single user mode (No Password).

In Solaris 10, you have the default Fail Safe boot option in the Boot Loader. With this you do not need a CD or JumpStart server as selecting it will launch the Single-User shell. I haven't used the early Solaris versions on an x86 system so if anyone knows it is availabe then please post your comments.

When you boot from the CD, select the option for the Sigle-user mode (Option 6 on Solaris 10 CD)

This will look for the currently installed Solaris OS on your system and prompts a Yes/No question to mount the Root filesystem as a Read/Write file system onto /a.

Select yes to mount the root file system (/dev/dsk/c0t0d0s0 here) on /a. If you select No, no harm as you still can mount it manually using:

mount /dev/dsk/c0t0d0s0 /a

NOTE: /a is a temporary mount point that is available when you boot from CD or a JumpStart server

Now, with the root file system mounted on /a. All you need to do is to edit the shadow file and remove the encrypted password for root.

# vi /a/etc/shadow

Now, exit the mounted filesysten, unmount the root filesystem and reboot the system to single-user mode booting of the disk.

# cd /

# umount /a

# init s

This should boot of the disk and take you to the single-user mode. Press enter at the prompt to enter a password for root.

This should allow you to login to the system. Once in, set the password and change to multi-user mode.

NOTE: Single-User mode is only to ensure that the root user without password is not exposed to others if started in multi-user mode before being set with a new password.

# passwd root

# reboot

This should do.

Click here for procedure on SPARC platform

How to reset the Root password in Sun Solaris (SPARC)

2007-03-30T15:32:00.000+01:00

Click here for procedure on x86 platform

There is every little chance that one loses or rather forgets the root password of his Sun Solaris servers. In the event, this happens, there is a way out of it. Well the way is to reset the password as there is no way to recover it

Ofcourse, it is recommeded that the security for the physical access to the server is restricted so as to ensure that there is no unauthorized access and anyone who follows this routine is an authorized personnel.

Boot the server with a Sun Solaris Operating System CD (I'm using a Solaris 10 CD but doesn't matter really) or a network boot with a JumpStart server from the OBP OK prompt.

OK boot cdrom -s

or

OK boot net -s

This will boot the server from the CD or Jumpstart server and launch a single user mode (No Password).

Mount the root file system (assume /dev/dsk/c0t0d0s0 here) onto /a

# mount /dev/dsk/c0t0d0s0 /a

NOTE: /a is a temporary mount point that is available when you boot from CD or a JumpStart server

Now, with the root file system mounted on /a. All you need to do is to edit the shadow file and remove the encrypted password for root.

# vi /a/etc/shadow

Now, exit the mounted filesysten, unmount the root filesystem and reboot the system to single-user mode booting of the disk.

# cd /

# umount /a

# init s

This should boot of the disk and take you to the single-user mode. Press enter at the prompt to enter a password for root.

This should allow you to login to the system. Once in, set the password and change to multi-user mode.

NOTE: Single-User mode is only to ensure that the root user without password is not exposed to others if started in multi-user mode before being set with a new password.

# passwd root

# reboot

This should do.

Click here for procedure on x86 platform

A good informative comparison of Sun Solaris Operating System features

2007-03-28T15:25:00.000+01:00

http://www.sun.com/software/solaris/whats_new.jsp

This page on the Sun website has a simple but informative comparision chart for Sun Solaris 7/8/9/10 operating systems based on the following features:

Availability
Data Management
Install/Upgrade
Networking/Interoperability
Observability
Performance
Platform Choice
Secuirty
Usability
Virtualization

Click here to go to the page on Suns website

Different Sun Solaris operationg system versions

2007-03-27T20:01:00.001+01:00

The table shown in the article shows all the versions and release details of the Sun Solaris Operating system. Call it a little simmary of the big long journey of Sun Solaris as an Operating System.

