Softpanorama

May the source be with you, but remember the KISS principle ;-)
Contents Bulletin Scripting in shell and Perl Network troubleshooting History Humor

Solaris vs. Linux: Framework for the Comparison

by Dr Nikolai Bezroukov


 

Prev Contents Next

1. Introduction

It is important to understand that large enterprises represent a different class of IT environments then home enthusiasts, startups or Internet service providers (ISP). For example there is no such thing as free enterprise Unix distribution: you need to pay one way or the other for such enterprise level features as regular patch updates and technical support (actually support-wise Red Hat Enterprise Server is  more expensive then Solaris).  Basically, large enterprises rightly consider Linux to be not free (as in beer) as it actually has a significant costs attached. TCO includes not only license costs, but also annual maintenance contract  cost, administrators costs, hardware cost, training costs, consultant related costs (in case of difficulties with important applications), etc.

Also hardware compatibility generally is not an issue: servers are purchased only from top vendors, vendors which guarantee compatibility with the OS (given the complexity of the drivers of modern OSes it's risky to use them on arbitrary hardware). Large enterprises are accustomed to pay a slight premium for hardware that provides better reliability and manageability. They are also the main consumers for mid-range and large ("mainframe") servers. The latter still is an important battlefield of IBM and SUN for the large enterprise customers.

System administrators in large enterprises work in very different conditions that their colleagues in small firms and startups and have different challenges. The main difference is that they usually need to administer several different flavors of Unix that got into the company either as a result of mergers and acquisitions or because of changing preferences of different generations of IT brass at different periods of company history. Actually mergers and divestitures create unique challenges in large enterprises system administrator world as you need to merge two different cultures not only two different IT infrastructures. Often that means the choice of the lowest common denominator and not only ""expenses rise to meet income" type of application of Parkinson law to IT.  The ideal of social stability often is viewed in the large corporate IT within the context of heavy spending for the purpose of setting up additional management positions.  Outsourcing and downsizing definitely interfere with this ideal, but not to the extent many think.

While Solaris is currently one of most common large enterprise flavors of Unix (along with AIX and HP-UX) it not uncommon to have two of three bland of Unix deployed in the infrastructure. At the same time linux is a newcomer and as such brings the danger of an managing additional brand of Unix with the same staff as before. So attitude to the linux introduction among rank and file Unix administrators are mixed and quite complex despite the fact that most of them use linux on their office workstations and/or home computers for many years. It is one thing to use it on workstations and the other to use it as a part of infrastructure when each minute of downtime costs several thousand dollars.  You need to be conservative in such an environment in order to survive. 

Like linux, Solaris is an important part of Unix Renaissance and is the oldest and the most prominent commercial brand of Unix (out of commercial brands only SCO Unix is older).  Now in addition to that Solaris became the most stable and well supported free OS. Sun (still) provides very good patch support (for Solaris 10 only security patches are free). For a free OS it has unsurpassed scalability. For only the cost of media and shipping, organizations can use the software on an unlimited number of computers with a capacity of 8 or fewer CPUs. Individual users can download Solaris for free.

For anybody who can use spreadsheets for calculation of savings it is clear that it not so much cheaper hardware and OSes but open source applications that provide real savings in large enterprise environment. And first of all major infrastructure applications (web servers, mail servers, DNS servers, etc) and scripting languages (Perl, PHP, Python, Ruby, JavaScript to name a few). PHP5 now is becoming a really cost-effective alternative to Java in large enterprise environment and is enjoying considerable IBM support (investments in Zend). Lower development and maintenance cost inherent in "very high level" nature of scripting languages gives them a tremendous advantages over traditional "Algol-level" languages on modern hardware. Usage of open source databases is also on the rise and savings for large SMP configurations and multicore CPUs (for example Suns T1 and IBM PowerPC) are very real. open source databases dominate Web hosting environment and gradually get a footprint at enterprise, especially as Web server backend. Not surprisingly all major commercial database vendors try to reverse this trend offering scaled down free versions of their products with specifications that exactly fit mainstream Web backend usage. Here essentially the compromise is structured around the possibility of buying better hardware due to dramatically lower cost of the software.

