From: ** http:// * ** osoft.com/ntserver/nts/news/msnw/LinuxMyths.asp * ** * BB ** Linux * B BB ** Poste * B B ** * BB B ** Wit * B BB B OO ** around Linux as an operating system, it's im * B BB O OO ** hype and look at the reality. First, it's worth n * BB B O O ** operating system. Linux fundamentally r ** BB B O O ** tem technology and architecture. Linux wa ** BB O O L ** o support symmetrical mult ** O O L ** nterfaces (GUI), asynchronous I/O, fine-g ** O O L ** er important characteristics of a modern o ** OO O L ** l limitations mean that as customers ** OO L L ** vely deploy scalable, secure, and robust a ** L L ** liver on the hype. ** LL L ** Myth: Linux perf ** LL L ** ** LL L OO ** Reality: Windows NT ** L O OO ** tomer Workloads ** LL O O ** ** LL O O ** The Linux community claims ** LL O O CC ** scalability in the latest ve ** O O C CC ** owever it's clea that Linux remains inferior to ** O O C C ** em. ** OO O C ** For File and Print services ** OO C KK ** conducted by PC Week Labs, th ** C K ** system delivers 52 percent bett ** C K KKKK ** processor system and 110 percent ** CC KKKKK ** 4-way system than similarly configu ** CC K K SS ** 4-way Linux/SAMBA systems. ** K K S SS ** For Web servers, the same PC Week tests ** K K S SS * 4.0 with Internet Information Server 4.0 ** K S * better performance on a single processor sy ** K SS * percent better performance on a 4-way system ** S * Apache. ** SS S * For e-commerce workloads using secure sockets (SS ** SS S * PC Magazine tests showed Windows NT 4.0 with Intern ** SS * Information Server 4.0 delivers approximately five ti ** * performance provided by Linux and Stronghold. ** * For transaction-orientated Line of Business applications, ** Windows NT 4.0 has achieved a result of 40,368 tpmC at a cost f $18.46 per transaction on a Compaq 8-Way Pentium III XEO ** cessor-based system. This industry leading pr * ** ormance result from the Transaction Processing P * ** Council (TPC) clearly shows how Windows NT * BB ** rld-class performance for heavy duty transaction * B BB ** interesting to note that there is not a single * B B ** database running on Linux, and therefore * BB B ** onstrate their capabilities as a database * B BB B OO ** * B BB O OO ** alability is architecturally limited in * BB B O O ** supports 2 gigabytes (GB) of RAM ** BB B O O ** mpared to 4 GB for Windows NT ** BB O O L ** ux supports is 2 GB versus 16 ** O O L ** 0. The Linux SWAP file is ** O O L ** nux does not support many o ** OO O L ** res that Windows NT 4.0 has ** OO L L ** , completion ports, and f ** L L ** hitecture constraints limit t ** LL L ** ast two processors. The Linux ** LL L ** or SMP and performance ** LL L OO ** sing these since the dev ** L O OO ** Delivering a scalable system ** LL O O ** ar that the Linux community c ** LL O O ** ckly. As D. H. Brown Associa ** LL O O CC ** ort,2 the Linux 2.2 Kernel rema ** O O C CC ** g a tuned SMP kernel. ** O O C C ** ** OO O C ** Myth: Linux is more reliable than ** OO C KK ** ** C K ** Reality: Linux Needs Real World Proof ** C K KKKK ** Stories ** CC KKKKK ** ** CC K K SS ** The Linux community likes to talk about Linu ** K K S SS ** operating system, yet there are no real world ** K K S SS * limited customer evidence to back up these claim ** K S * ** K SS * Windows NT 4.0 has been proven in demanding c ** S * environments to be a reliable operating system. ** SS S * such as Barnes and Noble, The Boeing Company, Chi ** SS S * Stock Exchange, Dell Computer, Nasdaq and many othe ** SS * mission-critical applications on Windows NT 4.0. ** * Linux lacks a commercial quality Journaling File System ** * means that in the event of a system failure (such as a po ** outage) data loss or corruption is possible. In any event, e syst must check the integrity of the file system during sys ** estart, a process that will likely consume an extended am * ** time, especially on large volumes and may require m * ** vention to reconstruct the file system. * BB ** ommercially proven clustering technologies to * B BB ** ilability for Linux. The Linux community may * B B ** projects and small companies that are * BB B ** gh Availability functionality. D. H. Brown * B BB B OO ** ese offerings remain immature and * B BB O OO ** demanding business world. * BB B O O ** vide uptime guarantees for Linux, ** BB B O O ** aq, Data General, ** BB O O L ** ys provide 99.9 percent ** O O L ** or Windows NT-based servers. ** O O L ** Myth: Li ** OO O L ** ** OO L L ** Reality: Fre ** L L ** w Total Cost of Ownership ** LL L ** ** LL L ** The Linux communit ** LL L OO ** ost nature of Linux. It's important to un ** L O OO ** y a small part of the overall decision-makin ** LL O O ** ** LL O O ** The cost of the ope ** LL O O CC ** ntage of the overall total cos ** O O C CC ** ndows NT has proven to have a ** O O C C ** Previous studies have sho ** OO O C ** lower TCO than UNIX. There ** OO C KK ** Linux is significantly differ ** C K ** when it comes to TCO. ** C K KKKK ** The very definition of Linux as a ** CC KKKKK ** that commercial companies like Red ** CC K K SS ** charging for