Solaris 10: tips and tricks for system administration - SysAdmin

Sun Cluster 3.2 is available for download

nospam@example.com (Gleb Reys) — Wed, 20 Dec 2006 13:40:02 +0300

I've just read at Ilya Voronin's blog (blog.curthread.org), that for all the Sun Cluster fans, today is definitely a lucky day, because the new version of Sun Cluster - 3.2 - is already available for download.

By the way, looks like from this version on the clustering solution from Sun is called Solaris Cluster, and includes the following components:
- Sun Cluster
- Sun Cluster Geographic Edition
- Sun Cluster Agents

To my shame and regret, I haven't got a chance to acquire any useful experience with Sun Cluster. In my company I'm working with Veritas Cluster Server, because it's a historical thing. But I have some theoretical knowledge of Sun Cluster, and would really like to give it a try some day.

Anyway. Here are just some of the Sun Cluster 3.2 features:

New command line interface
In new Sun Cluster version, you have separate commands for each type of cluster object, and all of them try to have similar names for command line options where possible. Alsom, both short and long names for each option are supported.

SMF support
Sun Cluster is now tightly integrated with SMF, and easily supports controlling SMF-managed services as Sun Cluster resources.

Sun Cluster Quorum Server
For any scenarios requiring quorum, you no longer have to use shared storage. Sun Cluster supports using a separate Solaris server running quorum-server module. All the atomic reservations will be done over TCP/IP.

Such an approach allows you to reduce the time needed for failing over from one cluster node to another.

Extended Solaris Zones support
Sun Cluster supports now a plethora of services running within non-global zones.

ZFS support
Starting from this Sun Cluster version, ZFS is fully supported as a local highly available filesystem (HA FS).

Multi-Terabyte Disk and EFI (Extensible Firmware Interface) labels
A quite traditional update for many other packages - full support of disks with capacity over 1Tb and new EFI disk labels.

Better support of Veritas products
In particular, components of Veritas Veritas Storage Foundation 5.0 for SPARC are supported, as well as VxVM 4.1 for x86/x64 (in case you didn't know, Veritas products of these versions are already Solaris 10 aware).

Live Upgrade
Now you can use Live Upgrade to update your OS with Sun Cluster on top ot if. From documentation I can see that Live Upgrade supports SVM only at the moment, not VxVM.

Command Logging
You can have all the cluster management commands logged now - this is good for both diagnostics and for reproduction of your cluster environment in a new configuration.

You can download Sun Cluster 3.2 here: Sun Cluster 3.2 download.

Also, you might like reading a Sun Cluster blog - Sun Cluster Oasis. And, as always. for more information you can consult the Sun Cluster section of docs.sun.com: Sun Cluster 3.2 Documentation.

Free equivalents to Windows software

nospam@example.com (Gleb Reys) — Tue, 19 Dec 2006 03:01:03 +0300

I've just learned through Lifehacker about a great page: The Linux Equivalent Project. Actually, it's a pretty impressive index of free open-source software, which will work on Solaris, Linux, or any other UNIX OS.

What I really like about this list is that you're given names of software products Windows, and they a few of open-source equivalents. Very useful, especially for all the beginners who are just switching to Solaris.

If you miss some particular software in Solaris world, please have a look: LinuxEq.com

Using Solaris 10 as DHCP client

nospam@example.com (Gleb Reys) — Sat, 29 Apr 2006 16:14:07 +0400

Today I just wanted to document here the quickest and easiest way of setting your Solaris box as a DHCP client.

1) Create the /etc/hostname. file
For instance:

bash-3.00# touch /etc/hostname.skge0

2) Create the /etc/dhcp. file
It can be an empty file, but you can also use it to specify how long ifconfig should be waiting for a DHCP-server reply before giving up and continuing with Solaris boot.

bash-3.00# cat /etc/dhcp.skge0
wait 60

3) Your system name
It has to be specified in /etc/nodename. This name is going to be used as your hostname in case your DHCP-server does not return your hostname in reply to your DHCP-request.

THat's all. Reboot your system, and it all should work! Enjoy! :)

HOW-TO: X11 forwarding using ssh, putty and Xming

nospam@example.com (Gleb Reys) — Fri, 21 Apr 2006 12:37:44 +0400

I've been using this combination on a daily basis for more than a year now, and I can't stress enough how really useful it is. I use it both at home and at work, and everything is very easy and - most importantly - absolutely free.

