DEVTOME.COM HOSTING COSTS HAVE BEGUN TO EXCEED 115$ MONTHLY. THE ADMINISTRATION IS NO LONGER ABLE TO HANDLE THE COST WITHOUT ASSISTANCE DUE TO THE RISING COST. THIS HAS BEEN OCCURRING FOR ALMOST A YEAR, BUT WE HAVE BEEN HANDLING IT FROM OUR OWN POCKETS. HOWEVER, WITH LITERALLY NO DONATIONS FOR THE PAST 2+ YEARS IT HAS DEPLETED THE BUDGET IN SHORT ORDER WITH THE INCREASE IN ACTIVITY ON THE SITE IN THE PAST 6 MONTHS. OUR CPU USAGE HAS BECOME TOO HIGH TO REMAIN ON A REASONABLE COSTING PLAN THAT WE COULD MAINTAIN. IF YOU WOULD LIKE TO SUPPORT THE DEVTOME PROJECT AND KEEP THE SITE UP/ALIVE PLEASE DONATE (EVEN IF ITS A SATOSHI) TO OUR DEVCOIN 1M4PCuMXvpWX6LHPkBEf3LJ2z1boZv4EQa OR OUR BTC WALLET 16eqEcqfw4zHUh2znvMcmRzGVwCn7CJLxR TO ALLOW US TO AFFORD THE HOSTING.

THE DEVCOIN AND DEVTOME PROJECTS ARE BOTH VERY IMPORTANT TO THE COMMUNITY. PLEASE CONTRIBUTE TO ITS FURTHER SUCCESS FOR ANOTHER 5 OR MORE YEARS!

Beyond 2014 Classic Computing - Microsoft Windows NT Server 4

Windows NT 4 Server Introduction

Welcome to this written presentation about classic computing in 2014 and beyond into the future. At this time we'll be getting many new ideas, and begin considering how older era software works. The focus of this topic is Windows NT 4 Server.

Around the early 1990's desktop and server pc operating systems did not typically have native support for networking. When Microsoft desktop operating system (DOS) was getting popular there was a large demand for clients that could be loaded and configured in DOS to connect to networks since DOS did not originally come with a built in networking client. Software companies that manufactured network operating systems had proprietary client’s specific to an operating system that could be used to communicate with the network server that the client was designed for.

An early popular client for desktop network connectivity was Novell. Various versions of the Novell network operating system (NOS) were especially popular before Microsoft came out with Windows NT Server. For example in Novell 3.x, a user with a PC would use a configuration file to load a Novell network client to facilitate communication with the Novell server. The Novell server was usually a network file server of some kind, or it was also often used for print services.

In general, Microsoft saw an opportunity to create its own server operating system similar to what Novell was doing for DOS and Windows 3.x. Microsoft released one of its first versions of a network server with Microsoft Windows NT 3.x server. It was rather nice to have all the software from one vendor and the server interface was a (GUI) graphical user interface that looked almost just like standard Windows 3.x desktop OS.

The close similarity of Microsoft Windows NT 3.x server and Windows 3.x desktop gave an edge to Microsoft because many companies liked the integration and dealing with one software vendor if possible. It was also easy to use the new Microsoft NT server network operating system because it was almost just like the desktop Windows 3.x with added services, tools, and features, so it was easy to learn. With Microsoft having both a desktop and server operating system combination, Novell quickly became overshadowed and Windows NT Server made itself strongly known in the network operating system market.

The Microsoft standard created a dual software combination. There was both a desktop and a server operating system in the Microsoft NOS line. For example, there was Windows 3.x desktop and Windows 3.x server. Both almost looked exactly a like except one was the desktop version of the software and one was the server version.

We can also see this in 2013 with a similarity between Windows 7/8 desktop operating system and Windows server 2012. Both look and feel rather similar in many ways but each is optimized to run on certain types of desktop and server systems. Also another few similar examples of matched pairs are Windows Vista and Windows server 2008/ Windows XP and Windows server 2003/ Windows 2000 Professional and Windows server 2000.

Novell in the 1990’s had a market edge with their Novell Directory System – NDS. NDS organized network resources in a hierarchy that distributed rights and permissions to network objects. This advanced method of organizing network resources was present in Novell 4.x NOS.

