• Introduction
• Exam Breakdown
• The Planning Section
• The Installation and Configuration Section
• The Managing Resources Section
• New Examination Format

Introduction

I intend to start this year’s MCSE Help guide with an indepth look at examination no. 70-067 – Implementing and Supporting Microsoft NT Server 4.0. The focus will be on how to pass the examination, explaining exam content, types of questions to look out for and the vital knowledge that you need to gain that all-important “Pass” mark.

As demonstrated when we looked at the NT Workstation exam, there’s not much difference between NT Server and NT Workstation, so some of what we will be looking at in this article and the next will be familiar to you if you have already revised for your Workstation exam. This is why it is a good idea to revise for both exams simultaneously. If you are really confident, you could even sit one exam straight after the other. The pass mark for the NT Server exam is 764. The different sections in the Server exam are laid out for you below, along with hints and tips for what to look out for in each section.

Exam Breakdown

The Server exam has six sections: Planning, Installation and Configuration, Managing Resources, (which we will consider this month), Connectivity, Monitoring and Optimisation and finally Troubleshooting, (which we will look at next month).

The Planning Section

What to revise:

When to use NTFS and when to use FAT volumes, Which protocols can be used over RAS, Fault tolerance strategies, Which network protocols to install, Types of drives that can include the system or boot partition, Which connectivity types will PPP support, When DNS, WINS and DHCP servers are required, When to implement a system policy, Performance benefit of stripe set (no parity).

Watch out for:

1. Which network protocols to install.

You must be aware of the following: To connect to UNIX or the Internet use TCP/IP, to connect to NetWare use NWLink, to connect to Macintosh use AppleTalk and to connect just to other Microsoft computers use NetBEUI. Remember however, NetBEUI is not routable and so limits traffic to the local subnet only.

2. Which upper layer network connectivity types will PPP support?

PPP will work with any network connectivity type, including Windows Sockets, NetBIOS, RPCs and Named Pipes. Effectively its job is to disguise the differences between a computer connected directly to the network and a computer connected to the network via a dial-up link. Its activities are not shown to the upper layers and so are not considerations in selecting which connectivity types to use.

3. Which protocols can be used over RAS?

A very similar question to the above, but looking now at the transport protocols. NT RAS supports the following three transport protocols using PPP: NetBEUI, TCP/IP and NWLink. NT RAS client can optionally use SLIP instead of PPP, in which case your choice would be limited to TCP/IP only.

4. Which types of drives can include the system or boot partition?

Only standard volumes or mirror/duplex sets can include the system or boot partition. Volume sets, stripe sets or stripe sets with parity can not.

5. Performance benefit of stripe set (no parity)

A small point, but one worth getting right in the examination. Stripe sets without parity are actually faster performing than stripe sets with parity. This is due to the fact that no parity information needs to be calculated. Stripe sets have faster read access than other types of volumes, due to the way that data is written in 64k stripes to each section in turn, thus involving all physical disks in the process of retrieving the data afterwards.

6. Fault tolerance strategies

This issue is by far the most important one to familiarise yourself with for the Planning section of the NT Server exam. First of all remember that NT Workstation does not support fault tolerant volumes, so before you answer a fault tolerance question be certain as to whether the question is referring to the Server or the Workstation version of NT – Microsoft likes to catch you out on this one!

A Stripe Set with Parity corresponds to RAID level 5. The word parity refers to a mathematical calculation performed on each data stripe and stored alongside the data itself within the stripe. If a hard disk fails, the data lost can be rebuilt by recalculating the parity information. In order to create a stripe set with parity you must use equal-size areas of free space on between 3 and 32 physical disks. Stripe sets with parity are less efficient in terms of speed and storage than stripe sets without parity. This is due to the overhead of calculating the parity information (speed) and storing it (storage). Parity information will take one disk area’s worth of space, so if you were using five disks, one fifth of space would be used for storing parity information and therefore only four fifths used for storing data.

A Mirror Set corresponds to RAID level 1, as does a Duplex Set. With a mirror set, all data is written twice, firstly to the original volume and then subsequently to a second volume that is a mirror image of the original and stored on a separate physical disk. If the first physical disk fails, NT data will still be accessible from the mirrored version on the second disk. In order to create a mirror set, you must use an existing volume on the first disk and an area of free space on the second disk that is of equal size to the existing volume. A duplex set is the same as a mirror set but with the added bonus of placing the second physical disk on a different hard disk controller. In this way, we have fault tolerance if either a disk or a controller fails.