X11 Forwarding with ssh is a wonderful feature which allows you get windows of a remotely started applications shown on your own desktop. For Windows, there are lots of pretty good albeit expensive products like Citrix, GoGlobal and XWin32, which allow you access your remote Unix desktop sessions. But in reality, if you don't need any sessions but only want to remotely start an application and get a window from it on your desktop, you won't need any of these expensive products - the easiest will be to use X11 forwarding and Xming.

Xming is an X Windows port for Microsoft Windows. Essentially it's an X-server which starts transparently on top of your MS Windows desktop. It allows you to redirect graphical output of applications you run on remote Unix servers and therefore see these applications windows on your MS Windows desktop.

Xming is very simple and easy to use. All you have to do is download its distribution archive from the project's page on SourceForge: XMing @ SourceForge.

After you install it, I recommend you have a quickstart panel shortcut created for it, and have your command line altered to something like this:

"C:\Program Files\Xming\Xming.exe" :0 -clipboard -multiwindow

Now that you have Xming installed, start it and it's time to take care of the ssh side of things. We have to alter the ssh daemon config file: /etc/ssh/sshd_config. Just ensure that it has the following:

# X11 tunneling options
X11Forwarding yes
X11DisplayOffset 10
X11UseLocalhost yes

If you already have similar parameters, don't do anything. But if you had to change the config file, it's time to restart ssh:

bash-3.00# svcadm restart svc:/network/ssh

All that's left now is to download the wonderful free Putty client, that is if you're not using it yet. You can get it here: Putty: a free telnet/ssh client.

I won't go into all the Putty configuration details, I'll only mention that for X11 forwarding, you have to do the following: in the main configuration window of Putty (you get it when you start Putty), select Connection section, then SSH, then X11, and make sure you tick the Enable X11 Forwarding option.

That's it, you're good to go! Good luck!

What would we do without crle?

nospam@example.com (Gleb Reys) — Thu, 13 Apr 2006 01:38:28 +0400

While compiling some of the stuff I use in 64bit lately, I've had another chance to use the wonderful crle tool and to acknowledge once again just how easier it made my life.

Here's what you can read from man crle:

The crle utility provides for the creation and display of a
runtime linking configuration file. The configuration file
is read and interpreted by the runtime linker, ld.so.1(1),
during process start-up.

The tool does exactly that - it speficies where ld should look first for the necessary libraries as you start any executable.

What would we do without this tool? The answer's simple:

1) we'd have to specify libraries directories implicitly every time we compile something - using -L and -R parameters.
2) we'd probably have to use the good old LD_LIBRARY_PATH environment variable, which again specifies where libraries should first be looked at.

Not only it's not cool because you have to waste time, but if you want someone else to use the program you'd get compiled using the tricks above, this someone would have to follow your steps and do the same - set LD_LIBRARY_PATH to the same values.

But with crle it's much easier:

to look at the current list of directories used by ld just do this:

-bash-3.00$ crle

Default configuration file (/var/ld/ld.config) not found
Default Library Path (ELF): /lib:/usr/lib (system default)
Trusted Directories (ELF): /lib/secure:/usr/lib/secure (system default)

and for 64-bit objects, do it like this:

-bash-3.00$ crle -64

Configuration file [version 4]: /var/ld/64/ld.config
Default Library Path (ELF): /lib/64:/usr/lib/64:/usr/sfw/lib/64
Trusted Directories (ELF): /lib/secure/64:/usr/lib/secure/64 (system default)

Command line:
crle -64 -c /var/ld/64/ld.config -l /lib/64:/usr/lib/64:/usr/sfw/lib/64

As you can see, adding more directories is quite easy. For instance, if I want to add another library directory of my own, /export/soft/lib/64, this command would help me:

-bash-3.00$ crle -64 -c /var/ld/64/ld.config -l /lib/64:/usr/lib/64:/usr/sfw/lib/64:/export/soft/lib/64

Compiling MySQL for amd64

nospam@example.com (Gleb Reys) — Mon, 10 Apr 2006 18:03:52 +0400

Yesterday I started compiling all the necessary software on my new server. Naturally, I'm trying to make it all 64bit. For lots of software packages I've used the same set of CFLAGS and LDFLAGS parameters, and everything worked just fine.