Microsoft was using a specialized method of proprietary domains to organize network resources in Windows NT. This would be phased out later in favor of their own implementation of an object oriented network system in Windows 2000 server using Microsoft Active Directory. Microsoft Windows NT was still very popular because it was less complex to manage than an Active Directory implementation.

Some servers in modern times are still Microsoft NT based. They may have been running important applications that were still being used. Also NT servers could be joined into the Active Directory and participate on the network in that way. Often older hardware ran great on NT and if drivers were available or used generically, newer hardware could run NT and be rather fast.

NT hardware resource requirements were designed with the lower CPU and memory resources on legacy servers in mind. Newer server hardware made NT fast because it did not have the overhead those more modern versions of Microsoft server software require. Keep in mind NT is an early implementation of Microsoft server and although it runs fast on some server systems it is phased out now and no longer supported. NT is also 16bit software architecture and does not run newer 32bit or 64bit software natively.

A few networks still have early legacy server programs they run on Windows NT. Most of these run the last NT version called NT 4. NT has been around for quite awhile now, and as the saying goes what’s old is new and what’s new is old. Some network administrators like NT because it is “old school” and not as vulnerable to modern hacks in some ways. Like any server though one still needs to lock it up good and back up important data.

Installation recommendations and advice

Consider the hardware specifications of Windows NT 4 server. Analyzing the hardware requirements helps to understand what is necessary as a baseline to get it running. Since it was designed of older legacy server hardware it may not run on some modern hardware used today. Drivers are the biggest issue on modern server hardware. Often the hardware used will run on preinstalled generic driver software or a compatible emulation already present in the NT operating system. The general hardware minimum requirements are: - CPU: Intel 486, Pentium or better. - Display: VGA or better. - Hard Drive: 124 MB free, 200+ or better. - Floppy Disk: 3.5” or 5.25”, or a bootable CD - CD ROM: None or 8x or better - Network adaptor: One or more. - Memory: 16MB, or 64MB + - Pointing device: generic mouse or other pointing device.

Using hardware with anything less than these specifications will not work right. As always, more resources are best. These are not flexible specifications and anything below them will probably lead to errors. Keep in mind USB will basically not be supported and an expansion card with older connector types may be needed on some newer hardware.

Checking a hardware compatibility list may be helpful but might be out of date, possibly last updated around 1996. Check out the available directories on the software disks. Depending on the NT 4 edition it might include tools, batch files, or other documentation that might be helpful. Evaluation versions of NT 4 are around but keep in mind they might expire and leave the software disabled.

Many experts installed Microsoft DOS first and got that stably running before trying to install NT 4. A possible configuration might include Microsoft DOS 6.x. If finding bugs in an install it is easy to use DOS to check out the files and directories. NT 4 can be difficult to install, and if not an advanced user someone might use a NT 4 manual or guide book for additional details and information.

Here are the basic steps for installing to consider:

1. Boot machine and prepare to access NT 4 installation files and directories.

2. Run the specific installation program for the operating system used. For DOS, Windows 3.x, or Windows 95 the file is WINNT.EXE. WINNT32.EXE is used for Windows NT machines using Workstation or Server 3.5 or higher. The installation program copies files and directories from the media to the local drive.

3. Ntdetect is built into the program and will help the installation program to detect and recognize the hardware onboard. Some devices and hardware may not be detected and might require additional device drivers to be used in the configuration. If the devices are not detected they will probably be ignored or produce an error message. Hardware or devices not needed to get the server up and running usually can be added later. Depending on if the installation is run in Windows or DOS the screen information will be different.

4. The installation will ask for a target to install the program. Select the partition. The partition used can be FAT or NTFS. Keep in mind there are many different versions of FAT and NTFS so the versions used will work best if they are of the proper type. Using too new of a files system version will probably lead to NT not understanding the way it works or lead to errors.

5. It is generally good to format the partition and create the file system with the NT installation program. Keep in mind to use caution because formatting or creating partitions will affect other partitions or file systems, possibly leading to lost or corrupted data. A good way to install is to use a test machine dedicated to only NT. Mixing partitions and file systems is possible but better to not try this and use only NT. After everything is installed a good configuration can be found and tested. Do not use a machine that is important for an install. As mentioned before NT is complicated and best done on a test machine so that if an error occurs it will not be important.

