Monday Jan 09, 2012

New Documentation for Common Sysadmin Tasks

The Oracle Solaris 10 documentation covers the most common sysadmin tasks in two main guides:

Oracle Solaris 11 uses one:

Oracle Solaris 11 Common Tasks Guide

The Common Tasks Guide covers:

  • Using Oracle Solaris man pages
  • Setting up and managing user accounts
  • Booting and shutting down Oracle Solaris 11
  • Working with the Oracle Configuration Manager
  • Managing services through SMF
  • Using the Fault Manager
  • Managing software packages
  • Managing disk use
  • Displaying and managing system processes
  • Managing system resources and configuration
  • Managing printing
  • Troubleshooting system and software problems, and managing core files

The rest of the Oracle Solaris 11 documentation is here:

Oracle Solaris 11 Documentation Library

By the way, there is no printing guide in the Oracle Solaris 11 library. That information appears in this part of the Common Tasks Book:

Setting Up and Administering Printers Using CUPS

- Juanita Heieck

Tuesday Jan 03, 2012

Next OTN Sysadmin Day is on January 18

Our next OTN Sysadmin Day will be held on January 18 in Salt Lake City, Utah. As usual, we will have two tracks of hands-on-labs:

Time Session
8:00 am System Shakedown
9:00 am Oracle's Dual OS Strategy / Overview of OTN

Oracle Solaris Track

Oracle Linux Track

10:00 am HOL: ZFS HOL: managing packages, configuring services
11:30 am HOL: Exploring OS, network, and storage virtualization HOL on Storage Part I: managing storage and file systems
1:00 pm Lunch Break
2:00 pm HOL: Managing software with IPS HOL on Storage Part II: Device Mapper, BTRFS
3:00 pm Presentation: Oracle Enterprise Manager Ops Center 11g
4:00 pm Discussion: What are the most pressing issues for sysadmins today?
5:00 pm We all go home

Participants of previous OTN Sysadmin Days found the hands-on labs particularly valuable. You get to learn by doing. And what you get to do is install, configure, and manage the technologies of Oracle Solaris 11 and Oracle Linux in the same way as you would in the real world.

OTN Sysadmin Day in Salt Lake City is free, but you must register. Please stay for the feedback session at the end. They tend to be pretty spirited, and you might win a neat prize. Address:

Salt Lake City Marriott City Center
220 South State Street
Salt Lake City, UT 84111

If you'd like to see some pictures from the Sacramento event, go to the "OTN Sysadmin Day Sacramento" photo folder on the OTN Garage on Facebook.

To find out what there is to do is Salt Lake City and Utah, click on the ski page above. It will take you to National Geographic's Guide to Utah.

- Rick

Friday Dec 16, 2011

Two Sysadmin Articles Make OTN's Top 20

In the OTN blog, Justin reports that two sysadmin-related articles made OTN's top 20 list for 2011:

Number 2
Taking Your First Steps with Oracle Solaris 11
- by Brian Leonard and Glenn Brunette

Number 11
How I Simplified the Installation of Oracle Database on Oracle Linux
- by Ginny Henningsen


The good work of Brian, Glenn, and Ginny makes those of us in the Systems Community of OTN particularly proud because the number of OTN readers who are system admins and developers is dwarfed by the number who are Java developers. Even making the top 20 is notable. To Brian, Glenn, and Ginny, a heartfelt:

- Rick


Thursday Dec 15, 2011

How to Survive the End of the World - Part I

If you've been paying attention you'll probably agree that Earth will be destroyed any day, now.

That used to concern me.

But the more I understand clustering, the more I realize we can simply reconstitute civilization from individual slices of other planets in the Universe. The first thing we need to do is identify the building blocks of an advanced civilization. That should be relatively simple:


Next, find planets that had excellent examples of each building block:

Building BlockBackup Planet
civilizationbackup civilization

Those four planets plus Earth would be easy enough to arrange into a high-availability cluster if we downloaded and installed Oracle Solaris 11 and Oracle Solaris Cluster 4.0 on each planet, including Earth.