SunOS version	Solaris version	Release date	Supported platforms
4.0.2	none	Sep. 89	386i
4.0.3	none	May 89	sun2, sun3/3x, sun4
4.0.3c	none	June 89	Sparc 1
4.0.3 PSR_A	none	July 89	Sun 4/470, 4/490
4.1	none	Mar. 90	sun3, sun4
4.1e	none	Apr. 91	sun4e
4.1.1	none	Mar. 90	sun3/3x, sun4
4.1.1B	1.0	Feb. 91	sun4
4.1.1.1	1.0	Jul. 91	sun3/3x
4.1.1_U1	1.0	Nov. 91	sun3/3x
4.1.2	1.0.1	Dec. 91	sun4, sun4m
4.1.3	1.1A	Aug. 92	sun4, sun4c, sun4m
4.1.3C	1.1c	Nov. 93	Sparc LX/Classic
4.1.3_U1	1.1.1	Dec. 93	sun4, sun4c, sun4m
4.1.3_U1B	1.1.1B	Feb. 94	sun4, sun4c, sun4m
4.1.4	1.1.2	Nov. 94	sun4, sun4c, sun4m
5.0	2.0	Jul. 92	sun4c
5.1	2.1	Dec. 92	sun4, sun4c, sun4m, x86
5.2	2.2	May 93	sun4, sun4c, sun4m, sun4d
5.3	2.3	Nov. 93	sun4, sun4c, sun4m, sun4d
5.4	2.4	Aug. 94	sun4, sun4c, sun4m, sun4d, x86
5.5	2.5	Nov. 95	sun4c, sun4m, sun4d, sun4u, x86
5.5.1	2.5.1	May 96	sun4c, sun4m, sun4d, sun4u, x86, ppc
5.6	2.6	Aug. 97	sun4c, sun4m, sun4d, sun4u, x86
5.7	7	Oct. 98	sun4c, sun4m, sun4d, sun4u, x86
5.8	8	2000	sun4m, sun4d, sun4u, x86
5.9	9	2002	sun4m, sun4d, sun4u, x86
5.10	10	2005	sun4m, sun4d, sun4u, x86

Install MySQL in Solaris

2007-03-27T19:59:00.000+01:00

I had a hard time in getting some proper knowledge base article to get MySQL installed on my Solaris today. However, I finally made up and running on my server and thought people should have access to this information to make sure they do not waste time like I did.

Installation of MySQL on Solaris is not that difficult but is a bit dragging and not as straightforward as is in any other Linux flavors.

To install MySQL on Solaris follow the procedures listed below:

I've used MySQL-5.0.24a for Solaris 8. However, this is a very generic procedure and I'm sure will work on other Solaris versions.

Download the MySQL Package for solaris from

http://dev.mysql.com/downloads/mysql/5.0.html

Once downloaded unzip using

# gzip mysql-standard-5.0.24a-solaris8-sparc-64bit.tar.gz

Now, comes the first problem, you need to have GNU Tar installed on your Solaris server as there is bug in getting the classic Solaris one to work.

To download and install GNUTar, please click here.

Untar using GNUTar by

# tar -xvf mysql-standard-5.0.24a-solaris8-sparc-64bit.tar

Now, install using

# pkgadd -d mysql-standard-5.0.24a-solaris8-sparc-64bit.pkg

This installs the software. You can be confirmed using

# pkginfo -l mysql

This will give a detailed output describing the installation.

Now, change dir to /opt/mysql/mysql/support-files

# cd /opt/mysql/mysql/support-files

Copy the my-middle.cnf as my.cnf and mysql.server to /opt/mysql/mysql/scripts

# cp my-middle.cnf ../scripts/my.cnf
# cp mysql.server ../scripts/

Now, change to /opt/mysql/mysql/scripts folder

# cd opt/mysql/mysql/scripts

Edit file my.cnf using vi

#vi my.cnf

Add the following lines to the file under [mysqld]

datadir=/opt/mysql/mysql/data
basedir=/opt/mysql/mysql
log
log-update

Similary, edit the mysql.server file and add the following lines:

# vi mysql.server

basedir=/opt/mysql/mysql
datadir=/opt/opt/mysql/mysql/data
conffile=/opt/mysql/mysql/scripts/my.cnf
mysql_daemon_user=mysql

Now, all done, run the mysql_install_db script as follows

# ./mysql_install_db --defaults-file=/opt/mysql/mysql/scripts/my.cnf

This creates the initial MySQL database.