So, I've used the same configure script parameters to compile the latest 64-bit version of MySQL - 5.0.19:

mysql#MAKE="/usr/sfw/bin/gmake" CFLAGS="-m64" LDFLAGS="-m64 -L/lib/64" ./configure --prefix=/export/soft/mysql-5.0.19 --with-charset=utf8

but during the compilation I've got this error:

Making all in sql
gmake[2]: Entering directory `/export/dist/mysql-5.0.19/sql'
if g++ -DMYSQL_SERVER -DDEFAULT_MYSQL_HOME="\"/export/soft/mysql-5.0.19\"" -DDATADIR="\"/export/soft/mysql-5.0.19/var\"" -DSHAREDIR="\"/export/soft/mysql-5.0.19/share/mysql\"" -DHAVE_CONFIG_H -I. -I. -I.. -I../zlib -I../innobase/include -I../include -I../include -I../regex -I. -O3 -DDBUG_OFF -fno-implicit-templates -fno-exceptions -fno-rtti -DHAVE_RWLOCK_T -MT gen_lex_hash.o -MD -MP -MF ".deps/gen_lex_hash.Tpo" -c -o gen_lex_hash.o gen_lex_hash.cc; \
then mv -f ".deps/gen_lex_hash.Tpo" ".deps/gen_lex_hash.Po"; else rm -f ".deps/gen_lex_hash.Tpo"; exit 1; fi
/bin/bash ../libtool --preserve-dup-deps --mode=link g++ -O3 -DDBUG_OFF -fno-implicit-templates -fno-exceptions -fno-rtti -DHAVE_RWLOCK_T -m64 -L/lib/64 -o gen_lex_hash gen_lex_hash.o ../myisam/libmyisam.a ../myisammrg/libmyisammrg.a ../heap/libheap.a ../vio/libvio.a ../mysys/libmysys.a ../dbug/libdbug.a ../regex/libregex.a ../strings/libmystrings.a ../zlib/libz.la -lpthread -lthread -lposix4 -lcrypt -lgen -lsocket -lnsl -lm -lpthread -lthread
mkdir .libs
g++ -O3 -DDBUG_OFF -fno-implicit-templates -fno-exceptions -fno-rtti -DHAVE_RWLOCK_T -m64 -o .libs/gen_lex_hash gen_lex_hash.o -L/lib/64 ../myisam/libmyisam.a ../myisammrg/libmyisammrg.a ../heap/libheap.a ../vio/libvio.a ../mysys/libmysys.a ../dbug/libdbug.a ../regex/libregex.a ../strings/libmystrings.a ../zlib/.libs/libz.so -lpthread -lthread -lpthread -lthread -lpthread -lthread -lposix4 -lcrypt -lgen -lsocket -lnsl -lm -lpthread -lthread -Wl,-R -Wl,/export/soft/mysql-5.0.19/lib/mysql
ld: fatal: file gen_lex_hash.o: wrong ELF class: ELFCLASS32
ld: fatal: File processing errors. No output written to .libs/gen_lex_hash
collect2: ld returned 1 exit status
gmake[2]: *** [gen_lex_hash] Error 1
gmake[2]: Leaving directory `/export/dist/mysql-5.0.19/sql'
gmake[1]: *** [all-recursive] Error 1
gmake[1]: Leaving directory `/export/dist/mysql-5.0.19'
gmake: *** [all] Error 2

Attentive reader or a seasoned developer will immediately see what my problem was. Turned out, I've forgotten to specify the -m64 parameter for g++.

So the following parameters worked for me:

mysql#MAKE="/usr/sfw/bin/gmake" CFLAGS="-m64" CXXFLAGS="-m64" LDFLAGS="-m64 -L/lib/64" ./configure --prefix=/export/soft/mysql-5.0.19 --with-charset=utf8

Update: One more error was found at a later stage of MySQL compilation (I've abridged the command output to save space):

gmake[3]: Entering directory `/export/dist/mysql-5.0.19/server-tools/instance-manager'
...
g++ -DDBUG_OFF -m64 -fno-implicit-templates -fno-exceptions -fno-rtti -DHAVE_RWLOCK_T -m64 -o .libs/mysqlmanager command.o mysqlmanager.o manager.o log.o thread_registry.o listener.o protocol.o mysql_connection.o user_map.o messages.o commands.o instance.o instance_map.o instance_options.o buffer.o parse.o guardian.o parse_output.o -L/lib/64 -L/usr/sfw/lib/amd64 ./.libs/liboptions.a -lpthread -lthread -lpthread -lthread /usr/sfw/lib/amd64/libstdc++.so -lgcc_s libnet.a ../../vio/libvio.a ../../mysys/libmysys.a ../../strings/libmystrings.a ../../dbug/libdbug.a ../../zlib/.libs/libz.so -lpthread -lthread -lpthread -lthread -lpthread -lthread -lposix4 -lcrypt -lgen -lsocket -lnsl -lm -lpthread -lthread -Wl,-R -Wl,/usr/sfw/lib/amd64 -Wl,-R -Wl,/export/soft/mysql-5.0.19/lib/mysql
ld: fatal: library -lgcc_s: not found
ld: warning: file /usr/sfw/lib/amd64/libstdc++.so: attempted multiple inclusion of file
ld: fatal: File processing errors. No output written to .libs/mysqlmanager
collect2: ld returned 1 exit status
gmake[3]: *** [mysqlmanager] Error 1
gmake[3]: Leaving directory `/export/dist/mysql-5.0.19/server-tools/instance-manager'
gmake[2]: *** [all-recursive] Error 1
gmake[2]: Leaving directory `/export/dist/mysql-5.0.19/server-tools'
gmake[1]: *** [all-recursive] Error 1
gmake[1]: Leaving directory `/export/dist/mysql-5.0.19'
gmake: *** [all] Error 2