The author argues that Linux is just "yet another" implementation of Unix and its developers has no monopoly over best technical solutions either in the area of technical innovations or in the area of large Unix server farms administration. For example Linus Torvalds completely missed virtualization train and here (as of end of 2005) Linux is inferior to other free OSes (both Solaris and FreeBSD). History of the first ten years of Linux development can be found in [Bezroukov2005a]. But the problem is that many Linux users do not know the history and deride Solaris not understanding that in many (but not all) important enterprise level features (for example the quality of light-weight VM support, the quality of NFS implementation, stability, scalability and manageability) Solaris is ahead of Linux.

Large enterprises less often make their choices on the base of religious or philosophical doctrines and more often base them on the capabilities and TCO of the products. Currently security, virtualization and reliability are three major areas were OSes are compared along with TCO. Solaris 10 is a pretty revolutionary OS in all those three aspects, the first flavor of Unix that deserves to be called XXI century Unix. Its ability to withstand hardware failures is probably the best in industry. It has stack overflow protection and light weight VMs called zones, two features that provide huge security boost for applications. It has kernel that is more technically advanced then Linux kernel and with the opening of Solaris 10 on X86 it definitely should be considered an option by those enterprise customers who already have significant Solaris deployment.  

The same is true for enterprise software which is certified for specific OSes or even a specific version of a particular OS. The initial cost of software is only one factor in the complex large enterprise picture. And as Microsoft successfully demonstrated driving it to zero does not necessary mean winning large enterprise accounts even for powerhouses like IBM with it relentless and well-oiled Linux propaganda machine.  The same is true for middleware.

The main problem for linux in large enterprise environment is that it's usually more cost effective to extend existing OS to a new platform (X86) then to add new flavor of Unix to the preexisting Unix flavours mix. The key problem here that adding linux to the enterprise Unix mix "disrupt exiting ecosystem" and overloads existing sysadmin staff.  Without elimination existing flavor of Unixes such move tend always be more expensive then extending Solaris to Opteron platform in organizations that already have significant Solaris presence and sysadmins who know Solaris well.

Still, linux has its own value in large enterprise environment, the value which go beyond cost savings. For example, the ability to fight corporate bureaucracy by creating prototypes using existing workstations is one of the key factors that favors linux in large enterprise environment. We should never ignore the fact that Linux represents a significant advantage in pilot projects, permitting to shortcut corporate bureaucracy. It is especially valuable for those projects that can benefit from using new or wider spectrum of existing software technologies. Here the level of support of open source applications as well as much better availability of prepackaged open source applications makes linux more attractive then Solaris as it allows more easy experimentation.  Unless Sun pays proper attention to this very important issue it might negatively  affect the level of Solaris 10 adoption despite all its obvious technical advantages. Being the development OS for most open applications is an "home field" advantage. It is not that easy to overcome linux home field advantage without major investments. Beyond development tools, large vendors like IBM routinely provide free entry-level editions of products as competitive weapons to crash smaller competitors by releasing them for linux.  As Gordon Haff  noted "The general approach is to pick a non-strategic product segment or area that just happens to be important to a key competitor." [Haff2007].   IBM Eclipse story is a nice illustration of this approach.  It really eclipsed other development tools for Java and served as a Trojan horse for selling IBM expensive, not that reliable and extremely complex Java middleware. Haff cites IBM’s release of its VisualAge software development tools to the open-source Eclipse Foundation in 2001 as a move that dealt a fatal blows to commercial Java IDEs such as Borland’s Jbuilder and Symantec’s Café. "Some collateral damage to competitors in the process is not something that IBM likely regrets," Haff wrote. Perversion of noble ideas and associated slogans for pretty nasty practices is a common metamorphose that occurred regularly in human history.  The ideas of open source and free software are not the first and not the last in this long line.

Similar maneuvering for a better shot on competitor is clearly visible in attempts by various vendors to push enterprise linux adoption. In any large enterprise infrastructure "savings nullification effect" from the introduction of a new flavor of Unix mentioned above happens independently of real or imaginable OS virtues or costs of the hardware on which it can run. Articles that boast about huge savings due to introduction of a new flavor of Unix in a large enterprise environment without complete elimination of at least two old flavors of Unix are very misleading and claims about savings are open to review. They probably should be treated mostly as marketing materials (with elements of self-promotion, as the presence of the requisite large photo of smiling CIO strongly suggests). In reality adding another flavor of Unix in a large enterprise IT environment usually increases not decreases the TCO.  One CPU architecture and one OS in large enterprise environment is like putting all eggs into one basket and the enterprise essentially became a hostage of a particular vendor, but too many architectures and too many OSes are also bad and very expensive to maintain. Extremes meet and here too an enterprise tends to become hostage, in this case the hostage of  unscrupulous consultants.  It is important to understand that  label "open source" is regularly used for gaining a competitive advantage by one large vendor against another with IBM as a flagship of such usage. That explains some attempts to push linux though the throat of large corporations IT infrastructures: the last thing such often highly placed pushers are interested is benefits for a particular organization. It's all about elimination of competitors although it is sold under the disguise of noble slogans about of advantages of open source and lower costs. The fact that in many cases expected costs savings never materialize should not come as a huge surprise; this is just the nature of the game.