Now, start MySQL using

cd . ; /opt/mysql/mysql/bin/mysqld_safe &

This will start MySQL.

Now, you need to set the password for root account

# /opt/mysql/mysql/bin/mysqladmin -u root password 'newpassword'
# /opt/mysql/mysql/bin/mysqladmin -u root -h password 'newpassword'

This completes a successful installation of MySQL on your Solaris Server.

How to install GNU tar in Solaris

2007-03-27T19:57:00.000+01:00

This is a simple article on how to get GNU Tar installed on a Solaris server.

There are scenarios where you cannot untar a certain package with the builtin tar utility. A best example would be when you use it to untar the MySQL Packages downloaded from http://dev.mysql.com/downloads/mysql/5.0.html

Download the GNU tar (tar-1.15.1) for your Solaris version and platform (SPARC/X86) from,

http://www.sunfreeware.com

Once downloaded, change to the directory where the file is downloaded (ex: /export/software/)

# cd /export/software

Now, unzip using,

# gzip -d tar-1.15.1-sol8-sparc-local.gz

NOTE: File Name varies as per your solaris version/platform

Once done, install the package using,

# pkgadd -d tar-1.15.1-sol9-sparc-local

NOTE: Be sure that you change to the directory and and just enter the package name after -d

Answer the prompts for confirmation and creation of new folder (if required).

This completes the installation.

To check the installation,

# pkginfo | grep tar

This will show up as

application SMCtar GNU tar..

To confirm with more details,

#pkginfo -l SMCtar

This should do.

Return codes during Patch Installation

2007-03-27T19:54:00.000+01:00

When you install Patches on Solaris, you will find quite often see error codes or Exit/Return codes when the patch installation fails.

These are the return codes that indicate the cause of the failure to install the specified patch and are very helpful as point to start with in troubleshooting.

The most commonly seen return codes are Return code 2, Return code 8, Return code 25, Return code 35.

Return Code 2 indicates that the patch you are trying to install is already been applied.

Return Code 8 indicate that the patch you are trying to install is for a package that is not already installed.

Return Code 25 indicates that a patch required for this patch to be installed in not found.

Return Code 35 indicates that the revision of the patch is older and that a later version of the patch is already applied to the system.

The following are a list of the Return/Exit codes that you may come across while patching your Solaris server:

Exit Code / Return Code	Description
0	No error
1	Usage error
2	Attempt to apply a patch that's already been applied
3	Effective UID is not root
4	Attempt to save original files failed
5	pkgadd failed
6	Patch is obsolete
7	Invalid package directory
8	Attempting to patch a package that is not installed
9	Cannot access /usr/sbin/pkgadd (client problem)
10	Package validation errors
11	Error adding patch to root template
12	Patch script terminated due to signal
13	Symbolic link included in patch
14	NOT USED
15	The prepatch script had a return code other than 0
16	The postpatch script had a return code other than 0
17	Mismatch of the -d option between a previous patch install and the current one
18	Not enough space in the file systems that are targets of the patch
19	$SOFTINFO/INST_RELEASE file not found
20	A direct instance patch was required but not found
21	The required patches have not been installed on the manager
22	A progressive instance patch was required but not found
23	A restricted patch is already applied to the package
24	An incompatible patch is applied
25	A required patch is not applied
26	The user specified back out data can't be found
27	The relative directory supplied can't be found
28	A pkginfo file is corrupt or missing
29	Bad patch ID format
30	Dryrun failure(s)
31	Path given for -C option is invalid
32	Must be running Solaris 2.6 or greater
33	Bad formatted patch file or patch file not found
34	The appropriate kernel jumbo patch needs to be installed
35	Later revision already installed

Platform tips for Solaris (one Liners)

2007-03-27T19:49:00.000+01:00

These are simple one liner tips for solaris.