With Solaris Cluster 4.0, we could create a nice five-node cluster. Not only would the cluster provide the disaster recovery we're looking for, but it would actually help us create an elastic cloud of sorts, in which we could, for instance, tap into the beer of planet Bud during the Super Bowl or other times of dire need. See What's New to read about other cool things you can do with Solaris Cluster 4.0.

Creating a five-node cluster can get a bit tricky, but you can build up your skills by creating a smaller one, using the instructions in this OTN article:

How to Install and Configure a Two Node Cluster

Once you have the two-node setup figured out, you can move to the five-node setup. But the resulting five-node cluster is actually more than what we want, isn't it? It's a cluster of five entire planets, when what we're looking for is a slice of each planet. In an upcoming blog I'll summarize how to create a cluster from the slices of those individual planets. That's called a virtual cluster or a zone cluster, and it's very cool.

- Rick


Tuesday Nov 22, 2011

Screwed Up Again, Did Ya?

Your turn to wear the Cantaloupe Cap of Shame? Here's how to keep it from happening again:

  1. Figure out what data you need to archive
  2. Create a solid archive someplace safer than your iphone
  3. Get wicked fast at recovering your system.

Jesse Butler explains how to do all three for a system running Oracle Solaris 11:

How to Recover an Oracle Solaris 11 System

- Rick


Monday Nov 21, 2011

Is Oracle Solaris 11 Really Better Than Oracle Solaris 10?

If you want to be well armed for that debate, study this comparison of the commands and capabilities of each OS before the spittle starts flying:

How Solaris 11 Compares to Solaris 10

For instance, did you know that the command to configure your wireless network in Solaris 11 is not wificonfig, but dladm and ipadm for manual configuration, and netcfg for automatic configuration? Personally, I think the change was made to correct the grievous offense of spelling out "config" in the wificonfig command, instead of sticking to the widely accepted "cfg" convention, but loathe as I am to admit it, there may have been additional reasons for the change.

This doc was written by the Solaris Documentation Team, and it not only compares the major features and command sequences in Solaris 11 to those in Solaris 10, but it links you to the sections of the documentation that explain them in detail.

- Rick


Friday Nov 18, 2011

New Shell In Oracle Solaris 11

In Oracle Solaris 11, Korn Shell 93 (/usr/bin/ksh/ or usr/bin/ksh93) replaces both the Bourne Shell (/usr/bin/sh or /sbin/sh) and Korn Shell 88 (/usr/bin/ksh).

There are some incompatibilities between the shells. They are described in:


If a script has compatibility problems you can use the legacy shell by changing the she-bang line:

If this doesn't work

Use This

#!/bin/ksh #!/usr/sunos/bin/ksh
#!/usr/bin/ksh #!/usr/sunos/bin/ksh
#!/bin/sh #!/usr/sunos/bin/sh
#!/usr/bin/sh #!/usr/sunos/bin/sh
#!/sbin/sh #!/usr/sunos/bin/sh

- Mike Gerdts


Friday Nov 11, 2011

How to Find Out Which Devices Are Supported By Solaris 11

Image of monks gathering on the steps of the main hall in the Tashilhunpo Monastery is courtesy of Alison Whitear Travel Photography.

In his update of Brian Leonard's original Taking Your First Steps With Oracle Solaris, Glynn Foster walks you through the most basic steps required to get a version of Oracle Solaris 11 operational:

  • Installing Solaris (VirtualBox, bare metal, or multi-boot)
  • Managing users (root role, sudo command)
  • Managing services with SMF (svcs and svcadm)
  • Connecting to the network (with SMF or manually via dladm and ipadm)
  • Figuring out the directory structure
  • Updating software (with the IPS GUI or the pkg command)
  • Managing package repositories
  • Creating and managing additional boot environments

One of the things you'll have to consider as you install Solaris 11 on an x86 system is whether Solaris has the proper drivers for the devices on your system. In the section titled "Installing On Bare Metal as a Standalone System," Glynn shows you how to use the Device Driver utility that's included with the Graphical Installer.

However, if you want to get that information before you start installing Solaris 11 on your x86 system, you can consult the x86 Device List that's part of the Oracle Solaris Hardware Compatibility List (HCL). Here's how:

  1. Open the Device List.
  2. Scroll down to the table.
  3. Open the "Select Release" pull-down menu and pick "Solaris 11 11/11."
  4. Move over to the "Select Device Type" pull-down menu, and pick the device type. Or "All."