The paper also draws attention to such important but often neglected in OS comparisons area as frequency of patching and its cost for large enterprises. An OS which allows to use less frequent patching schedule has a tremendous advantage in such an environment. This issue might well be an Achilles' point of linux. I would like to stress that TCO for any OS in large enterprise environment now significantly depends on the frequency of patching. And that fact alone explains the rationale behind many Windows to Unix migrations, migrations that happen despite high quality of Windows Server 2003. 

As Solaris has regular quarterly updates that have extremely high level of compatibility with previous versions, individual users generally can substitute quarterly updates for patching and thus, the absence of free recommended patch cluster in Solaris 10 is not critical for this category of the users. In this sense Solaris is as democratic OS as Debian and much more democratic then Red Hat or Suse with their artificial designation of alpha versions as "free community distributions" and more debugged versions as "commercial" that require support contract to use. 

But the most important feature of large enterprise environment that the paper tried to stress is that Linux needs to be "implanted" into pre-existing Unix mix that usually already consists of two or more flavors of Unix. Enterprise Unix environment behaves more like a complex ecosystem and like in nature invasion of new species has often unpredictable effects. The author argues that this fact alone often nullify all the savings from adding any new flavor of Unix including Linux. Also with Solaris 10 availability on Opteron there generally should be other arguments then hardware cost savings: they are almost zero. Solaris 10 runs well on Opteron based servers both from Sun and HP. Also with the introduction of Intel Duo and Quattro CPUs  high-end server (8 CPUs or more, 32G of RAM or more) are reality and Solaris has a very good track record in this segment.  While there are differences between Solaris 10 on Opteron and Solaris 10 on UltraSparc they are much less then between any enterprise class Unix and linux as well as between any two linux distributions, for example Red Hat (as well as now Oracle Unbreakable Linux which is based on Red Hat) and Suse. Still due to the level of differences between UltraSparc and Intel platforms the cost of deploying Solaris on Opteron is definitely not a zero and is close of equal to the introduction of a new Unix flavor that increases the diversity of enterprise Unix ecosystem.  Like is the case with linux there is no free lunch here despite low initial cost of the OS itself (actually the cost of OS is zero and the price we are talking about is the annualized price of maintenance contract).

All-in-all the paper is intended to provide "slightly skeptical" information about relative merits of linux and Solaris. Each has strong and weak points and as always weak points often represent the logical continuation of the strong. Neither is a dream OS for system administrators and we are mostly concerned about determining "lesser evil" then "clear winner". Also much depends of preexisting UNIX flavors mix: the real headache starts when you need to administer them in addition to one other Unix flavor. 

The author hopes that the framework of comparison which was developed in the paper has value independently of whether the readers agree or disagree with some assessments of the author. I also hope that it might help current and future decision makers and might inspire other authors to revisit the topic and/or to extend it to the comparison of different pairs of enterprise class OSes.

Linux  and linux

Linux is a loaded word. It denotes a particular free "almost POSIX-compatible" OS kernel developed with a huge PR campaign for the last 15 years and, simultaneously, multiple semi-compatible distributions based on this kernel.

Here and below we will try to distinguish between Linux as an OS and Linux as a kernel by using lowercase word for the OS ("Red Hat is  one of many distribution of linux" ) and upper case word for the kernel ("Linux 2.6"). 

We will use the word "Linux" to mean "Linux kernel" and "linux" to mean "Linux distribution"

When we are talking about linux as an OS we will generally mean Red Hat as the dominant linux enterprise distribution. In case we need to compare different enterprise distributions or stress the internal fragmentation of the enterprise linux market we will use the respective names of distributions, for example Red Hat and Suse.  When talking about Red Hat we generally mean RHEL 4.x and about Suse, Suse 10.x: two versions that were current and were preloaded by Dell on its servers in the second half of 2006 when the bulk of the paper was written.  The choice of Suse 10 is a little bit arbitrary as it was pretty new distro launched in August 2006 and I understand that, but I really prefer it to Suse 9 and I am reluctant to write about Suse 8: the last "good" Suse distribution before 10. In this sense the paper is slightly pro-Novell biased ;-).