Processor Information:

The psrinfo command displays the Processor info on the Sun Server

# psrinfo -v

Status of virtual processor 0 as of: 08/31/06 12:27:07
on-line since 08/28/06 01:46:34.
The sparcv9 processor operates at 296 MHz,
and has a sparcv9 floating point processor.

Open Boot Prompt (OBP) revision:

The prtconf command displays the Open Boot Prompt revision

#prtconf -V

OBP 3.29.0 2000/12/20 18:41

3.29.0 indicates the revision of the Open Boot Prompt.

Patch Revision info:

The showrev command displays the revision of a particular patch. This helps when you install a version of the patch to check if it is the latest and/or check if it is obsolete.

#showrev -p |grep

System Type:

The uname command can be used to find the hardware type.

#uname -i

SUNW,Ultra-60

The above command is run on a SPARC Ultra60

Solaris revision:

Use the uname command to find the revision number of your Solaris Operating System.

#uname -v

Generic_108528-29

where 108528-29 is the Solaris patching revision.

Please note, Newly installed patch revisions will not show until the server is rebooted.

Solaris Server Version:

The uname command can be used to find the Solaris operating system version.

#uname -r

5.8

#uname -a

SunOS myserver 5.8 Generic_108528-29 sun4u sparc SUNW,Ultra-60

where 5.8 indicates Solaris 8.

RSH: Permission Denied error

2007-03-27T13:57:00.000+01:00

When you run a command remotely from one Solaris Server (say SERVERA) to a remote server (say SERVERB) using RSH you might encounter the error

RSH: Permission denied

This could happen even though you have the host entry in the .rhosts file of the user.

To resolve this problem. Try to do a RSH / Telnet / SSH onto the other server entering a username and password.

Once logged in, run the command

#who am i

This will show the machines actual name/ip that the remote machine sees. Enter this name in the .rhosts file and that should fix the problem.

However, it is not very advisable to use RSH for security reasons and suggest to migrate to SSH on your Solaris boxes which more stable and secure.

/var/adm/messages

2007-03-27T13:42:00.000+01:00

The /var/adm/messages is the file to which all the messages printed on the console are logged to by the Operating System. This helps to track back check the console messages to troubleshoot any issues on the system.

Syslog daemon also writes to this /var/adm/messages file.

The /var/adm/messages file monitored and managed by newsyslog and its configuration file is /usr/lib/newsyslog

click here for a sample view of the newsyslog file.

This script runs as the roots cron job everyday, checks the /var/adm/messages file and copies/moves it to /var/adm/messages.0,1,2,3,4,5,6,7. In other words, it does the Log Rotation for the /var/adm/messages.
In an event the /var file system is running out of space, these files needs to checked and can be removed (not the actual /var/adm/messages itself) to free up space on the file system.
However, care has to be taken, if you decide to empty the /var/adm/messages itself for any reason. This process is called Truncation.

Truncate /var/adm/messages

The best way to truncate the /var/adm/messages is as follows:

# cp /var/adm/messages /var/adm/messages.bak

# cp /dev/null /var/adm/messages

NOTE: >/var/adm/messages does the same as what the Step 2 does.

To explain, all you do is to copy the contents of the existing /var/adm/messages to a new file say /var/adm/messages.bak and then empty the contents of messages by overwriting it with /dev/null which is empty. This will allow syslog to still continue writing to /var/adm/messages file.

There is another way to do it but is not as graceful to syslog as the previous one wherein you move the current /var/adm/messages file to a new one say /var/adm/messages.bak. When this happens, the /var/adm/messages no longer exists and this troubles syslog wherein you need to re-initialize the syslogd again. Not sure why, if you check the /usr/lib/newsyslog, Solaris follows this procedure than the safer earlier one.