The table will list all the devices of that type that are supported by Solaris 11, including PCI ID and vendor.

In the coming days the Solaris Hardware Compatibility List will be updated with more Solaris 11 content. Stay tuned.

- Rick Ramsey

Monday Nov 07, 2011

5 Commands That Make Solaris Administration Easier

Face it. Not all of us have got it figured out.

If the Service Management Facility in Oracle Solaris 10 happens to be one of those areas that you didn't quite understand as well as you had intended, you may be interested to know that it's not as complicated as, say, the interplay between geopolitics and energy policy.

In fact, SMF only has five commands:

svcs Get detailed views of the service state of all service instances in the service configuration repository
svcadm Perform common service management tasks, such as enabling, disabling, or restarting service instances
svcfg Display and manipulate the contents of the service configuration repository
svcprop Retrieves property values from the service configuration repository with an output format appropriate for use in shell scripts
inetadm Observe or control services controlled by inetd

The svcfg and svcprop commands deal with the service repository (maintains configuration info and run-time data for services). The inetadm command focuses on inetd-controlled services. You can get more information about these three commands in the Oracle Solaris 10 Basic System Administration Guide, available from The Solaris 10 System Administration Documentation Collection. Or, if it hasn't moved, use this link.

If you want to learn about the many things that you can do with the first two commands, svcs and svcadm, read this technical white paper:

Easier System Administration with the Solaris Service Management Facility

It describes how you can use the svcs and svcadm commands to:

  • Display all the services currently running
  • Display information about individual services
  • Display dependencies between services
  • List the individual processes that constitute a service
  • Perform common administrative tasks such as starting a service and then monitoring it
  • Investigate system faults
  • It's a good way to become familiar with real-world uses of the SMF. And, perhaps, put you in the practice of fully developing your perspective before you are moved to unleash it upon the world.

    - Rick Ramsey and Cindy Swearingen

    Thursday Nov 03, 2011

    Solaris Security Resources on OTN

    image courtesy of Faisal's photo stream on Flikr

    An Overview of Oracle Solaris 10 Security Controls

    Glenn Brunette describes how to more easily secure ZFS file systems compared to UFS file systems in this white paper, along the following lines:

    UFS file systems have the following characteristics:

    • UFS file systems are directly tied to disk slices
    • Disk slice space is not easily expanded to increase capacity for UFS file systems because the disk generally contains other disk slices for active file systems
    • In some cases, you have to reinstall the OS to increase the size of the UFS root file system
    • UFS file system space is controlled by using UFS quotas

    ZFS file systems have the following advantages:

    • ZFS uses a pooled storage model where all the file systems in pool use available pool space.
    • No relationship exists between ZFS file systems and disk slices except for the ZFS root file system.
    • A long-standing boot limitation is that a ZFS root file system must be created on a disk slice.
    • During installation, you define the size of the root pool disk slice or mirrored slices that contain the root file system.
    • The root file system contains separate directories of system-related components, such as etc, usr, and var, unless you specify that var is separate file system.
    • You can put a reservation and a quota on the /var file system to determine how much disk space is reserved for /var and how disk space it can consume.

    For example, you might consider configuring a separate /var file system when installing a system that will be used as a mail server. This way, you can control the size of var with a quota so that root pool's space capacity is not exceeded.

    In addition, if the ZFS root file system and the /var file system begin to exceed the pool's capacity, you can easily replace the root pool disk with a larger disk without having to unmount, restore a backup, or reinstall the root file system.

    How should you configure your ZFS data sets for optimum security? Read Glenn's paper to find out. He not only provides security-based recommendations for ZFS, but also for:

    • Software installation clusters
    • Minimization
    • Non-executable stacks
    • Filesystems
    • USB Support
    • Plugable Authentication Modules
    • Service Management Facility
    • Cryptographic services management
    • Zones
    • And lots more

    If you're inclined to read more about security, try these other two papers we published recently, plus OTN's security collection.

