Term
| What are the relationships among AIX, SCO, Novell & IBM? |
|
Definition
•IBM licensed source code for System V Unix in 1986 and called their operating system AIX.
•SCO licensed source code for System V Unix in 1989, from this they developed SCO Unix and later called it OpenServer.
•Novell developed a distribution called UnixWare in a partnership with AT&T's Unix System Laboratories in 1991. Novell acquired USL and became the owners of the UNIX source code and trademark.
•SCO acquired the UnixWare product from Novell, but did not acquire the ownership rights of the UNIX source code or the trademark in 1995.
•SCO was acquired by Caldera in 2001 and changed its name to the SCO Group.
•The SCO Group sued IBM (Linux distribution) in 2003 claiming that Linux used some UNIX code illicitly. The lawsuit still continues today.
•The SCO Group sued Novell over the ownership of the UNIX source code and trademarks in 2004. It was judged that Novell is the owner of both properties in 2011.
•The SCO Group is now bankrupt. |
|
|
Term
| What are three traditional, proprietary Unix manufacturers? |
|
Definition
•SCO, IBM (AIX) and HP (HP-UX). Univel
•ESIX
•Micro Station Technology
•Microport |
|
|
Term
| What's the difference between MySQL and Oracle? |
|
Definition
MySQL: Open Source, Mainly web-based apps
Not scalable, Does not support distributed databases, Does not support save points, Does not allow programs within database
Suited for small to medium scale,Allows programs within the database, If you put through too much information, may break and may need to be repaired, 32bit, Most used but only for small applications
Oracle: Commercial—can only be use with the proper paid license, Backend database, Scalable, More powerful, Supports distributed databases, Supports save points,Suited for enterprise deployments, Cannot outgrow, One of the biggest databases along with IBM IMS,64 bit
Similarities: Relational database management system, SQL language, Oracle owns both |
|
|
Term
| How do companies like Red Hat & Suse charge money for their distributions of free software? |
|
Definition
| A subscription includes more than just the software that people can get for free. It includes integration and certification and training and updates. It includes early access to new technology, which often includes access to the developers who are still building it. It also includes guaranteed support turnaround times. It’s the time that is spent dealing with the actual deployment that people pay for. |
|
|
Term
| What are 'fault tolerant' or 'highly available' computers? How do they differ from a stand-alone server-class machine? |
|
Definition
| Fault tolerant or highly available computers are a two or more servers set up on two separate locations (redundant parts). In the event that one server fails, another will take over. In highly available systems, this will be instantaneous or nearly so, so that from a user perspective there is no degradation in service or availability of the application. When a stand-alone server fails, it must be taken offline. |
|
|
Term
| How can inexpensive, relatively unreliable server-class machines be used to make a fault tolerant environment? |
|
Definition
| Unreliable server-class machines can be used to make a fault tolerant environment by setting up two servers at two separate locations. If one fails or goes offline or needs to be repaired, you can repair the problematic server without taking the system offline. The other sever will simply take over. Clustering software or server farm software that can detect when a machine goes out of service and it detects it automatically and activates the other server and reboots the problematic server. You make the unreliable hardware reliable by running a reliable software. Less expensive more effective approach rather than pouring money into large machine that are fault tolerant. I.e. Google |
|
|
Term
| Considering server farms, mid-range or mainframes, what is required to provide as close to 100% availability of computing resource as is possible? |
|
Definition
Having systems in more than one geographical location will provide as close to 100% availability of computing resource. N+1 or N+2 redundancy. In mainframes they pretty much always have an extra one (sometimes an extra two) of everything. All the way down (fractally). Just in case. And if you're running a high stakes production application there is often an entire backup machine that syncs with the main ready to go.