To perform this operation:

# mv /var/adm/messages /var/adm/messages.bak

# touch /var/adm/messages

# kill –1 `cat /etc/syslog.pid`

If you do not want Solaris to do the Log rotate and use a tool logrotate then you may do so by simply disabling the newsyslog from the roots cron and your logrotate tool to do the job.

/dev/rmt0 Device busy

2007-03-27T00:34:00.000+01:00

If you are an Administrator and especially a backup administrator there is a very high chance that you get stuck with this error message with your backup.

If you do a

# mt -f /dev/rmt0n status

(infact any option with the mt or related command)

and end up with the error

/dev/rmt0n: Device Busy

then the most probable problem could be your Tape drive is still being used by the backup program or some related process.

The first thing to check would which process has locked your Tape drive. This can be done using the fuser utility.

# fuser -u /dev/rmt/0n

This will output the process ID that is actually holding the /dev/rmt0n device. The fuser can also be used against files.

Now, use

# ps -ef | grep <PID>

This will show the process. Now, kill the process using

# kill <PID>

If this fails, then use

# kill -9 <PID>

if still fails, use

#pkill -9 <PID>

That should do. You can check the status by using

# mt -f /dev/rmt0n status

How to flush DNS in Solaris

2007-03-27T00:27:00.000+01:00

It may be sometime required to flush your DNS Cache mostly when trying to troubleshoot a problem. There is no direct way or a command that can do this however, you are just lost yet. yes, still have a way to do it.

The NSCD daemon in Solaris and ofcourse in UNIX and most of the Linux falvors maintains the system cache including the DNS/Name service cache.

All it needs is to restart the Daemon to flush the DNS cache.
Do a

# ps -ef | grep nscd

This will find the PID for the NSCD daemon.

Now, do a

#pkill <PID>

#kill <PID>

Now, check if the nscd process is killed by using

# ps -ef | grep nscd

Now, start the nscd daemon by using

#/usr/sbin/nscd

This should start the nscd daemon and there you go, you have flushed the DNS Cache in your Solaris Server.

Click here for NSCD daemon man page.

du and df show different values for a file system

2007-03-26T22:37:00.000+01:00

If you are a UNIX/LINUX administrator then you would have possibly come across or will see quite regularly that du and df do not always agree on the size of the filesystem when measured with most of the commands.

In most known scenarios, df will be reporting less available space than the du command.

The du comman counts file by file and thereby count the total space on it. On the other hand, the df command directly calls the filesystem and queries the current disk space used. The super blocks of the file system provides this information.

The possible causes for the difference could be as follows:

Processes with open file which does not exist or has been deleted from the filesystem. A most common example would be a log file to which a process keeps writing continously. When the file is deleted, the process keeps running and the space is not freed (the process keeps the file open). The disk blocks associated with the file are actually deleted and made available for reuse when the last reference to the file is removed.

When a Unix process opens a file, the reference count to that file is incremented. Subsequently, if the file itself is removed from the filesystem, the data blocks remain in use until the process closes the file, either either when stops itself or is killed.

Under these conditions, du will be unable to see the file in the filesystem and therefore will not count its size, but df in getting the answer from the filesystem beleives the file still exists.When the process closes the file either when it quits or is killed, or when the machine is rebooted, the disk blocks will return to the free list and du and df will see the actual size. In most cases, an unmount and remount (or a reboot) will fix the problem.

Inconsistent file system requiring fsck

If the filesystem becomes corrupt/inconsistent for some reason then there can be a problem. What can be seen by a process looking at the filesystem (du), does not match up with the view the filesystem itself has (df). In this situation a fsck should help.

Boot from the CD and run a fsck or take down the system to Single User mode and run a fsck.