    Oracle Solaris 11 Security: What's New for Developers

    Recommendations for Creating Reduced or Minimal Solaris Configurations

    OTN's Security Collection

    - Rick Ramsey and Cindy Swearingen

    Tuesday Oct 18, 2011

    Solving the Tape Storage Space Problem

    Because I came of age professionally in the shorts-and-tshirt smarter-than-you culture of Silicon Valley, I always assumed tape storage was used only by retired British spies with names such as Baratheon and Brewster and Cameron who lived in dank mansions on the rocky coast of Scotland and still dressed in tweed jackets for dinner. They spent their days engrossed in the struggle to keep their Dunhill pipes lit and their hair piece in place against fury of the North Wind. Every few months a carrier pigeon would arrive from MI5, and Baratheon or Brewster or Cameron would slowly descend the stairs to a back room. A week later he would return with a name written in code, and hand it to a man in a dripping wetsuit and spear gun who would jump off the cliff without looking and swim it back to a submarine waiting off the coast.

    Turns out I was behind the times. In fact, tape for archiving has several advantages that make it economically feasible in today's digital pack-rat economy. Such as durability. And much, much lower power consumption. You can read about them in this paper by Horison information strategies:

    Tape: The Digital Curator of the Information Age (registration required)

    If you're a storage admin or IT manager considering tape, there's another paper that may interest you more. Published on OTN in July, it describes very clearly the limitations of data that is written in a stream to tape, and how Oracle technologies overcome them. For instance, once you write a block of data to a stream of tape, that particular bit of data not only becomes inefficient to target for access, but updates to the data become clumsy and cumbersome. And as tape cartridges grow to store a terabyte of data, the problem becomes even more pronounced.

    Oracle's StorageTek In-Drive Reclaim Accelerator avoids this problem by simply breaking up the serial data on a tape into smaller, more manageable chunks that are grouped together and managed as logical volumes. Find out how in this well-written white paper:

    How it Works: StorageTek Reclaim Accelerator

    For more information about Oracle tape drive products, visit OTN's Tape Storage product page.

    - Rick

    Monday Oct 17, 2011

    Networking Services You Can Run Inside a Oracle Solaris 11 Express Zone

    Oracle Solaris 11 Express introduced a new network stack architecture previously known as “Crossbow”. It lets you combine virtual NICs into flexible virtual networks that are tightly integrated with zones. In addition, the new architecture introduces the ability to manage your network resources by controlling bandwidth and flow.

    As a result, you can now run these services inside a Solaris 11 Express zone:

    • DHCP client
    • DHCP server
    • Routing daemon
    • IPsec
    • IPfilter
    • IP Multipathing (IPMP)
    • ndd commands
    • ifconfig with set or modify capabilities (usage of dladm and ipadm is recommended

    This is just one of the changes between Oracle Solaris 11 Express and previous versions. For more info, see the Oracle Solaris 11 ISV Adoption Guide.

    - Rick

    Tuesday Sep 27, 2011

    Linux-Related Content and Roadmap at Oracle OpenWorld

    Interested in the Oracle Linux strategy and roadmap direct from Wim Coekaerts, VP of Linux Engineering at Oracle? Find out where and when, plus how other companies like Cisco and Intel are using Oracle Linux. Here's the summary of Linux-related content at Oracle OpenWorld:

    Focus on Oracle Linux

    The summary covers:

    • Keynotes
    • General Session
    • Oracle Linux and Oracle VM Customer Forum

    - Rick

    Solaris-Related Content and Labs at Oracle OpenWorld

    Did you know that you can virtualize the big honkin' Oracle Exalogic Machine with Oracle Solaris Zones? I didn't.

    You can learn about other things you can and can't do with Oracle Solaris at this year's Oracle OpenWorld, Oct 2-5, by perusing this handy summary the Solaris folks put together:

    Focus on Oracle Solaris

    The summary covers:

    • Keynotes
    • General Sessions
    • Hands-On Labs
    • Demos
    • Meet the Experts

    - Rick

    Friday Sep 16, 2011

    The Confederate Hellcat and Other Minimal Configurations

    I've been looking for a reason to use this picture of the Confederate HellCat for a while, now. A souped-up Harley engine in a radical sportbike chassis. Makes you want to run into the garage and roll around in dirty oil rags, doesn't it?