6. Select the default Windows NT directory \winnt. Most of the steps in the installation program are best left to accepting the defaults.

7. Enter name and organization or enter a simple character like a period if not important.

8. Select a licensing mode. For a test machine it will probably be “per server”. Per seat allows users to connect to many servers. Per server is usually ok. Choose carefully because changing per server to per seat is permitted only once.

9. Enter a computer name.

10. Set a server mode, security role. PDC- primary domain controller, BDC- backup domain controller or stand alone server. In more modern versions of Windows server there is no BDC.

11. Set the Administrator password when prompted.

12. Create the emergency repair disk. Consider doing this later in NT after it is running so the machine is configured as much as possible correctly.

13. Choose to install components. Each one will add to the resources used. Additional components can be added later. Components like games, accessibility options, and multimedia can be chosen.

14. Choose the networking role. Wired to the network or remote access to the network.

15. Add IIS, Internet information server. This is a very basic early version of this program but it is good to use for research or testing.

16. Network adaptor detection. Select the protocols. TCP/IP, NWLINK IPX/SPX, and NetBeui. NWLINK IPX/SPX is used for Novell connectivity.

17. Add network services like RPC, Netbios, Workstation, Server, etc. Services can also be added later also when NT is up and running.

18. Manual adjustment may be required for IRQ, DMA, etc. No plug and play, just NT detection. These can be adjusted later in some configurations. 19. Adjust and supply protocol settings. These can be adjusted later in some configurations.

20. Set Computer and Domain name. This is needed if role of machine is a PDC.

21. Configure IIS if installed.

22. Set date and time. Set display properties. These can be adjusted later in some configurations.

23. Create ERD, emergency repair disk if desired at this time.

This is what will be done in install. After NT is loaded there will be more adjustments needed to tune and configure. Once the program is installed the machine can be loaded with more programs and customized.

Another consideration before installing would be using switches and options. They are not required but might be needed depending on the way the software needs to be installed. See further documentation for additional details. The Microsoft website is great to find older documents helping to explain options or use a good resource manual to review before installing.

There may be a machine available with an older version of NT than NT 4. It is possible to upgrade the machine. Keep in mind it will modify the configuration and it might be a better choice to load a fresh installation than try to change another computer to a more recent software version. Always keep backups if choosing to upgrade or preserve data for later.

The Windows NT 4 Server boot up process

The first phase is the bootstrap process. First turn on power to the server. It will then perform a power on self test or POST. After verifying with the POST, NT will load the runtime environment to bring up the operating system.

The files needed to boot up are NTLDR, BOOT.INI, NTDETECT.COM, etc. NTLDR will facilitate the loading of NT, BOOT.INI will initialize the configuration, and NTDETECT.COM will detect important necessities. Reference further information for more details on what each file does if you like.

There will be variations with different machine configurations. For example, a RISC CPU will have its own different things it does. Reference Microsoft documentation for more details.

A boot loader process will run. Hardware will be detected, and then OS loader will load NT with the OS kernel. If a blue screen loads it will then show information regarding processor, memory, build number, and NT version.

After the OS is loaded the machine will display a message to logon with CTL+ALT+DELETE. The BOOT.INI file can be adjusted in NT with a text editor to customize settings if needed. Most experts recommend adjusting the BOOT.INI file in control panel but it can be modified by most text editors. To backup the file consider changing the file extension to .OLD if you like and create a new BOOT.INI file. Use notepad or any standard compatible text editor.

Supported File Systems

Windows NT 4 server supports the FAT - file allocation table and NTFS - the new technology file system. FAT has the advantage of being versatile when different types of clients will be accessing the NT 4 server. NTFS tends to be best for Microsoft only clients and when an administrator or user needs advanced file features and permissions. NTFS also has better performance on the average.

