Solaris vs. Linux: Framework for the Comparison

by Dr Nikolai Bezroukov

Dangers inherent in development and business models

Among the issues that concern large enterprises that adopt open source software and OSes, the following are the most prominent:

1. The level of fragmentation

2. The level of financial stability of the vendor (chances for long term survival)

3. The level of interoperability with Microsoft

4. The level of stability of market position of the vendor.

All of them are of course related, but we will treat them separately. 

There is also a natural tendency for larger OSS project toward bureaucratization and excessive rigidity of command and control structure tuned to a particular superstar developer (cult of personality). A large community inhibits innovation and bold architectural decisions so any OSS   project that reached significant level of popularity may stagnate architecturally.  As Jamie Zawinski put it:

"There exist counterexamples to this, but in general, great things are accomplished by small groups of people who are driven, who have unity of purpose. The more people involved, the slower and stupider their union is."

Even without complete stagnation with time large projects like linux tend to became more and more conservative in architecture (problem typical for most "vendor cooperatives" projects and probably one of the reasons of why linux missed the virtualization train). Of course those  tendencies are not absolute and much depends on the personalities involved. 

The level of fragmentation

There is no risk of fragmentation with the Solaris OS, since Sun is the only designated party managing production version of the OS and is the key player in the open source version of the OS. That provides Solaris a strategic advantage in comparison with linux. The latter repeats "Unix curse" on a new level and is split between multiple partially incompatible enterprise distributions (mainly Red Hat and Suse with some minor presence of Debian and its derivatives). This advantage of a single version should never be overlooked by enterprise customers.  As we mentioned multiple times there is a significant cost in supporting additional flavors of OS in enterprise environment and those cost increase dramatically as the number of distributions exceeds three. Paradoxically adoption of Linux almost guarantee that this "magic number" will be exceeded.

Although generally open source development model should probably be consider as a variation of  well known "applied science" model, the commercial model that is used by Red Hat and Novell are still pretty new and the long term prospects of it are difficult to predict with enough certainty. Several years ago Red Hat generated a lot of bad will by abandoning its desktop distribution. There are also some inherent limitations of scalability of commercial open source development due to the difficulty of attracting large pool of developers on the basis of the attractiveness of the project itself without sufficient monetary rewards. In a way programming for large ,complex project is as harmful for health as working for chemical industry and the question of compensation is not mute.  So volunteer and semi-volunteer (business cooperative open source development) labor pool, the labor pool that is behind most applications packaged with Red Hat can bring RH only as far as it is ready to go.  

Linux is notoriously fragmented distribution.  Initially Red Hat was the king of the hill (and it is still is). But recently Red Hat fall out of favor in IBM and Dell headquarters and Suse start rapidly gaining ground. Pack with Microsoft solidified the position of suse as the more suitable for large enterprise environment. Suse also have some technological advantages over Red Hat (among them I would like to mention Yast and AppArmor). Another dark horse is Ubuntu which definitely try to get a foothold in enterprise space but so far was not very successful.

Here is one telling comment from the discussion "Why is there no Open Source SLES ?"

Linux Fragmentation hurts

Submitted by Phasor Burn (not verified) on Sun, 2007/11/11 - 01:57.

At my workplace, our datacentre is 99.99% Linux. Our distribution of choice has always been SLES but over time we have had to install CentOS, RedHat, OpenSuse for various reasons (some commercial apps required platforms, etc).

As part of looking at what to do with this dogs breakfast, we are seriously considering going to Solaris x86, whether it is OpenSolaris or genuine Solaris is still up in the air.

If there's someone that understands how to build and support enterprise operating systems, it's Sun.

The level of financial stability of the vendor

Unique financial problems exists for Red Hat as a market leader due to usage of  "GPL-based" commercial model.  While Red Hat generally can be classified not as a development company but as a mutual fund, it might not be scalable much beyond its current market share of about 10% as GPL model does not protect the vendor from other companies that want to sell "copycat" distributions undercutting Red Hat price.

That also necessitates increased charges for support (as we mentioned above, Red Hat has very expensive support contracts) and can leave Red Hat (and also, possibly, Novell which currently enjoys more favorable position of No.2  in this market share race) with cash flow problems that might negatively influence the quality of support.