    Here's another minimal configuration:

    Recommendations for Creating Reduced or Minimal Oracle Solaris Configurations

    Some sites use OS minimization to reduce the security footprint of their Oracle Solaris installations. Others do it to reduce the administrative burden of patching and updating software. But minimization has both risks and benefits. Glenn Brunette provides his recommendations for mitigating the risks and reaping the benefits. Covers initial installation, package removal, patching, and what to watch out for. Applies to Oracle Solaris 10 and prior releases.

    And since we're talking about simplification, this article might also be apropos (that's French for "I like American beer"):

    How I Simplified the Installation of Oracle Database on Oracle Linux

    Ginny Henningsen describes how she simplified the installation of Oracle Database 11g by automatically pre-configuring Oracle Linux with the required software packages and correct kernel parameters. Hint: using the "oracle-validated " RPM package.

    - Rick

    Thursday Sep 15, 2011

    How Content Management Makes Tape Drives More Efficient

    Guest blog by Steve Aaker and Jamie Giovanetto

    Oracle StorageTek's Enterprise Library Software (ELS) is the storage management software available for these Oracle StorageTek Tape Storage mainframe products:

    In addition to its hardware enablement functions, ELS provides a rules-based content manager that can significantly increase the efficiency of your tape drives and systems. It's called the Library Content Manager or LCM for short, and it is described in the documentation for the ELS software.

    For physical tapes, which can be native data or scratch cartridges, Multiple Volume Cartridges (MVCs), or cleaning cartridges and empty (free) storage cells, LCM places the tapes in the best location to:

    • Maximize the efficiency of the enabling software selections
    • Minimize the robotics activity at job mount time.

    In the virtual tape environment, LCM controls whether a volume is in Oracle's StorageTek Virtual Storage Manager System buffer or only on an MVC at the appropriate time.

    This content management capability can dramatically increase the efficiency of your tape storage.

    An Example

    Let's take the case where a StorageTek SL8500 modular library system from Oracle is totally full of cartridges and has no content management. All mounts and dismounts for the tapes in the library occur when required, but the question is whether that is sufficient. It is the minimum expected action, to be sure, but is it being accomplished in an efficient manner? If the cartridge to be mounted is in Library Storage Module (LSM) 03 and the drive where it is to be mounted is in LSM 00 (because that is the only location for that type of drive), the mount will take five major robotics actions. At dismount time, because there are no free cells, it will take another five major robotics actions to return the cartridge back to its source cell. In both directions, at least one of these robotic actions will be a move through the entire length of the library. When complete, the cartridge is back in LSM 03 and the drives it can be mounted on are still in LSM 00. Each time the volume is mounted, this scenario is repeated.

    Now, let's look at what the same mount would look like in a content-managed StorageTek SL8500 modular library system. First, the cartridge to be mounted would have been placed in LSM 00 by a prior management run, which would have placed it in the LSM where the drives on which it could be mounted are located. Because of this management activity, the mount would require only two robotics actions, and the dismount would require the same. The reduction of three robotics actions, including that move through the entire length of the library, significantly reduces overall robotics time and results in much faster mount time. The cumulative mount times directly affect overall production time, so proper placement of volumes has a significant impact on production performance.

    However, you might ask, "Didn't the cartridge have to be moved to the right place at some time? Didn't that take robotics activity?" Of course, it did. The difference is that content management activity would have been done outside the production job's execution time. In addition, the cartridge would have been placed where it can be mounted many times without requiring the many robotics actions required in the unmanaged example.

    Note that this is just one of many examples where significant performance improvements can be obtained by active, rules-based content management. At its fullest implementation level, StorageTek LCM can bring scheduled production mounts down to sub-second levels in the virtual environment and, in some cases, to an average mount time that is approximately half the average mount time for an Oracle StorageTek SL8500 modular library system in the physical environment. Combined with its other capabilities and the underlying hardware and enabling software environment, Oracle's StorageTek Library Content Manager offers great enhancements to the automated tape environment.

    - Steve Aaker and Jamie Giovanetto

    Wednesday Sep 07, 2011

    Oracle OpenWorld for 50 Bucks - Really?