So I had to go into /usr/sfw/lib/amd64 and do this:

ln -s libgcc_s.so.1 libgcc_s.so

If I come across anything interesting while compiling 64bit things, I'll be sure to mention it here, but for now I just hope this entry will keep you from making the same mistake. Good luck!

Asus A8N-SLI networking in Solaris

nospam@example.com (Gleb Reys) — Fri, 07 Apr 2006 03:01:27 +0400

Finally, I've found some time to finish off the on-board network adapters configuration in Solaris 10 for my Asus A8N-SLI motherboard.

As I've already told before, this motherboard comes equipped with 2 network adapters:nForce4 and Yukon Gigabit Ethernet 10/100/1000Base-T Adapter. Both adapters are based on Marvell chips - they are Marvell 88E8001-LKJ1 (PCI Gigabit Ethernet) and Marvell 88E1111-RCJ (PHY) respectively.

So, the first one of these adapters is properly found and supported in Solaris 10 out of the box. In fact, Solaris 10 even finds and attempts to configure nge0 during the initial OS install.

To make the Yukon Gigabit (Marvell) adapter work, you should go to the SysKonnect website and navigate to the drivers downloads page (skge is there as well). What's REALLY cool is that there is a 64bit driver already! So I've got myself this file: skgesol_x64v8.12.1.3.tar.Z

After you unpack the archive, you should install the SKGEsolx, and that's it - the pre-install script even allows you to configure skge0 adapter - set its IP and network mask.

This is where my two network adapters are found (taken from /etc/path_to_inst):
"/pci@0,0/pci10de,5c@9/pci1043,811a@c" 0 "skge"
"/pci@0,0/pci1043,8141@a" 0 "nge"

Attention: due to unknown reasons, prtconf -vp clearly shows only one of the adapters - skge. That's the one represented by the "Ethernet controller" model name. Don't know why, but nge is called "Bridge device" there.

CISecurity: Solaris 10 security scoring tool

nospam@example.com (Gleb Reys) — Tue, 04 Apr 2006 16:42:22 +0400

At last! CISecurity has finally completed their Solaris 10 securiy scoring tool, and now you can download it here: Sun Solaris Seecurity Benchmarks.

pS: There's probably been a lot of work put into this project, as it certainly took its time - I've mentioned about this scoring tool for the first time back in July'05: Solaris mini-digest, #2

How to quickly install the Solaris 10 OS

nospam@example.com (Gleb Reys) — Mon, 13 Mar 2006 12:40:04 +0300

sun.com has a wonderful how-to on installing the Solaris 10 1/06 OS. The article contains plenty of screenshots and explanations, so it's very easy to follow. I was going to write a similar how-to myself, so this saves my time as I couldn't possibly do it better!

Here is the link: How to Quickly Install the Solaris 10 1/06 OS

HOW-TO: using truss, ldd and nm to observe binary files

nospam@example.com (Gleb Reys) — Fri, 03 Mar 2006 11:07:50 +0300

I've decided to write down a few useful tips on observing binary files in Solaris. The commands used are available since Solaris 8, so hopefully you'll find these tricks useful.