Mount point containing data
If a mount point already contains data, the du process will be unable to see this data (sees only the mounted filesystem), but the underlying filesystem will still keep track of this data, consequently df will report the extra disk space in use.Unmounting the filesystem will reveal the data. However, if the mounted filesystem is being used by running processes it will not be possible to unmount it. Either identify and kill the processes (fuser), or reboot in single user mode to check the mount point directory.

Error: ! cron cannot fork

2007-03-26T22:34:00.000+01:00

Error: ! cron cannot fork Wed Nov 2 11:10:00 2005! rescheduling a cron job Wed Nov 2 11:10:00 2005

This could happen, if the cron daemon has failed or crashed. The easist solution would be to restart the Cron daemon.

The following are the ways that you can restart Cron daemon. Perform them in the order and proceed to the next step only when the previous fails to fix the problem.

Use Cron stop / Start

# cron stop

will try to stop Cron cleanly. And, then run

#cron start

This should start the cron daemon cleanly. Sometime, this may not help in which case, try the next step suggested below:

#/etc/init.d/cron stop

This will try to stop the cron daemon cleanly.

Start the daemon back by using:
#/etc/init.d/start

The /etc/init.d/cron script will have the proper handling for the FIFO file during the startup.

This should help. Should this still fail, try the following:

#/etc/rc2.d/S75cron stop

and

#/etc/rc2.d/S75cron start

This should help.

Run levels in Sun Solaris

2007-03-26T22:31:00.000+01:00

Know the Run levels on Sun Solaris Operating System

The following lists the 6 Run Levels in Sun Solaris. These are often different mostly from the linux run levels

0 - Open Boot prompt (OBP) also called OK prompt. The system is taken down to PROM monitor or Security monitor state. It is safe to power off the hardware.

1,s,S - Single-User or Administrative mode where user logins are disabled. Also minimal Kernel functions and minimal required file systems are mounted (/ and /usr).

2 - Multi-User mode without Networking (No NFS)

3 - Multi-User mode with Networking (Default)

4 - Not in use

5 - Shut-down and power OFF. Power OFF supported on Sun 4m & Sun 4u architecture. More like a boot -a where the system is taken down to run level 0 and then an interactive boot.

6 - Reboot. takes the system down to run level 0 and then back to the default run level (3)

SSH without password in Solaris

2007-03-26T22:17:00.000+01:00

A Trust relationship can be established for users on multiple servers running OpenSSH to allow a password free ssh session. This is sometime important when you want to run scripts or commands remotely.

Most part of this article is common for most if not all of the UNIX/LINUX flavors except the following which steps to uncomment a few lines from the ssh_config file in Solaris:

Uncomment the following lines from the /usr/local/etc/ssh_config file:
RSAAuthentication yes
IdentityFile ~/.ssh/id_rsa

Now, let's assume ServerA and ServerB both run the ssh daemons.To allow ServerA to SSH to ServerB without password,

please try the following:

# ssh-keygen -t rsa

Note: User here is root.

This generates two files id_rsa.pub and id_rsa

Now, this needs to be copied to the authorized_keys file on ServerB

# scp id_rsa.pub ServerB:~/.ssh/authorized_keys

Enter password when prompted.

BEWARE: If the ServerB is already having a trust relationship with more that one hosts already then the above will wipe the contents and write this key alone. In which case, copy the file to the remote server as something like ServerA_rsa.pub and then append the contents to authorized_keys as follows. This will allow the existing authroized_keys from being wiped off.

# scp id_rsa.pub ServerB:~/.ssh/ServerA_rsa.pub
# cat ServerA_rsa.pub >> authorized_keys

Thats it. Test if you are able to do a ssh from ServerA without a password:

# ssh serverB uname -a

This will run the command "uname -a" on ServerB and returns the result on ServerA.

The same procedure has to be followed in the reverse to allow ServerB to talk back to ServerA

And, if there is anyone other server to be added to the existing list follow the same procedure ensuring the key is appended to the remote servers authorized_keys file and not by overwriting it. without any password.