    If you're an OTN Member, you can get a Discover Pass to Oracle OpenWorld for $50.00. The Discover Pass won't let you into the technical sessions and hands-on labs, but it will let you into the keynote and executive sessions, the user group meetings, the exhibition halls and Demo Grounds and, of course, the OTN Lounge. And that's not all. Find out how it compares to full registration here:

    Discover Pass vs Full Registration

    Be sure to use this priority code: DOTN11.

    Here's how:

    • First, become an OTN member:
      1. Go to the OTN Membership Page
      2. Select the "join today!" link (underlined, in the middle of a paragraph)
      3. Go to "My Community Memberships"
      4. Select "Oracle Technology Network"
      5. Identify yourself as a "sysadmin."
    • Register for Oracle Open World and use priority code DOTN11

    By the way, that's my nephew Anthony on the right. Coolest kid ever.

    - Rick

    Thursday Aug 25, 2011

    Next OTN Sysadmin Day is in Sacramento

    Bottom Line: next OTN Sysadmin Day is September 8 in Sacramento.
    Free, but registration is required.

    One of the most subtle yet powerful skills I learned from The Missus is how to tell the difference between what was better and what I preferred.

    Motorcycle magazines, for instance, spend an awful lot of time telling us why one motorcycle is better than another. They do this by employing their best interpretation of objective criteria: acceleration, braking, cornering, carrying capacity, conveniences, ergonomics, and sometimes even the personal riding experience of the reviewer.

    There's nothing wrong with that. Very useful stuff. And fascinating to read. Except that it gets me all confused. Or confuzzled, as my daughter calls it. The objective criteria makes me think one bike is better than the other. But when I buy it, why am I less than fulfilled?

    Because features and capabilities don't often match personal preference.

    Although we don't usually put personal preference at the top of the list when choosing an operating system, we shouldn't ignore it. Personal preference is not just whim. It takes into account the job we're trying to accomplish, the way we have to handle it, and the tools we prefer to use. In fact, ignoring our personal preference has an impact on productivity: if we hate "our ride," we're not going to get very good at using it, are we?

    OTN's second Sysadmin Day (read about the first) will begin with an overview of Oracle's dual OS strategy. Then we'll split off into two tracks, one with hands-on labs for Oracle Linux, one for Oracle Solaris.

    Time Session
    8:00 am System Shakedown
    9:00 am Oracle's Dual OS Strategy

    Oracle Solaris Track

    Oracle Linux Track

    9:30 am Overview of Oracle Solaris Overview of Oracle Linux
    10:00 am HOL: ZFS HOL: managing packages, configuring services
    11:30 am HOL: Exploring OS, network, and storage virtualization HOL on Storage Part I: managing storage and file systems
    1:00 pm Lunch Break
    2:00 pm HOL: Managing software with IPS HOL on Storage Part II: Device Mapper, BTRFS
    3:00 pm Presentation: Oracle Enterprise Manager Ops Center 11g
    4:00 pm Discussion: What are the most pressing issues for sysadmins today?
    5:00 pm We all go home

    Consider it an Oracle Demo Day, if you will. It's free, but you must register to attend.

    As for me, after spending a couple of years riding the canyons on a 2005 Ducati 800SS, a 2003 Ducati ST4 with a 4-valve 996 engine, Ohlins shocks, and Marchesini wheels, a 2005 BMW K1200S, and the bikes of a few friends, I went back to a 2006 Harley Davidson Fat Boy. Even though the other bikes performed much better, I prefer the Fat Boy. And my wicked smaht Missus.

    - Rick

    Friday Aug 19, 2011

    Friday Relapse - Maradona and The Hand of God

    Three minutes after Diego Maradona scored a goal against England in the 1986 World Cup in Mexico City with a move so blatantly illegal that even an NBA official would have called him for palming the ball, he scores a solo goal so magnificent it became the Goal of the Century (video).

    See you next week.


    Thursday Aug 18, 2011

    The Impact of Oracle Optimized Solutions on a Sysadmin's Job

    When Oracle acquired Sun we kept it simple. In doing so we came up with Oracle Optimized Solutions. Of course they're great for business because they are assembled from a pre-configured stack of Oracle products that we test and use ourselves, but are they be great for Sysadmins? I decided to ask.