There are two major types of open source software vendors: those who try to protect their profits at the expense of the open source community, and those who try to protect the community while leaving profits aside.  The first type of vendors in the most numerous and includes both Red Hat and Novell. It also has better survival characteristics (Red Hat is already seven years old company).  Canonical Ltd (Ubuntu vendor) currently belongs to the second type.  I do not know how long it will stay in this role.

"Rulez" and usage of slang and other idioms

For the rest of the paper it is assumed that readers understand the differences in approaches between enterprise linux vendors as exemplified by Red Hat and Suse distributions and community distributions such as Debian, Gentoo, or the newest kid on the block, Ubuntu. Enterprise linux vendors need to survive and they consciously or unconsciously need to practice "free as in free lunch" approach to the distributions within the limits permitted by GPL (here they actually proved to be very inventive, beating RMS it his own game).  

As I mentioned before the paper was written with the explicit goal to serve as an antidote to primitive reviews on Linux self-congratulation sites styled like "I found old PC in the closet, dusted it off, tried to install Solaris on it; my God what a crap Solaris is in comparison with Linux". Such reviews are not only misleading, they disorient open source enthusiasts (especially among IT staff of large companies) substituting wornout and most false clichés for critical thinking and conditioning them against a more stable and in several areas (virtualization is one) more advanced server OS that has a lot to offer. While Linux is a great achievement and now has an undisputable large enterprise penetration, with the current complexity of Unix it can benefit from the specialized hardware tuned to the OS, not the OS written with all the compromises inherent in the achieving the capability to run on mass-produced hardware and attract mass users.

That is another reason why in my opinion Solaris still plays and will continue in the foreseeable future to play an important role in the large enterprise IT environment. 

 
Prev Contents Next



Etc

FAIR USE NOTICE This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available in our efforts to advance understanding of environmental, political, human rights, economic, democracy, scientific, and social justice issues, etc. We believe this constitutes a 'fair use' of any such copyrighted material as provided for in section 107 of the US Copyright Law. In accordance with Title 17 U.S.C. Section 107, the material on this site is distributed without profit exclusivly for research and educational purposes.   If you wish to use copyrighted material from this site for purposes of your own that go beyond 'fair use', you must obtain permission from the copyright owner. 

ABUSE: IPs or network segments from which we detect a stream of probes might be blocked for no less then 90 days. Multiple types of probes increase this period.  

Society

Groupthink : Two Party System as Polyarchy : Corruption of Regulators : Bureaucracies : Understanding Micromanagers and Control Freaks : Toxic Managers :   Harvard Mafia : Diplomatic Communication : Surviving a Bad Performance Review : Insufficient Retirement Funds as Immanent Problem of Neoliberal Regime : PseudoScience : Who Rules America : Neoliberalism  : The Iron Law of Oligarchy : Libertarian Philosophy

Quotes

War and Peace : Skeptical Finance : John Kenneth Galbraith :Talleyrand : Oscar Wilde : Otto Von Bismarck : Keynes : George Carlin : Skeptics : Propaganda  : SE quotes : Language Design and Programming Quotes : Random IT-related quotesSomerset Maugham : Marcus Aurelius : Kurt Vonnegut : Eric Hoffer : Winston Churchill : Napoleon Bonaparte : Ambrose BierceBernard Shaw : Mark Twain Quotes

Bulletin:

Vol 25, No.12 (December, 2013) Rational Fools vs. Efficient Crooks The efficient markets hypothesis : Political Skeptic Bulletin, 2013 : Unemployment Bulletin, 2010 :  Vol 23, No.10 (October, 2011) An observation about corporate security departments : Slightly Skeptical Euromaydan Chronicles, June 2014 : Greenspan legacy bulletin, 2008 : Vol 25, No.10 (October, 2013) Cryptolocker Trojan (Win32/Crilock.A) : Vol 25, No.08 (August, 2013) Cloud providers as intelligence collection hubs : Financial Humor Bulletin, 2010 : Inequality Bulletin, 2009 : Financial Humor Bulletin, 2008 : Copyleft Problems Bulletin, 2004 : Financial Humor Bulletin, 2011 : Energy Bulletin, 2010 : Malware Protection Bulletin, 2010 : Vol 26, No.1 (January, 2013) Object-Oriented Cult : Political Skeptic Bulletin, 2011 : Vol 23, No.11 (November, 2011) Softpanorama classification of sysadmin horror stories : Vol 25, No.05 (May, 2013) Corporate bullshit as a communication method  : Vol 25, No.06 (June, 2013) A Note on the Relationship of Brooks Law and Conway Law