Quite often it happens that the programs we run don't work as they should, showing us errors like "referenced symbol not found" or this:
ld.so.1: ./thunderbird-bin: fatal: libmozjs.so: open failed: No such file or directory

So in order for us to find out what exact librabries are being used by a given binary, or where it expects these libraries to be found, and what libraries are not found at all, we have to take the following steps:

1) First we have to see what libraries the ld linker is going to use for our file. For instance, this is how you would find our what libraries are used by the main executable of the Thunderbird mail client:

bash-3.00$ ldd thunderbird-bin

        libmozjs.so =>   (file not found)

        libgtk-x11-2.0.so.0 =>   /usr/lib/libgtk-x11-2.0.so.0

        libgdk-x11-2.0.so.0 =>   /usr/lib/libgdk-x11-2.0.so.0

        libatk-1.0.so.0 =>       /usr/lib/libatk-1.0.so.0

        libgdk_pixbuf-2.0.so.0 =>        /usr/lib/libgdk_pixbuf-2.0.so.0

        libm.so.2 =>     /lib/libm.so.2

...

As always in my examples, the command output is abridged, but you get the idea. It's clearly seen that one of the files - the libmozjs.so library - isn't found. This means, there is no such file in either of the standard library paths or in the current working directory.

2) Now it's about time we find our where the linker expects this libmozjs.so to be found. To do this, we're going to use the truss command. All we're interested in at this stage are the stat and open system calls: stat is used to verify whether some file exists, and open is used to open this file for the later access.

bash-3.00$ truss -f -t stat,open ./thunderbird-bin

13070:  stat("/export/soft/thunderbird/thunderbird-bin", 0xFFBFFAD0) = 0

13070:  open("/var/ld/ld.config", O_RDONLY)             Err#2 ENOENT

13070:  stat("/export/soft/thunderbird/libc.so.1", 0xFFBFF600) Err#2 ENOENT

13070:  stat("/export/soft/thunderbird/../libc.so.1", 0xFFBFF600) Err#2 ENOENT

13070:  stat("/usr/sfw/lib/libc.so.1", 0xFFBFF600)      Err#2 ENOENT

13070:  stat("/opt/sfw/lib/libc.so.1", 0xFFBFF600)      Err#2 ENOENT

13070:  stat("/usr/local/lib/libc.so.1", 0xFFBFF600)    Err#2 ENOENT

13070:  stat("/usr/openwin/lib/libc.so.1", 0xFFBFF600)  Err#2 ENOENT

13070:  stat("/lib/libc.so.1", 0xFFBFF600)              = 0

13070:  open("/lib/libc.so.1", O_RDONLY)                = 3

13070:  stat("/export/soft/thunderbird/libxpcom.so", 0xFFBFF418) = 0

13070:  open("/export/soft/thunderbird/libxpcom.so", O_RDONLY) = 3

13070:  stat("/export/soft/thunderbird/libmozjs.so", 0xFFBFF398) Err#2 ENOENT

13070:  stat("/export/soft/thunderbird/../libmozjs.so", 0xFFBFF398) Err#2 ENOENT

13070:  stat("/usr/sfw/lib/libmozjs.so", 0xFFBFF398)    Err#2 ENOENT

13070:  stat("/opt/sfw/lib/libmozjs.so", 0xFFBFF398)    Err#2 ENOENT

13070:  stat("/usr/local/lib/libmozjs.so", 0xFFBFF398)  Err#2 ENOENT

13070:  stat("/usr/openwin/lib/libmozjs.so", 0xFFBFF398) Err#2 ENOENT

13070:  stat("/lib/libmozjs.so", 0xFFBFF398)            Err#2 ENOENT

13070:  stat("/usr/lib/libmozjs.so", 0xFFBFF398)        Err#2 ENOENT

ld.so.1: ./thunderbird-bin: fatal: libmozjs.so: open failed: No such file or directory

13070:  stat("/export/soft/thunderbird/libgtk-x11-2.0.so.0", 0xFFBFF398) Err#2 ENOENT

...

From this output you can see very well how linker goes through one standard libraries directory to another in a cyclic search for ever library our binary needs. And for our libmozjs.so we can see that logically this library would be found in the current directory where I'm running the thunderbird-bin from, /export/soft/thunderbird. There is no such file there, simply because I've renamed it for this demo.

And this is how the same ldd command would look when all the files are in place:

bash-3.00$ ldd thunderbird-bin

        libmozjs.so =>   /export/soft/thunderbird/libmozjs.so

        libgtk-x11-2.0.so.0 =>   /usr/lib/libgtk-x11-2.0.so.0

        libgdk-x11-2.0.so.0 =>   /usr/lib/libgdk-x11-2.0.so.0

        libatk-1.0.so.0 =>       /usr/lib/libatk-1.0.so.0

        libgdk_pixbuf-2.0.so.0 =>        /usr/lib/libgdk_pixbuf-2.0.so.0

        libm.so.2 =>     /lib/libm.so.2

...