    The best person to ask anything about Oracle Optimized Solutions is Marshall Choy. Not only is Marshal the director of the engineering group that selects and assembles these systems, but he began his high tech career as a sysadmin working on both Solaris and Linux systems. Marshall agreed to let us put him on the spot, so Justin interviewed him on OTN TechCast. In addition to talking about a sysadmin's job, Marshall explains the patching strategy for these types of stacks. Here are some of the questions we asked him.

    1. What's the difference between Oracle Optimized Solutions and Oracle Engineered Systems?
    2. What will this mean my job as a sysadmin - will my skills become obsolete? Will I be replaced by someone less skilled?
    3. I'm not sure I want to outsource my sysadmin skills to Oracle - how will these optimized solutions change what I spend my time doing?
    4. Aren't we just turning back the clock 20 years - why did you decide to build a proprietary vertical stack?
    5. What if I want to change something in the stack, how will it affect my support contract?
    6. How often do you update the components in the stack, and do I get those updates for free?
    7. How do I install updates and patches?

    It's a good show. It lasts 14 minutes. Don't miss it.

    - Rick

    Tuesday Aug 16, 2011

    Ginny Had A Bright Idea

    image courtesy of Twenty Words

    In the Olden Days before most of us were born, if a woman got a bright idea she got an immediate spanking from John Wayne. Thank goodness John Wayne has stopped doing that, or we wouldn't get to reap the benefits of the research Ginny Henningsen did with Oracle Solaris 11.

    When Ginny read about all the different ways to download, install, patch, and manage updates in Solaris 11, she wasn't sure where to start. So she drew on her personal experience, the experience of other sysadmins and systems engineers, the documentation, and the related technical articles posted on OTN.

    The result? These three very practical articles.

    Article 1
    Best Way To Update Software Using IPS in Oracle Solaris 11

    The SVR4 packaging and patching systems in earlier versions of Solaris were designed by the Chosen for the Faithful. If you loved SunOS you could recite package nomenclature in your sleep and you always, always used the command line. Alas, nobody loves software for its own sake any more. At least, not enough of us do. And so, the latest version of Solaris does away with the mystery, the animal sacrifice, the practice of witchcraft, and the other requirements for mastery of earlier versions. Read how Ginny put away her potions and figured out the best way to use the new tools.

    Article 2
    Best Way to Automate ZFS Snapshots and Track Software Updates in Oracle Solaris 11

    Boot environments in Solaris 11 perform a function similar to Live Upgrade environments in Oracle Solaris 10. Except that they're implemented with ZFS. Which means you can generate snapshots of your boot environments at every point you'd like to record. And the beauty of that is, of course, that you can return to any snapshot of the boot environment that you want to use. In this article, Ginny introduces TimeSlider, shows you how to configure it to take automatic snapshots, and explains how to keep a record of the software updates that have been made to the current boot environment.

    Article 3
    Best Way to Update Software in Oracle Solaris 11 Zones

    Before the Zone there was the Container. And before the Container, the Zone. This is The Way of Software. In her third "Best Way" article, Ginny figures out the best way to manage software updates in Solaris 11 zones which, as you might expect, are different from Solaris 10 zones. After showing you those differences, she shows you how to create, configure, install, and clone a Solaris 11 zone, then how to upgrade both the global and non-global zones. As a bonus, you get to find out what to do if something goes wrong.

    We're expecting more "Best Way" articles from Ginny down the road. So read these, try out their recommendations yourself, and tell us what you think.

    And don't forget to save the lemur!

    - Rick

    Wednesday Aug 10, 2011

    Join OTN Or ...

    ...The Lemur Gets It!

    Turns out Oracle wants more sysadmins (Linux, Solaris, systems, storage, or network) to become official, bonafide, full-blown OTN members. I explained that sysadmins aren't really the "joining" type, but I lost. Oracle wants more sysadmins to join OTN. Period.

    So I've been wondering how I could convince the more reluctant among you to become official, bonafide, full-blown OTN members. After all you, your bookie, and your bookie's mother-in-law can read our technical articles, view our OTN videos (may take time to load), and visit just about every part of the OTN Systems website without signing up for anything.

    But there is a bunch of very cool stuff you can't do unless you're a member. This month I'll tell you about one.