There are many versions of FAT and each one has its plusses and minuses. There is 16bit FAT, 32bit FAT, and VFAT to name a few. Each is related to a specific version of Windows. It is possible to mix around FAT types but they are different and it is not recommended getting to complex with a configuration. Multi boot machines are possible but tend to be complex to manage. Other factors to consider are partition and volume sizes, maximum file size, and long name support. Generally 16bit FAT does not directly support long file names and will limit them to 8 characters with an extension limit of 3, this is called 8.3 file names. More advanced file systems do not have the same limits as short file names. When viewing long file names in DOS they will be truncated to 8.3.

NTFS comes with NT 4 as a file system option. It is recommended for higher levels of security and permissions. Do not confuse “old school” NTFS and modern NTFS. Modern NTFS will often vary in version depending on the version of Windows and one need’s to be aware what they have and what they need to do the job right. Keep in mind some file systems on partitions on the same disk will reach out and corrupt other nearby file systems on the same hard disk.

It is not recommended to try to use FAT32 with NT 4. Experts say for best results and file performance use NTFS in NT 4. Dual boot is possible in Windows NT 4 but usually the limitation is using 16bit FAT.

Share permissions can be used on directories of files for security. No access, read, change and full control. Depending on what groups have access the share permissions will have affect. The most restrictive permission is the effective permission. For example if no access is combined with the read permission the result is no access even though it is combined with the read permission. Test different combinations and plan what is wanted for share permissions in a network server environment.

NTFS also has its own set of directory and file permissions. Folder permissions - No access, full control, add, add & read, list, and change. NTFS permissions can get complex and need careful planning so access is not inhibited. Files have their own set of NTFS permissions, full control, no access, change, and read. Like share permissions, the NTFS permissions add together except no access overrides all others used.

Other factors to consider involve permission inheritance where child files inherit folder permissions they are contained in. Also groups have their own permissions assigned. Groups are resources in NT 4 domains that often can group users of a similar type, like administrators. Planning and rule sets are best to keep things strait and clear. It is good to try giving permissions to groups and add users to groups to control access. Assigning permissions to individual users is possible but can be hard to keep track of without groups.

HPFS is the high performance file system. Used in OS 2/x and LAN manager it was a great file system. It is similar to NTFS in regard to features. Now rather obsolete HPFS is not directly supported in Windows NT.

Methods of Windows NT 4 Server fault tolerance

Fault tolerance is a large subject and consists of many environment options and variables. In relation to Windows NT the main focus is on disk fault tolerance including disk mirroring, disk duplexing, and disk stripping. Areas to be aware of are backups, server clustering, backup power supply, and many others.

The disk administrator utility is a great way to monitor and keep track of NT server disks. It has a graphic user interface GUI like most NT utilities. One can use it to perform operations like creating partitions, creating volume sets, view file system types, and disk size & status. Be sure when making changes with the disk administrator to click commit changes now because this is mandatory to change configuration.

Most user admin tools like the disk administrator are contained in start > programs > administrative tools. In more recent Windows server versions the tools are assembled together in the Microsoft management console MMC. The MMC can be customized to snap in many utilities but the MMC is not included in Windows server NT 4.

The server hard drives are separated into partitions, and volumes are organized structures of one or more partitions that support file storage. In NT a volume set has two to thirty two partitions. External storage is possible and different manufactures sometimes include their own methods of fault tolerance for their systems. In some servers that have a large enough capacity all the hard drives are self contained within it. Disk fault tolerance can be very handy on small simple servers in Windows NT 4. Common methods supported in NT are Disk mirroring and disk duplexing. Disk mirroring is just like it sounds, an exact duplicate of one physical disk is replicated to another in case the disk crashes the other disk is ready to go.

Disk duplexing is like disk mirroring. Not like mirroring it will often use its own controller for each disk so if the hard disk controller goes bad it can keep operating. Disk mirroring uses disks on the same hard disk controller, so if the controller goes bad in a mirror both disks go down. Disk stripping uses a parity stripe written across sets of multiple hard disks. Disk striping is fast because multiple disks combine together to work. Disk striping does not have limitations like mirrors and duplexing in the way the duplication of these methods cut disk storage in half because half the total disk space is used to back itself up. Disk striping uses all disks in combination to increase storage capacity and speed. Parity can be added to disk stripping to regenerate lost information if a disk goes bad.