But it is mainly Red Hat who is involved in a costly "war of clones" with CentOS and other "rebuild" of its Enterprise Server.  Novell's problems currently are due the decision of many (but not all) major Netware customers to move to Windows 2003 instead of Suse as well as problems in positioning Suse as a legitimate and technically superior successor of Novell Netware.  Still the portfolio of products that Novell has make is unique and promising enterprise distribution, the distribution that enjoys considerable "goodwill". Recent pact with Microsoft made SUSE much more attractive for enterprise customers especially those who use Active Directory, so in this respect Red Hat needs to deal with two dangers simultaneously: hijacking support revenue by Oracle and better integration with Microsoft from Suse. Red Hat, then, could be a casualty in a larger fight: the fight between Oracle and Microsoft.

Actually only during writing of this paper I started to appreciate Suse as a contender in Unix enterprise space that in several areas of our framework is superior to Red Hat.  Novell actually represents more serious competitor for Solaris too as Sun brass missed the opportunity to capitalized on its pact with Microsoft several years ago. For example, for Solaris 10 the integration with Active Directory (AD) is rather weak and badly documented as Sun tries to sell its own directory services which is a rather difficult task in case customers already use AD.  Suse 10 is compatible with AD out of the box.

Dangers connected with open source development model are partially applicable to Sun too, especially in view of its recent flirt with GPL, but so far the company does not use GPL-based model in Solaris. Currently Sun uses GPL for UltraSparc chips, Open Office (as a second license) and Java (also as a second license).

BTW despite my strong allergy to GPL I should admit that GPLing Java was a pretty interesting move that put IBM on the defensive and might give Java technology a second breath. Unlike Red Hat, Sun can partially  rely on the revenue from its hardware business to compensate for losses in OS and software development area but its capabilities to absorb such losses are limited and make sense if and only if such move substantially increases hardware sales.

Financial considerations might weight both for Sun, Novell and another major linux supporter (at least in PR dimension) IBM.

Sun and its servers were very popular amoung larger dot-com companies in late 1990s; the company even dubbed itself the "dot in dotcom". But when the bubble burst at the end of 2000, Sun's share price sank to less than 10% of its peak - and stayed there.

Novell suffered a difficult time when in 2001 Eric Schmidt before his move to Google acquired a technology consulting company, Cambridge Associates, under the fuzzy logic that "Cambridge accelerates Novell's adoption of a solutions-selling model that supports customers and partners transforming their businesses" . Eric Schmidt, once viewed as the company's savior essentially sunk his former company with $266 million acquisition of technology consultant Cambridge Technology Partners. After he left Jack Messman for CEO of run Cambridge Technology Partners run the combined company. Messman engineered Suse acquisition in 2003 (for 210 Millions). He was ousted from CEO position in June 2006 and soon has quit the company's board of director. Just before that Novell sold off its Celerant consulting unit which came to Novell as part of its acquisition of consultancy Cambridge Technology Partners for pennies on dollar (for $77 million.) To please Wall Street it also bought up $400 million in its own shares as it reported its financial results for its fiscal second quarter ended April 30, 2006.  Novell recovered recently after Suse 10 was released and sales improved but still turnaround is far from complete. As Reuters noted in March 2007 Novell in every quarter for more than a year has either missed analysts’ targets, made disappointing forecasts or both. Of course, Novell’s losses aren't universal and SuSE Linux business is growing fast: quarterly Linux revenue grew 46% to $15 million But as TechIQ noted "Unfortunately that’s only about 7% of Novell’s annual revenue—which the company says will range from $945 million and $975 million. Translation: SuSE Linux sales just aren’t growing fast enough to offset plummeting NetWare sales."

Some financial observers claim that IBM might play a dangerous game with its shares, artificially keeping the prices of shares growing. As Jim Jubak noted in his May 8, 2007 column in MSN Money:

Even big players such as IBM Corp. (IBM, news, msgs) are piling on debt to repurchase their own shares.

Since 2003, IBM has purchased 203 million of its own shares at a total cost of $30.7 billion. That's a huge percentage -- about 52% -- of the company's total operating cash flow of $59.5 billion during the period. It looms even larger if you add in the $17 billion IBM spent during this period on capital expenditures, the $8.8 billion it spent acquiring businesses and the $5.3 billion it spent paying dividends to investors. All that -- added to the spending on buying its own shares -- comes to 104% of operating revenue.