services. Therefore, com ** K K S SS ** for Linux will be fee-based and will li ** K K S SS * premium. These costs have to be factored ** K S * model. ** K SS * Linux is a UNIX-like operating system and is ** S * to configure and manage. Existing UNIX users ma ** SS S * transition to Linux easier but administrators for ** SS S * Windows®-based or Novell environments will find it ** SS * difficult to handle the complexity of Linux. This ret ** * l add significant costs to Linux deployments. ** * Linux is a higher risk option than Windows NT. For exampl ** how many certified engineers are there for Linux? How easy it t find skilled development and support people for Linux? Who ** orms end-to-end testing for Linux-based solutions? Th * ** ors and more need to be taken into account when c * ** latform for your business. * BB ** Myth: * B BB ** ure than Windows NT * B B ** Real * BB B ** el Is Weak * B BB B OO ** Al * B BB O OO ** ecurity issues, however it's important to note t * BB B O O ** model as the original UNIX i ** BB B O O ** ot designed from the ground up to be ** BB O O L ** ** O O L ** ** O O L ** s for files and directories. In c ** OO O L ** , from files to operating sys ** OO L L ** control list and its use can b ** L L ** Linux s ** LL L ** rators cannot delegate administr ** LL L ** any administrative capability ** LL L OO ** which compromises b ** L O OO ** , Windows NT allows an admin ** LL O O ** an exceptionally fin ** LL O O ** Linux has not suppo ** LL O O CC ** ndards. Every member of the W ** O O C CC ** 3.5 has been evaluated ** O O C C ** Government's evaluation p ** OO O C ** under the British Governmen ** OO C KK ** Linux products are listed on ** C K ** product list. ** C K KKKK ** Linux system administrators must ** CC KKKKK ** understanding the latest Linux bugs ** CC K K SS ** do about them. This is made complex d ** K K S SS ** there isn't a central location for secu ** K K S SS * and fixed. In contrast Microsoft provides ** K S * repository for notification and fixes of se ** K SS * Configuring Linux security requires an admini ** S * expert in the intricacies of the operating syst ** SS S * components interact. Misconfigure any part of the ** SS S * system and the system could be vulnerable to attack ** SS * NT security is easy to set up and administer with too ** * the Security Configuration Editor. ** * ** Myth: Linux can replace Windows on the desktop Reality: L ** Makes No Sense at the Desktop * ** Linux as * ** operating system makes no sense. A user would end up with a * BB ** s fewer applications, is more complex to use and manage * B BB ** uitive. * B B ** * BB B ** de support for the broad range of * B BB B OO ** Windows NT 4.0 currently supports over * B BB O OO ** es on the Hardware Compatibility * BB B O O ** rt important ease-of-use ** BB B O O ** d Play, USB, and Power ** BB O O L ** ** O O L ** ating system and ** O O L ** GUIs would make r ** OO O L ** ing and would add sig ** OO L L ** Linux ** L L ** d, meaning that custome ** LL L ** horizontal and vertical ** LL L ** Forrester Research highlighted ** LL L OO ** enterprise ISVs develop appli ** L O OO ** 13 percent develop for Lin ** LL O O ** ** LL O O ** Summary ** LL O O CC ** The Linux operating system i ** O O C CC ** by business or home users. Today ** O O C C ** be confident in delivering applicat ** OO O C ** reliable--yet cost effective to de ** OO C KK ** a long way to go to be competitive with Win ** C K ** e Windows 2000 operating system, Microso ** C K KKKK ** superiority of the platform even further ** CC KKKKK ** r the next generation applications to solve ** CC K K SS ** ** K K S SS ** More information ** K K S SS * ** K S * Customer Testimonials ** K SS * See how these leading companies and organizations ha ** S * Windows NT Server 4.0: ** SS S * ** SS S * Nasdaq ** SS * Barnes & Noble ** * Dell Computer Corp ** * The Boeing Company ** Chicago Stock Exchange Gartner Gr ** eports New repor * ** Gartner raise important questions about the future role of Linux. * ** * BB ** * B BB ** able Competition for Windows Desktops? * B B ** * BB B ** w Linux as a serious competitor for * B BB B OO ** op, Linux will not disappear from the * B BB O OO ** ugh 2004." * BB B O O ** x Face-Off ** BB B O O ** tant niche roles, it will not gain ** BB O O L ** te for Unix and Windows in the ** O O L ** ** O O L ** " ** OO O L ** success in the Linux mar ** OO L L ** ot a foregone conclusion, despi ** L L ** ** LL L ** Performance Data ** LL L ** See Industry Bench ** LL L OO ** tperforms Linux ** L O OO ** ** LL O O ** Footnotes ** LL O O ** 1. Siemens & SuSE announce ** LL O O CC ** nd to 4 GB, although this is not p ** O O C CC ** ributions. ** O O C C ** 2. Linux: How Good Is It? D. H. ** OO O C ** ** OO C KK ** 3. Forrester Research, Software Vend ** C K ** 1999 ** C K KKKK ** |--------------------------------------------------------------------| | This file has been borrocksed by Borrocks 95. | | (c) 1999 Scott A. Belmonte, Lachlan M. D. Cranswick | | http://www.ccp14.ac.uk/people/lachlan/borrocks/ | |--------------------------------------------------------------------|