With disk striping with parity all partitions in a set must be of equal size or close to it. Each created partition must be on a separate hard disk. Here NTFS must be used for the stripe set with parity; FAT 16bit can not be used. Three to thirty two disks can be used with a stripe set with parity. The stripe set with parity in Windows NT 4 is considered RAID level 5. Disk striping with or without parity can not be used on Windows NT boot and system partitions.

RAID stands for redundant array of inexpensive disks. RAID level 0 is disk striping without parity and RAID level 1 is disk mirroring. RAID 0, 1, and 5 are used in Windows NT 4. There are other levels of RAID implementation but only three work with an NT 4 machine unless they are made by a third party manufacturer.

It is also good to use an uninterruptible power supply. If power fails and NT crashes, there may be errors on the hard drives that can not be easily fixed. Best to have a steady power supply ready at all times.

The Windows NT Performance Monitor

The performance monitor tool is very handy for server optimization. Performance on a server is something vitally important to be aware of. Slowdowns and poor resource access can cost time and money. Also server slowdown can be a sign of a virus, malware, or other abnormality in the network operating system.

Great performance gains can be realized when analyzing performance metrics on the server. Fine tuning and tweaking can be done to make a relatively slow server much faster once significant adjustment is made. The performance monitor is very important for many reasons and since it is built into the operating system no additional money is necessary to acquire a similar monitoring tool.

System components in Microsoft Windows NT 4 are represented by objects in performance monitor. A computer is easily selected from a GUI menu and then the object representing the component can be added to the interface chart. Counters and colors can be used for different objects then they are displayed over time on the screen. Trends relating to different component objects can reveal things like bottlenecks and underperforming devices. Once a baseline performance is known and established, adjustments improving devices or components can be done and re-measured to optimize server network performance. Server characteristics can also be observed remotely. Alerts, charts and reports can be used to enhance network analysis.

For example a memory object can be chosen and counters like available bytes and pages/ second can be displayed to give insight on a server concern. One great thing about this is it is a great way for new administrators to use a basic and easy to use tool of this type to practice real life systems support skills. Monitoring the server is a terrific idea and such specific information can be gathered quickly without effort.

To start a performance monitor chart from the start menu go to programs >administrative tools >performance monitor. From the menu select chart, on the edit menu choose add to chart. Select a computer to be monitored, and then choose objects and counters to monitor. That’s all there is too it. Save your configuration to a file if you want to easily monitor the same data again.

The best reason to use the report view in performance monitor is it’s a lot easier to read if many objects and counters are too messy to display on a chart view. Adjusting the update interval allows one to receive data updates to the display over a period of time needed to best illustrate what needs to be known.

Ways to manage resources on Windows NT 4 Server

If you have heard of Active Directory in Windows 200 server and above, then this will seem like a light version of that. Very basic network administration will be explained here from the point of view of Windows NT 4. First there are user accounts for network users to logon and authenticate into the network. Compute accounts are used for computers to allow access to the network domain. The network domain is like a controlled area for granting access to network resources.

Windows NT network domains are not the same as domains in other IT realms. The Windows NT domains are like containers for network resources and users. The domain can be used to manage the network environment. Computers with accounts can join the domain to participate on the NT network. Users can use their accounts to logon to the computers to access domain resources like file servers and printers. Groups are used to assemble users together and assign them rights to network resources. Restrictions can also be applied to groups.

In Windows NT 4 there are local groups and global groups. Local groups are available only on the local domain. Global groups are available across trusted domains in an enterprise network. In the user manager tool one can add a new user or group account. On more modern Windows server versions this tool is similar to the Active Directory users and computers tools. It is rather basic and easy to use user manager, just click on the icon to open it then navigate the menu and choose the option for the task you want to perform. For example to create a new user select the option then complete the dialog box with the information requested. Many versatile options include standards like setting a password and hours allowed to logon to the network. An account template can also be used to create users easy and quickly. A script can be created to run to perform functions like user administrative tasks. The scripting capability for administrative tasks has been more perfected in current Microsoft server versions.

In Windows NT 4 server a domain must have a domain controller or PDC - primary domain controller. Other servers can be added to the network as BDC’s or backup domain controllers also. The domain controllers help establish the network domain and secure its resources. The domain controllers work to authenticate users into the network domain. If more than one domain exists they can form trusts to allow sharing of network resources.