Or look at it another way. In 2006, IBM used the equivalent of 67% of its total net income to buy back shares. In 2005, the percentage was 82%. In the two years before that, 64% and 42%, respectively.

If you add in dividends, 2006 payouts to investors came to 85% of total net income at IBM.

That's the level of payout ratio that sends up a red flag to investors, I've been taught, because it's clearly not sustainable over the long run. In fact, from its financial statements it looks like IBM is borrowing to keep up this level of payout while keeping its business running as usual. Cheap money makes that possible.

Internally the company might have problems too. As Slashdot noted in May 13, 2007 article "In an e-mail worthy of the Dilbert Hall of Fame, IBM execs responded to Robert X. Cringely's Project LEAN layoff rumors, reassuring employees by pointing out that they've already wiped out too many U.S. jobs to be able to lay off another 150,000. Big Blue's employment peaked around 1985, when it had about 405,000 workers who were acclimated to a long tradition of lifetime employment. IBM puts its current global workforce at 355,766, with a 'regular U.S. population' of less than 130,000.". While Robert X. Cringely  might get the number wrong some posts of IBM employees are pretty revealing:

• I, Cringely . Pulpit . Lean and Mean . Comments PBS
• I, Cringely . Pulpit . The Final Daze . Comments PBS

Here is a letter from IBM executives stating that rumors of massive layoffs are false:

Subject: Rumors of massive layoffs

We have received many inquiries regarding the subject. If IBM
responded to every rumor, we would get distracted from the important
work of delivering value to our clients.

However, a recent external blog report suggesting that IBM is
planning a massive layoff is causing unneeded activity. If this blog
is generating concern in your unit, please feel free to use this
information to assure your teams and business leaders that the blog is
inaccurate, and relies on gross exaggerations.

The blog suggested we would be letting go more IBMers than we
currently employ in the U.S. The facts are that our regular U.S.
population is just under 130,000 IBMers -- a number that has remained
relatively stable in recent years, as we have divested and acquired
businesses and continued to invest through new hiring.

We said when we released 1Q results we would be putting in place a
series of actions to address cost issues in our U.S. strategic
outsourcing business. We have undertaken efforts toward that, and
recently implemented a focused resource reduction in the U.S. While
any such reduction is difficult for those employees affected, these
actions are well within the scope of our ongoing workforce rebalancing
efforts.

The blog also completely misinterpreted our efforts around Lean. To
fully understand Lean, you have to view it in a strategic context -- a
key part of what we're doing to reinvent service delivery to provide
more value to clients and make IBM more competitive. We are using
Lean, which is a commonly used methodology to conduct process design
and development, to make informed decisions about how to improve and
streamline processes. We are going about that in a disciplined and
rigorous way, and the intent, as it has always been, is to improve our
speed, quality and responsiveness to clients.

Pat Kerin Joanne Collins-Smee Patt Cronin

The level of stability of market position of the vendor

Large enterprises tend to prefer the leading vendor with the most strong marketing position. but the question arises how solid is this market position.  for Solaris the question does not make sense as there is only vendor and only one flavor of Solaris. But for linux situation is more complex. For example Red Hat could be a casualty in a larger fight: the fight between Oracle and Microsoft.

Until recently Red hat was the dominant linux vendor and as such enjoyed the largest enterprise penetration. but the situation changed after Novell acquired Suse. There is a lt of ground-level support of Novell in large enterprises and it is unclear whether Red Hat will be able to present its dominant position for another three-five years.  Recent events like Novell pact with Microsoft and Oracle decision to provide support for Red Hat distribution does not increase confidence in Red Hat to say the least.  BTW when I checked the prices in late April, 2007, I noticed that Dell dropped Red Hat support from its PowerEdge 2950 servers for small business. Only Suse is now available for pre-instillation [Dell2007]:

Windows Server® 2003 R2, Standard Edition, Includes 5 CALs [add $799]

 

Windows Server® 2003 R2, Enterprise Edition, Includes 25 CALs [add $3,295]

 

Windows Server® 2003 R2, Standard x64 Edition,Includes 5 CALs [add $799]