So you can be assured now that our library is found and loaded from the /export/soft/thunderbird/libmozjs.so location.

Thus, by using this simple yet very powerful combination of ldd and truss commands, you can stop guessing and find out exactly which libraries are loaded in a particular order and where they are loaded from.

3) For the "referenced symbol not found" errors, we have to take one more step. These errors happen mostly because the linker finds the library our binary needs, but this library somehow doesn't have the functions we're interested in. So, once we know the exact library file after following the steps 1 and 2 of this manual, it's time for us to look at this particular file and see what functions are available. Many libraries have thousands of functions accessible, so it would make sense to use grep if you know the function you're looking for.

For instance, the following command shows all the functions of the libgdk-x11-2.0 library, which have gdk_threads in their names:

bash-3.00$ nm /usr/lib/libgdk-x11-2.0.so.0 | grep gdk_threads

[3632] | 86512| 68|FUNC |GLOB |0 |11 |gdk_threads_enter

[219] | 86648| 104|FUNC |LOCL |0 |11 |gdk_threads_impl_lock

[220] | 86752| 104|FUNC |LOCL |0 |11 |gdk_threads_impl_unlock

[4329] | 86856| 212|FUNC |GLOB |0 |11 |gdk_threads_init

[3507] | 86580| 68|FUNC |GLOB |0 |11 |gdk_threads_leave

[4095] | 532720| 4|OBJT |GLOB |0 |17 |gdk_threads_lock

[3579] | 532716| 4|OBJT |GLOB |0 |17 |gdk_threads_mutex

[4391] | 87068| 144|FUNC |GLOB |0 |11 |gdk_threads_set_lock_functions

[4443] | 532724| 4|OBJT |GLOB |0 |17 |gdk_threads_unlock

As you understand, if you get a "referenced symbol not found" error, this is quite often because an incorrect (incompatible) version of some library is used - it could be very old or just too new a version of the library, so the functions your binary is looking for are simply not there anymore or called differently.

Well, that's all for today. If you have any additional info to add to this short manual, you're very welcome to leave comments. I'll be happy to hear your opinion.

Solaris services: replacing SunSSH with OpenSSH

nospam@example.com (Gleb Reys) — Thu, 23 Jun 2005 02:36:34 +0400

Today I was going to spend some time writing up an example of adding a service to Solaris 10 and explain working with manifest by replacing the standard SSH-server which comes with Solaris 10 FCS with the latest build of OpenSSH - 4.1p1. But after I'd quickly looked into the /var/svc/manifest/network/ssh.xml file, I realised that I wouldn't write a really good example today. Things turned out to be much easier than I expected. To replace an SSH server, you simply have to change paths to ssh-keygen and sshd files in /lib/svc/method/sshd. That's the file used by the manifest ssh.xml.

pS: of course, you're going to have to download and compile OpenSSH before doing anything else.

NVidia OpenGL driver for Solaris x64/x86

nospam@example.com (Gleb Reys) — Sat, 04 Jun 2005 15:07:49 +0400

Wonderful news! Finally, we have an OpenGL driver for x86!
As Alan already said, NVidia has finally released an OpenGL driver. This driver will work only in Solaris 10 and newer Solaris releases. Everything works just fine, and just like Sean had mentioned, we've already tried it with Quake II! :)

Solaris x86 driver page on NVidia's website can be found here.

HOW-TO: Installing Solaris with PXE/DHCP

nospam@example.com (Gleb Reys) — Thu, 26 May 2005 14:44:09 +0400

Next Solaris Express build will probably contain one of the long awaited features: Solaris NewBoot architecture for x86 systems. Among other boot architecture improvements, NewBoot introduces GRUB with ufs support, and a pxegrub for PXE booting. Casper Dik describes it here:
The End Of Realmode Boot.
In this entry I'll try and explain how to set up an install server to allow you do a network install of Solaris with PXE.

The procedure described below describes steps necessary for PXE booting and installing
a client with Grub bootloader, which is the default one in Solaris since Nevada b14. Obviously, the very first requirement is that you absolutely must have a PXE-enabled network card installed in the client machine you're about to install.