History:

Fifty glorious years (1950-2000): the triumph of the US computer engineering : Donald Knuth : TAoCP and its Influence of Computer Science : Richard Stallman : Linus Torvalds  : Larry Wall  : John K. Ousterhout : CTSS : Multix OS Unix History : Unix shell history : VI editor : History of pipes concept : Solaris : MS DOSProgramming Languages History : PL/1 : Simula 67 : C : History of GCC developmentScripting Languages : Perl history   : OS History : Mail : DNS : SSH : CPU Instruction Sets : SPARC systems 1987-2006 : Norton Commander : Norton Utilities : Norton Ghost : Frontpage history : Malware Defense History : GNU Screen : OSS early history

Classic books:

The Peter Principle : Parkinson Law : 1984 : The Mythical Man-MonthHow to Solve It by George Polya : The Art of Computer Programming : The Elements of Programming Style : The Unix Hater’s Handbook : The Jargon file : The True Believer : Programming Pearls : The Good Soldier Svejk : The Power Elite

Most popular humor pages:

Manifest of the Softpanorama IT Slacker Society : Ten Commandments of the IT Slackers Society : Computer Humor Collection : BSD Logo Story : The Cuckoo's Egg : IT Slang : C++ Humor : ARE YOU A BBS ADDICT? : The Perl Purity Test : Object oriented programmers of all nations : Financial Humor : Financial Humor Bulletin, 2008 : Financial Humor Bulletin, 2010 : The Most Comprehensive Collection of Editor-related Humor : Programming Language Humor : Goldman Sachs related humor : Greenspan humor : C Humor : Scripting Humor : Real Programmers Humor : Web Humor : GPL-related Humor : OFM Humor : Politically Incorrect Humor : IDS Humor : "Linux Sucks" Humor : Russian Musical Humor : Best Russian Programmer Humor : Microsoft plans to buy Catholic Church : Richard Stallman Related Humor : Admin Humor : Perl-related Humor : Linus Torvalds Related humor : PseudoScience Related Humor : Networking Humor : Shell Humor : Financial Humor Bulletin, 2011 : Financial Humor Bulletin, 2012 : Financial Humor Bulletin, 2013 : Java Humor : Software Engineering Humor : Sun Solaris Related Humor : Education Humor : IBM Humor : Assembler-related Humor : VIM Humor : Computer Viruses Humor : Bright tomorrow is rescheduled to a day after tomorrow : Classic Computer Humor

The Last but not Least


Copyright © 1996-2016 by Dr. Nikolai Bezroukov. www.softpanorama.org was created as a service to the UN Sustainable Development Networking Programme (SDNP) in the author free time. This document is an industrial compilation designed and created exclusively for educational use and is distributed under the Softpanorama Content License.

The site uses AdSense so you need to be aware of Google privacy policy. You you do not want to be tracked by Google please disable Javascript for this site. This site is perfectly usable without Javascript.

Original materials copyright belong to respective owners. Quotes are made for educational purposes only in compliance with the fair use doctrine.

FAIR USE NOTICE This site contains copyrighted material the use of which has not always been specifically authorized by the copyright owner. We are making such material available to advance understanding of computer science, IT technology, economic, scientific, and social issues. We believe this constitutes a 'fair use' of any such copyrighted material as provided by section 107 of the US Copyright Law according to which such material can be distributed without profit exclusively for research and educational purposes.

This is a Spartan WHYFF (We Help You For Free) site written by people for whom English is not a native language. Grammar and spelling errors should be expected. The site contain some broken links as it develops like a living tree...

You can use PayPal to make a contribution, supporting development of this site and speed up access. In case softpanorama.org is down you can use the at softpanorama.info

Disclaimer:

The statements, views and opinions presented on this web page are those of the author (or referenced source) and are not endorsed by, nor do they necessarily reflect, the opinions of the author present and former employers, SDNP or any other organization the author may be associated with. We do not warrant the correctness of the information provided or its fitness for any purpose.

Created Jan 2, 2005.  Last modified: June 04, 2016