A stand alone member server in Windows NT 4 is neither a PDC nor BDC, it is just a server unto itself possibly to run applications or serve a network service. A PDC in NT 4 can not be demoted to a member server without reinstalling the operating system. In recent Microsoft server versions there are no BDC’s, all domain controllers are PDC’s. Some experts say Windows NT 4 server is like a Microsoft server NOS without an Active Directory.

NT has built in groups that can be used for common tasks. Some useful groups all ready to use are administrators, users, domain guests, print operators, and server operators. Keep in mind groups also have the capability to have assigned to them their own access permissions.

System policies control user environments and actions they can perform. In NT 4 the system policy editor could be used to customize user desktops, menus, and specify network settings. User policies are an easy way to apply restrictive settings or customizations across a group of users.

Similar to system policies are user profiles. Basically they work in their own way to set up a user environment. They can also be roaming and follow the user around the network or they can be made mandatory to enforce certain rules.

Like system policies and user profiles one can also use logon scripts. Some administrators not wanting to put the effort into maintaining large schemes of system policies or mandatory user profiles might use logon scripts. They seem to mostly be used when not a lot of environment options are needed, just things like mapping certain network connections and stuff.

Early ideas for replicating elements of the basic domain directory before Active Directory were evident in NT services like the directory replication service, which is now sort of obsolete because of Active Directory. Another useful tool in Windows NT 4 server was the server manager. When managing network shares it was easy to perform many everyday tasks with the server manager.

Computer networking and Windows NT 4 Server

Windows NT server comes with various network communication protocols. Some might be useful and others may never be used. NetBios, NetBeui, TCP/IP, IPX/SPX, DLC, and Appletalk are some protocols used in NT 4. NetBios is used mostly for basic, fast NT operations, and it is not routable. NetBeui is very simple and similar to NetBios. Like NetBios, NetBeui is not routable.

TCP/IP was and is still the standard routable network protocol in the world today. IPX/SPX - NWlink was a protocol used for Novell networks. In modern times it is rarely seen on Microsoft networks. TCP/IP was much more popular than IPX/SPX. Appletalk was sort of replaced by TCP/IP also. The DLC protocol was included mainly for HP printing application.

Windows NT 4 server came with the ability to run protocol routing as a service. In the early 1990’s this was seen as somewhat acceptable in performance for the time. As router hardware became faster and more advanced it became impractical to use routing software on NT server. When running routing services on Windows NT 4 server it would slow the server down considerably, and dedicated external hardware routers were the most practical solution. Also NT networks would commonly run multiple protocols. There was just too much resource overhead on the NT server to run routing services and network services at the same time, especially with several protocols at once. Software routing services are fun to research and test, but not practical in the real world.

Some of the most widely used services for Windows NT server are TCP/IP related. DHCP or dynamic host configuration protocol is used for dynamically assigning TCP/IP address on the network. The TCP/IP assigned addresses is within a range addresses in a pool. When the client requests an address the DHCP, dynamic host configuration protocol service running on the NT 4 server arranges for the TCP/IP address to be leased to the client host.

Windows NT 4 server typically needs a fixed or static address that does not change. The stable TCP/IP address used by the server provides reliability in referencing and calling its network resources. TCP/IP is the most used networking protocol.

Windows Internet Naming Service, WINS enables network names to be resolved to IP addresses. This is run on the NT server as a service to facilitate Windows naming instead of naming by TCP\IP address only. Be sure to have WINS in use on NT servers.

Serial Line Internet Protocol, SLIP was mostly used to connect to internet providers or Unix based hosts and networks. SLIP is not in use as much as it was because it had no error checking built into it. Point to Point protocol, PPP came along and began to replace connections formerly handled by the SLIP protocol. PPP has additional features compared to SLIP like error checking, encrypted logon, and additional transport protocol support.

Point to Point Tunneling Protocol, PPTP shared the benefits of PPP and combined them with creating secure connections between private networks. PPTP also facilitated this type of connectivity with good cost savings, compared to implementing this technology with other methods at the time. It also worked well in low bandwidth type connections.