The Installation and Configuration Section

What to revise:

The AppleTalk and DLC protocols
WINNTEXE Options
Converting FAT to NTFS
Modifying the PATH statement
Threads and processes
The default shares
Using Disk Administrator
TCP/IP parameters to configure
Updating printer drivers
Modifying the boot delay time
The ARC Naming Convention
Printing from DOS Applications
Configuring the page file
REGEDIT & REGEDT 32
NTFS permissions and file compression
Installing a tape backup unit
Configuring UPS
Printer properties
Using Server Manager
Working with Network Client
Administrator
Control Panel utilities
Backing up the registry
Minimum hardware requirements

Watch for:

1. Installation Methods

Know the difference between WINNT and WINNT32. Be familiar with the /B and /OX switches.

2. AppleTalk and DLC

Remember that AppleTalk is installed when you install Services for Macintosh, not from the Protocols tab in Network

3.Server Manager

Server Manager has a multitude of uses: it creates accounts for computers, promotes BDCs, demotes the PDC (in the case of two PDCs only), synchronises the domain, remotely start and stop services, remotely create, edit and delete shares, view share and connected user information, set administrative alerts and configure replication.

4. Working with Network Client Administrator

This utility allows for several different disk sets to be created, so make sure that you are familiar with the process involved for each one. With Network Client Administrator, you have four options: Make Network Client Installation Startup Disk, Make Installation Disk Set, Copy Client-based Network Administration Tools and View Remoteboot Client Information. The Network Installation Startup Disk allows you to automatically connect to a server and install Windows NT Workstation, Windows 95 or Microsoft Network Client for DOS. To use it, you must specify a path to the shared source files and provide a floppy disk formatted as a DOS system disk. If you choose Make Installation Disk Set, you can create the following: Microsoft Network Client 3.0 for MS-DOS, Microsoft LAN Manager 2.2c for MS-DOS or Microsoft LAN Manager 2.2c for OS/2. The Network Administration Tools come for either Windows NT Workstation or Windows 95. Make sure you know which tools you use for each platform:

5. Updating Printer Drivers

An important point: To update all computers that print to a shared printer with a new printer driver, you only need to install the new driver to the printer server. Clients will automatically download the new driver when they next connect to the server to print.

6. The ARC Naming Convention

I know it’s a pain, but it’s a useful thing to learn for troubleshooting and you’ll certainly get at least one ARC name crop up in the Server exam. The ARC name refers to a volume on a physical drive and this is specified as follows: multi –or scsi – (0) disk (0) rdisk (0) partition (0)\<path>. Each section is described below:

The Managing Resources Section.

What to revise:

Managing User Accounts
Default rights for built-in groups
Changing Groups Membership
User profiles
Creating a server based working directory for each user
Gateway Services for NetWare
Promoting BDCs to PDC status
The Directory Replication Service
Differences between Local and Global Groups
NTFS and Share permissions
The Behaviour of an Expired Account
Copying user profiles to multiple users

Watch for:

1. Managing User Accounts

You get one or two scenario-based questions here, so be comfortable with what happens when you delete and recreate an account and how to use template accounts.

2. User Profiles

Be familiar with the differences between local, roaming and mandatory profiles. Remember with roaming profiles, if a user logs on to a computer that he/she has logged on to before and the server profile is unavailable, the user will be able to use the locally saved version of the profile.

3. Default Rights for Built-in groups

You’ll get at least two questions on this point so make sure you know what rights each built-in group has by default, as well as which groups are different for domain controllers. In particular remember that backup operators only have backup and restore rights when logged on locally.

4. The Directory Replication Service

Remember that you can use NT Server as an export computer and NT Server, NT Workstation and Windows 95 as import computers. Remember that system policies (NTCONFIG.SYS) and logon scripts must be replicated to all domain controllers.

5. The Behaviour of an Expired Account

What happens when an account expires depends upon a setting in your Account Policy. If Forcibly Disconnect Remote User When Logon Hours Expire is selected, the user will be disconnected as soon as the account expires. Otherwise, the user will be able to continue working but not be able to logon again once they have logged off. If an account has expired, select a new date in the future or select that the account is not to expire in the Account Information dialog box.

6. Changing Group Membership

Remember that the user will not be able to access a resource if he/she is a member of any group with specific No Access given. Remember that changes to group membership only occur when the user logs off and on again (when their access token is rebuilt).

7. Gateway Services for NetWare

To use GSNW, create a group called NTGATEWAY on the NetWare server and assign it supervisor permissions. Add the user that you are going to specify on the NT server’s GSNW utility to the group.