    Download Software For Free

    You can't download our software for free unless you're an OTN member.

    I know Oracle's license terms are not the same as Sun's were, but you still get to download and horse around with world-class software for free. If you're anywhere within a decade of your mid-life crisis, you'll clearly remember when you had to actually pay a lot of money for good software. All we ask is that you be honest about when you deploy our software. That's only fair.

    For all the details, read the OTN developer license.

    You can read about other benefits of membership here.

    So, if you really want that lemur to have a future, sign up here, check the "Oracle Technology Network" box under "My Community Memberships," and identify yourself as a sysadmin.

    And while you're at it, sign up for our newsletter. It'll highlight the best content we've published over the previous month, in case you weren't paying attention.

    - Rick

    Monday Jul 18, 2011

    Women Soccer Players Don't Flop

    I'm now a rabid fan of women's soccer. Men's soccer? Phht! Men flop. Really. They could recruit Vlade Divac straight out of the NBA without letting him even see what a soccer ball looks like, and he'd be a superstar before mid-season.

    Sunday's match between Japan and the US was one of the best matches I've ever seen, in large part because the women played clean soccer. And because they exhibited none of the honor-sapping flopping that is prevalent in the men's game.

    What's this got to do with Oracle Solaris Cluster? Think about it. You can hit the Japanese women with a veritable tsunami of shots on goal, you can overwhelm them with skill, speed, power, and intensity. You can jump over them, power through them, score debilitating goals on them. More than once. And you can run them into the ground for 120 minutes. But they never go down. They never give up. They keep fighting. And they keep coming at you.

    Wouldn't you want that on your side?

    We just released version 3.3 5/11. Here are some resources to help you become familiar with it.

    Resources for Oracle Solaris Cluster 3.3

    By the way, that picture is not from the USA-Japan match, but from

    - Rick
    System Admin and Developer Community of OTN
    OTN Garage Blog
    OTN Garage on Facebook
    OTN Garage on Twitter

    Wednesday Jul 13, 2011

    Suicide by C++ And How to Avoid It

    photo courtesy of To Be A Pilgrim blog.

    You may have seen The Onion's story about some recent ground-breaking research ...

    New Study Shows People With Panic Disorders Respond Poorly To Being Locked In Underwater Elevators

    If you get the same reaction every time you realize your C++ code is hopelessly tangled up with your C++ libraries, you may find some relief in this series by Darryl Gove and Stephen Clamage.

    • Introduction to Libraries and Linking - How to make sure that your Oracle Solaris application links to the libraries it needs correctly and in the right order.
    • Part II - Resolving Symbols in Libraries - How to use the -z defs flag in your code to make sure the runtime linker links your application to the correct C++ libraries.
    • Part III - What Happens When An Application Starts - How investigate run-time application linking problems by using the LD_DEBUG environment variable.
    • Part IV - Avoiding Linking Problems - How to identify duplicate symbols and circular dependencies in your C++ code that would lead to linking problems between your application and its libraries.
    • Part V - Libraries in C++ - A detailed demonstration of how your C++ compiler can bind to a symbol from the incorrect library at compile time, and how to use the -g compiler flag to detect it.
    • Part VI - Resolving the Initialization Order Problem - Using link *order* to resolve C/C++ library dependencies is quick, but not optimal. Better to use either -instlib to keep the compiler from generating multiple templates, or use the -Bdirect linker flag to record dependencies at run time. This article explains how.
    • Part VII - Using Symbol Scoping to Avoid Linking Issues - By default, a symbol defined in a library is visible to other libraries and executables. You can limit a symbol's scope in a number of ways, including symbolic binding, hidden scope, and interposing. How to and examples described here.
    • Part VIII - Concluding Remarks and Summary of Best Practices - Building an application as a combination of executables and library calls has many advantages but potential problems with the links between the executable and its libraries. By using the techniques described in this series of articles, you can identify problems and make the risks manageable.

    - Rick
    System Admin and Developer Community of OTN
    OTN Garage Blog
    OTN Garage on Facebook
    OTN Garage on Twitter


    Rick Ramsey
    Kemer Thomson
    and members of the OTN community


    « September 2015
    Blogs We Like