Here is the full list of server-side components you need to have in order to be able
to PXE-boot and install a client machine over network:

1. DHCP server setup and configured
2. Network install OS image
3. Jumstart configuration files for a given client

I'll go through this list first, and then will give a list of actions you
have to take to configure and PXE-boot a client:

1. DHCP SERVER

1.1. Setup
Start off by doing this:

bash-3.00# dhcpconfig -D -r SUNWbinfiles -p /var/dhcp

This configures your DHCP server: it creates an empty resources table for
its use in /var/dhcp, and creates an /etc/inet/dhcpsvc.conf file with a
contents like this:

CODE:

DAEMON_ENABLED=TRUE
RESOURCE=SUNWbinfiles
RUN_MODE=server
PATH=/var/dhcp
CONVER=1

Your default DHCP daemon (/usr/lib/inet/in.dhcpd) is started at this stage, and
the above shown dhcpsvc.conf file ensures your dhcp server gets started up automatically upon reboot from /etc/init.d/dhcp script.

1.2. Networks

Here you have to look at your topology and decide, which networks are going to be served by your DHCP server. If you have a DHCP server configured by someone else, here's how you look at the list of networks served:

pntadm -L

It returns you a list, on our server it looks like this:

CODE:

192.1.3.0
192.1.4.0
192.1.1.0
192.1.2.0
192.1.9.0

To create a table for another network, do this:

pntadm -C 192.1.10.0

Make sure you have /etc/netmasks updated, if your network is subnetted.

Update your /etc/inet/dhcpsvc.conf file to include a line with all the
DHCP active interfaces. On our server it looks like this:
INTERFACES=hme1,qfe0,qfe1,qfe3

2. NETWORK INSTALL OS IMAGE

Follow the instructions from this page:
Creating an install server with CD media

In short: create a directory for your netinstall image, and use setup_install_server and add_instll_server
commands to copy necessary files from Solaris_10.1/Tools directory of your install media (CDs or DVD).
Share your netinstall image out.

Checklist:
- make sure the netinstall directory is shared out to your clients
- if you have more than one network, make sure all your networks have
access to this directory.

Share command should return you something like this for your install directory:

- /export/install2 ro=@192.1.1:@192.1.2:@192.1.3:@192.1.4:@192.1.9 ""

3. JumpStart CONFIGURATION

Follow the instructions from this page:
Creating a profile for networked system.

On our server, the JumpStart directory is /export/install/config. Then each machine has a directory there, like /export/install/config/perf-123.

Okay. Assuming you've done everything above, you've got a DHCP server configured, netinstall image ready and JumpStart files created for your client machine.

Now we're ready to reinstall your client. Here's what to do:

1) Generic add_install_client

Go to your netinstall image directory, and go to Solaris_10.1/Tools dir, where you have add_install_script. On our server, for Nevada b14 it's this directory:

/export/install2/nv/snv_14/Solaris_10.1/Tools

Now, do this:

./add_install_client -d SUNW.i86pc i86pc

It's a very important commant. It creates, if necessary, /tftpboot directory, or updates it with generic files for the netinstall OS image we're using. This command will also enable TFTP server on your machine. At last, it will create a /tftboot/pxegrub.I86PC.Solaris_10.1-1 file, which will be used to PXE boot each client you want to install with this build.

2) Client-specific add_install_client

Then, for our client, we have to do this:

CODE:

./add_install_client -d -e $ETHER \
-c $SERVER:/export/install/config/$SYSNAME \
-p $SERVER:/export/install/config/$SYSNAME \
-s $SERVER:$SOLARIS_PATH \
-b "console=$TTY" i86pc

It's a line from our script which automates the installation, and it shows clearly what is being passed to the add_install_client command: $ETHER is an MAC-address of the client box. Bear in mind, that for most MAC-addresses you have to add "01" to the beginning of the address. Also, you
have to get rid of the colons usually found in MAC address. So, for instance, if your client has this entry in /etc/ethers file on your server:
00:09:3d:00:68:f3 perf-123

The MAC-address you'll pass to the add_install_client will be this: 0100093D0068F3.

Other parameters are:

$SERVER - IP address of your server's interface on the subnet your client is on
$SYSNAME - a hostname of the client
$SOLARIS_PATH - is a full path to the netinstall image you're using.
$TTY - is an input-output device used during pxegrub boot. You can specify ttya/ttyb here to get an output on a console, connected to your client. You may also not specify this parameter at all for the default device to be used (screen/keyboard).