One other interesting thing that came to mind is that many large systems also have a flywheel setup with them. Basically a small amount of energy generates and maintains a high speed wheel as insurance. If the main power fails, it might take a little while to get power from a generator. So the flywheel can act as a generator with its existing momentum for about 15-90 seconds depending on the model until a more permanent supply is working. Server that picks up from the last server is called a ‘hotsite’ and has the most recent copy of data. |
|
|
Term
| Where did UNIX come from and when? Where did Windows come from and when? Windows 3.11 for workgroups and Windows NT marked Microsoft’s entry into the server market place—when did that happen? |
|
Definition
UNIX originated in Bell Labs in 1969, developed by Ken Thompson and Dennis Ritchie as a multi-user time-sharing operating system. The name for the new operating system was actually coined in 1970 as a pun, playing on the name "MULTICS", the name of the complicated experimental operating system that it came to replace.
Windows was developed by the Microsoft Corporation in 1985 as a graphical shell for their Microsoft Disk Operating System (MS-DOS). Later developments would turn it into it's own operating system
Dave Cutler , an OS Architect, was already famous. He designed several Operating Systems while working at DEC (Digital Equipment Corp). This included RSX-11M and vms.
He is the principal architect behind Windows NT. The basic design persists to this day.
It is interesting to note that Windows NT has nothing to do with earlier versions of Windows. The NT stood for "New Technology". Serveral emulation layers were provided to interface with the NT kernel. One of those emulation layers was the Windows Interface. Another was Posix (a Unix like interface) There was also a clean separation between these emulation layers and the NT kernel. This was later abandoned in favor of promoting the Windowssubsystem at the expense of the other two subsystems. (The third subsystem was OS/2). The Windows subsystem had hooks directly into the kernel.
NT and 3.11 for workgroups gave Microsoft that client-server relationship. All you had to do is connect the computers to a hub, and you would have a network. |
|
|
Term
| How are machine language and assembly languages similar and how are they different? |
|
Definition
Machine code is the only form of program instructions that the computer hardware can understand and execute directly. All other forms of computer language must be translated into machine code in order to be executed by the hardware. Machine code consists of many strings of binary digits that are easy for the computer to interpret, but tedious for human beings to interpret. Machine code is different for each type of computer. A program in machine code for an Intel x86-based PC will not run on an IBM mainframe computer, and vice versa.
Assembly language is a symbolic representation of machine code, which allows programmers to write programs in machine code without having to deal with the long binary strings. For example, the machine code for an instruction that adds two numbers might be 01101110, but in assembly language, this can be represented by the symbol ADD. A simple assembler program translates this symbolic language directly into machine code. Because machine code is specific to each type of computer hardware, assembly languages are also specific to each type of computer. However, all machine languages and assembly languages look very similar, even though they are not interchangeable.
High-level language is a language that is convenient for human beings to understand. High-level programming languages must be translated into machine code for execution, and this process is called compilation. A program that carries out this translation is a compiler. High-level language may bear no resemblance at all to machine code. The compiler figures out how to generate machine language that does exactly what the high-level-language source program specifies. Languages like C++, Algol, COBOL, etc., are all compiled high-level languages. They usually work more or less the same across all computer types, which makes them much more portable than assembly language.
Some high-level languages are interpreted rather than compiled. This means that they are not translated into machine code. Instead, when the program is executed, another program, called an interpreter, looks at the source code and does what it says, without actually translating it into machine language. Visual Basic, Java, Javascript, BASIC, and many other languages work this way. It's more flexible, but it is far slower than a compiled language.
Both reference computers instruction set.
They are different because assembly is an abbreviation for machine code which is binary.
Three different assembler languages are 8286, 8088, Power System Assembler; every CPU has its own assembly language. Usually backwards compatible. Intel 8086 is backwards compatible with IBM. |
|
|
Term
| Assembly vs. Machine Language |
|
Definition
Assembly: Describe a data type
Describe a code address or function with a label.
(x86 specific) Ignore some segment registers which can be implicitly determined from a data decelaration.
Write macros |
|
|