 

Windows Server® 2003 R2, Enterprise x64 Edition, Includes 25 CALs [add $3,295]

Windows® Server 2003, Web Edition [add $349]

 

SUSE Linux Enterprise Server 10, Up to 32 CPUs, 1 YR Subscription, FI [add $280]

 

SUSE Linux Enterprise Server 10, Up to 32 CPUs, 3 YR Subscription, FI [add $698]

 

VMware VI 3.0.1, Standard 2 CPU,1Yr Support and Upgrade Subscription, No CD [add $4,688]

 

VMware VI 3.0.1, Standard 2 CPU,3Yr Support and Upgrade Subscription, No CD [add $6,563]

 

No Operating System [Included in Price]

The level of interoperability with Microsoft

The level of interoperability with Microsoft is important for large enterprise for many reasons, the main of which is the presence of large Windows server park and almost universal deployment of Windows clients for staff.

Linux traditionally have very strong level interoperability with Microsoft because it shares the same hardware platform. That happens despite "anything but Microsoft" stand of many linux evangelists. And among linux distributions the most "Microsoft-friendly" is Novell Suse. Unlike Novell, Sun does not have extensive experience with Windows interoperability and its pact with Microsoft essentially failed to provide any compatibility improvements for Solaris. At the same time Novell Netware was for many years the standard of interoperability with Microsoft products and the recent pact with Microsoft might help Novell utilize this advantage on Linux side of its business. As a company Novel has years of experience in interoperating with Microsoft products. Sun just does not has level of experience. Especially dangerous for Solaris 10 is the growing gap in the ease and scope of integration with Active Directory. In February 2007 Novell and Microsoft announced they are working together to develop a method for using existing protocols for bridging network access between eDirectory and Active Directory, with complete details to come sometime during the first half of this year.  Samba4 will makes Active Directory even more linux-friendlier. For example, suppose I need to set up a new an account for an additional server administrator who needs to have an ssh access to subset of servers, access to root accounts on those servers, maybe a couple of other miscellaneous local accounts on other servers. The full usage of Active Directory gives a single point to control who is authorize to do what and promptly revoke those privileges in case the person left the company. Otherwise we have  a problem of dormant accounts. Integration with AD also reduces the number of things administrators have to accomplish in order to grant somebody the access to particular resources. Here is another business case: Linux pro turns to Active Directory identity management

Research in Motion (RIM) systems architect Ian Brown wanted to give his administrators the same central identity management authentication functionality for their Linux and Unix machines that they enjoyed with Microsoft Windows Server and Active Directory -- but it just wasn't happening.

While the Windows boxes enjoyed the central authentication management hub in Microsoft Active Directory (AD), the Red Hat Linux, Sun Solaris and HP UX boxes in his data center needed to be administered on a server-by-server basis.

"Essentially, we wanted to be able to use a central authentication source for all of our servers for our half dozen administrators," Brown said. "We could get that through Unix, but that would have required us to have a central login server and we basically can get that for free already with Active Directory." The login server would have been redundant; Brown wanted to leverage what AD had to offer.

Central authentication would reduce the "unnecessary overhead" of server-by-server password upkeep, Brown said. In some cases, administrators find scrubbing Linux and Unix servers of employee login credentials so time consuming that it is not done at all, leading to security threats. That wasn't the case at RIM, where turnover was relatively light and a series of checks and balances assured every server was scrubbed when employees left the company, Brown said. The bigger threat, Brown said, was actually a looming Sarbanes-Oxley (SOX) compliance audit.

See also  Authenticating to Windows Active Directory 2003 - SUSE Wiki
 

Etc

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 quotes :  Somerset Maugham : Marcus Aurelius : Kurt Vonnegut : Eric Hoffer : Winston Churchill : Napoleon Bonaparte : Ambrose Bierce :  Bernard 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 DOS :  Programming Languages History : PL/1 : Simula 67 : C : History of GCC development :  Scripting 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-Month :  How 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 Technology is dominated by two types of people: those who understand what they do not manage and those who manage what they do not understand ~Archibald Putt. Ph.D

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 Softpanorama society. We do not warrant the correctness of the information provided or its fitness for any purpose. 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.

Created Jan 2, 2005.  Last modified: March 12, 2019