So, for our client perf-123 the final line looks like this:

CODE:

./add_install_client -d -e 0100093D0068F3 \
-c 192.1.1.254:/export/install/config/perf-123 \
-p 192.1.1.254:/export/install/config/perf-123 \
-s 192.1.1.254:/export/install2/nv/snv_14 \
-b "console=ttya" i86pc

This line creates three files in your /tftpboot directory:

menu.lst.0100093D0068F3 - grub menu for your client's PXE boot
rm.0100093D0068F3 - a standard shell script for cleaning up /tftpboot files for your client
0100093D0068F3 - a symbolic link to a build's pxegrub file in /tftpboot

For perf-123, this link looks like this:

CODE:

bash# ls -al /tftpboot/0100093D0068F3
lrwxrwxrwx 1 root other 28 May 23 17:27 \
/tftpboot/0100093D0068F3 -> pxegrub.I86PC.Solaris_10.1-1

If you check now, the grub menu file will look like this:

CODE:

bash# more /tftpboot/menu.lst.0100093D0068F3
default=0
timeout=30
title Solaris_10.1 Jumpstart
kernel /I86PC.Solaris_10.1-1/multiboot kernel/unix -B \
console=ttya,install_config=192.1.1.254:/export/install/config/perf-123, \
sysid_config=192.1.1.254:/export/install/config/perf-123, \
install_media=192.1.1.254:/export/install2/nv/snv_14
module /I86PC.Solaris_10.1-1/x86.miniroot

Unfortunatelly, at this early stage, we have to manually update this file so that the machine not only boots off the network, but starts JumpStart install automatically. To do this, you have to pass a "- install dhcp" as a boot argument to Solaris kernel. So we update one of the menu file lines to look like this:

CODE:

kernel /I86PC.Solaris_10.1-1/multiboot kernel/unix - install dhcp -B \
console=ttya,install_config=192.1.1.254:/export/install/config/perf-123, \
sysid_config=192.1.1.254:/export/install/config/perf-123, \
install_media=192.1.1.254:/export/install2/nv/snv_14

I'm sure later there will be a command line option, so I'll update this how-to accordingly.

As you can see, most of the add_install_clients parameters end up in this file. A menu file is also a good way to do basic sanity checks, cause it shows you where your client will try to mount a netinstall image from, and where JumpStart will expect to have your client's configuration files to be found.

3) DHCP - IP address and a macro for our client

First, add a fixed IP for your client, which will be issued to it by your server upon DHCP request:

pntadm -f 01 -e 12/31/2010 -A $IPADDR -i $ETHER -m $SYSNAME -s $SERVER $NETWORK.0

Most of the parameters are self-explanatory, and you know most of them from add_install_client command. So I'll mention only the new ones:

$IPADDR is the IP address you want DHCP server to assign to your client
$NETWORK is the network your client belongs to. Obviously, $IPADDR should be in this network as well.
-f 01 means that this IP lease is permanent
-e 12/31/2010 sets the expiration date for this lease (we sure have got some time before 2010)

Now, what this line does apart from assigning an IP address to our client, is that it also assigns the client a DHCP macro. By following the macro specified, your DHCP server will give the client all the additional info the client needs to PXE boot.

Here's the line:

dhtadm -A -m $SYSNAME -d ":BootSrvA=$SERVER:BootFile=$ETHER:"

This line tells the client to try and PXE boot off a specified server with IP $SERVER, and suggests that the bootfile will be $ETHER. This file, as you remember, essentially points to a pxegrub file for the netinstall build.

That's it! Make sure your BIOS points to a PXE-enabled network card on your client, reboot the box, and you should see it downloading the Solaris miniroot onto the client and then booting Solaris. As always, snoop is your friend with network troubleshooting.

Solaris 10 presentation from the authors of Solaris Internals

nospam@example.com (Gleb Reys) — Wed, 06 Apr 2005 18:38:41 +0400

Solaris Internals, a website of a famous book with the same name, has posted a presentation on Solaris 10 inner workings, and I strongly suggest you read it right away.

The authors of this presentation are the highest level engineers of Sun Microsystems, and judging by their book popularity, the presentation is going to be one one of the best information sources on Solaris 10 so far.

The presentation is available as PDF file download here.

vim colour schemes

nospam@example.com (Gleb Reys) — Thu, 06 Jan 2005 15:22:30 +0300

Simply put, I adore vim!

Today I finally found some time to compile its latest version on my laptop, and it was decided to find some new color sheme for it. I know, I know: there are tons of scripts for vim which demonstrate different schemes, but to do that it is necessary to download all them schemes files first, and I just didn't want that. I had spent just few seconds looking for it, and found a wonderful page with tons of vim color sheme tables. What I liked most is that tabled views are so convenient - it really is much easier to compare and pick schemes this way.

So, if you feel like choosing a new scheme for your vim and got a minute to spare, visit this page: Vim Color Scheme Test.