File Transfer Protocol, FTP is used for fast file transfer from a local host to a TCP/IP network FTP server. Windows NT server came with the ability to run a FTP server service for file transfers. For administrators it was especially handy to have this service readily available on the server for using as a file transfer utility.

Gopher is a legacy internet service included in the NT server package. Gopher was an early method of using the internet before HTTP services were used. This protocol is good for research and testing but is not normally used anymore because it is a text based system that displays basic network links.

Hypertext Transfer Protocol, HTTP is included as a service in the NT server package. A WWW, World Wide Web protocol still in use as of 2013. It allows for the transfer of HTML documents over the internet or private intranet.

To install TCP/IP in Windows NT 4 server first open the control panel icon. Click on the network icon and choose the protocols tab. Click on the add button and select TCP/IP from the list of available protocols for network use. Windows NT might prompt you for additional files to install the protocol so be ready to supply a path needed for the installation files or files it is possibly asking for.

Close the network window and choose to obtain an IP address from a DHCP server or enter in the TCP/IP address of the computer. If a DNS server is available on the network enter its information and address. If a WINS server is available on the network also enter its information and address.

When installing the TCP/IP address of the computer manually there will be information needed to enter into the network window to complete the configuration. You need the NT server IP address uniquely assigned to the machine, a subnet mask number to identify the subnet the server will participate on, and default gateway information.

The server default gateway is important because it enables communications on the network outside the LAN, local area network. If a host or resource is not present on the local network the information request is sent to the address of the default gateway. The default gateway might be a security filter address, firewall address, router address, wireless router address, or some type of format translator. One sure thing to know is if the default gateway information is not present or incorrect the server will be isolated from network resources outside the LAN.

The subnet mask is very important also. It must be created and entered accurately into the server configuration, or else if the subnet information is not accurate the server network communications will not work because the server will be on the wrong subnet. Calculating subnets and planning TCP/IP address ranges requires an understanding of binary math and is not easy for most people to do, even systems administrators. Be sure you either receive accurate TCP/IP information from a network administrator or have the skill and understanding to create your own address scheme. Subnet calculators can be useful but if you do not understand subnets then you will not recognize calculation errors. Subnet calculation requires an understanding beyond just knowing about the TCP/IP address of the computer.

There is not much choice when using TCP/IP on servers because a static address manually entered is rather mandatory. Dynamic addresses are not commonly used on servers because if the automatically assigned address expires the server might receive a new address. On a workstation this is not normally an issue but in the server environment you do not want servers changing addresses on their own unpredictably.

In modern networks Windows servers still commonly serve DHCP information to the network, just like in the NT 4 era. Very fast routing and switching equipment like Cisco is available to perform DHCP as a service, so the DHCP service can be unloaded from the server overhead and free up server resources. In any network environment try to be free thinking about how the network will grow and develop several years into the future. For example, if a company had a Windows NT 4 server image that does not have hardware specific information they might now choose to run a NT machine in a virtual environment like VMware.

If the old NT software is left on old hardware without saving the server image for migration, when the server hardware goes bad the disk image will not work because it is tied to the hardware. Often old legacy hardware is hard to come by in an exact old server replacement. If the server image is specific and tied to the hardware it will not run on another server hardware of different specifications. The solution for now would be to disk image the Windows NT 4 server before the old hardware breaks down and use an imaging program to create a image or snapshot that can work on any hardware type and not be limited to being loaded on the same model server hardware.

That’s why today running Windows NT 4 server can be done well in a virtual hardware environment. A large number of resources can be allocated to increase performance. Also the server can be tested in a sandbox environment that keeps it isolated so if it crashes you are not worried about it affecting other network devices directly. Fault tolerance in the virtual environment makes sense too. A virtual machine running NT 4 can quickly and easily be restored and running again much faster than using a dedicated hardware server directly. Virtual environments also can greatly expand storage capacity and capability because a virtual machine running NT 4 does not need to manage local backup and storage when the virtual top level server handles all the network resources.

software | systems


QR Code
QR Code beyond_2014_classic_computing_-_microsoft_windows_nt_server_4 (generated for current page)
 

Advertise with Anonymous Ads