Service and Support

Section 1: Introduction to Service and Support

1. List and describe tasks related to providing service and support on a NetWare network
1. Configuring network hardware and software
2. Installing network hardware
3. Troubleshooting, which includes:
1. investigating whether a problem is software- or hardware-related
2. diagnosing the problem
3. fixing the problem
4. documenting the problem
4. Upgrading hardware and software
5. Network management
2. Describe the troubleshooting model
1. Gather basic information
1. determine the symptoms and who is affected
2. determine usage and activity levels
3. check network logbooks and records to determine baseline performance
2. Develop a plan
1. decide where the problem most likely comes from, and develop hypotheses
2. prioritize the hypotheses
3. Execute the plan
1. break down the first hypothesis into the smallest reasonable, testable concepts
2. test out the hypothesis by changing only one thing at a time
3. start working at the source device and move toward the destination device (forward chaining)
4. use only reliable test equipment, software, and procedures
5. use NESPro and user groups for help
4. Document the solution
1. record the nature of the problem and solution in the network logbook
2. prevent a recurrence of the problem
3. Describe good service and support techniques, and how to use diagnostic tools
1. Avoiding problems with electrostatic discharge
1. normal movements such as lifting a foot or moving a chair will generate charges of 1,000 volts
2. computer components may be destroyed by discharges as low as 20 or 30 volts
3. about 90% of the time, ESD events cause the component to degrade but not fail testing procedures
2. Benefits of an ESD control program:
1. less need for spare hardware inventory
2. less downtime
3. fewer difficult-to-trace intermittent problems
4. fewer unnecessary service visits
5. fewer disgruntled customers
3. Rules of static prevention:
1. ground yourself and equipment with a proper wrist strap and mat (not when working on monitors)
2. never touch components or ICs by their electrical leads
3. do not allow anyone to touch you when working on boards that contain ICs
4. always transport and store boards and ICs in perfect static-shielding (not anti-static) bags
5. keep nonconductors, such as plastic and Styrofoam, away from open computers and components
6. never place components on any conductive surface, such as metal
7. keep humidity at 70 to 90% (static problems more likely occur in low humidity)
4. Documenting and recording network problems
1. LAN system - documenting to help solve physical network problems
1. map - detailed graphic display of the LAN
2. LAN inventory - record of the hardware, software, and peripherals
3. cabling documentation - record of the actual cabling
4. workstation documentation - description of each workstation, its configuration, and network role
5. chronological change log - record of changes in the LAN configuration
2. history of the LAN - keeping track of old problems and baselines
1. business environment - use of the LAN and its relationship with the company
2. user information - users’ names, locations, duties, and training records
3. log of past problems and trouble reports - details over the past two years
4. baseline information and usage patterns - statistics on network traffic, CPU usage, errors, bandwidth
3. resources used with the LAN - keeping track of people resources
1. technology documentation - technical materials, protocols, routing, and LAN architecture
2. technical support - technical service contacts and bulletin board numbers
5. Using diagnostic and system information software
1. CheckIt PRO (DOS-based) and WINCheckIt:
1. determine facts about the hardware and operating system
2. inventory internal components
3. benchmark system performance
4. check IRQs, I/O addresses, memory addresses
5. edit CMOS
6. collect system information to determine software conflicts
7. use hardware tests to fully diagnose components
2. in addition, WINCheckIt uses a clean-up utility, memory tune-up utility, and un-install utility

Section 2: Using Research Tools in Troubleshooting

1. Use NESPro to find solutions to service and support problems
1. Windows startup icons:
1. Network Systems Support - NetWare products and programs (i.e., NetWare 3.12, NetWare 4.1)
2. Business Applications Support - Novell business applications and programs (i.e., WordPerfect)
3. Workgroup Applications Support - Novell workgroup applications and programs (i.e., GroupWise, UnixWare)
4. Network TIDs - technical information documents by Novell engineers and technicians regarding support calls
5. Novell Labs Bulletins - test results on Novell products, and third-party hardware and software compatibility
6. User Guide and Folio Help - training and help information on using the Folio Bound Views software
2. Types of information in each volume on the bookshelf:
1. What’s New - new items since the last update, accessed directly through a hyperlink
2. Service and Support - useful in troubleshooting
1. Technical Information Documents (TIDs)
2. Files, Patches, and Drivers
3. Novell Application Notes
4. Novell Professional Developer Bulletins
5. Novell Labs Bulletins
6. Third-Party Files
7. Training
8. Top Issues
9. Printing Decision Trees
3. File Updates - all available files, patches, drivers, and enhancements categorized by product
1. downloadable files are tagged with the disk download icon
2. place SET NSE_DOWNLOAD=x:\DOWNLOAD\ in workstation’s AUTOEXEC.BAT
4. Product Manuals - dozens of product manuals
5. Sales and Marketing - Novell Buyer’s Guide of Novell products and sales tools
6. Novell Programs - Novell Education Certification Programs, Authorized Service Centers, NPA, NUI
7. New User Info - introduction to NESPro for the new user, including Usage Tips and NESPro features
3. Using the Toolbelt (get help by <F1>)
1. Query - opens the Query dialog box and searches the infobase
1. single-word queries - searches for a single word
2. binary searches - searches for more than one word using And and Or
2. Clear Query - clears the current query
3. Next - advances the cursor to the next line of the query hit
4. Previous - returns the cursor to the previous line of the query hit
5. Backtrack - backs up through the search performed and links followed
6. Trail - displays a map of the links followed and the queries performed (can select any item and jump directly)
7. Contents - toggles between the Table of Contents hit (record) list and the document
8. Print - opens the Print dialog box
1. can save a selection as a file and bring it into a word processor
2. select Print from the File menu or click on the printer icon
4. Finding information in NSEPro
1. navigate the database headings - follow the book titles, topic headings, or icons to the information
2. perform a search - click on the Query button and enter search criteria
1. record is a portion of text that focuses on a single topic
2. view is a collection of records
2. Use NetWire on CompuServe to find solutions to service and support problems
1. Accessing NetWire on CompuServe
1. load a CompuServe browser, such as WinCIM or NovCIM
1. if new to WinCIM, download NOV1.EXE and NOV2.EXE
2. to upgrade, download only NOVELL.EXE
2. connect and log in to CompuServe
3. activate GO NETWIRE
2. Finding information in NetWire
1. navigate the NetWire Services interface:
1. New User Info
2. What’s New
3. Technical Services
4. File Updates
5. Sales and Marketing
6. Novell Programs
2. use the GO feature - select GO from the Services menu, or click on the GO icon
1. type a forum name to access and press <Enter>
3. Use NetWire on the Internet to find solutions to service and support problems
1. Accessing Novell Internet services
1. connect to the Internet
2. load TCP/IP at the workstation
3. use an Internet browser, such as Netscape or Mosaic
1. access Novell through http://www.novell.com, gopher.novell.com, or ftp.novell.com
4. Use Micro House Technical Library to find about system boards, hard disks, and network interface boards
1. Encyclopedias (updated quarterly):
1. Hard Drives - performance specifications, jumper settings, component locations
2. Main Boards - jumper settings, memory and cache configurations, connections, component locations
3. Network Cards - jumper settings, specifications
4. I/O Cards - jumper settings, configuration
2. Search features:
1. Keywords - use up to 10 words, accessed from the Key icon on the Main screen
1. keyword search, or text word search, assumes a binary AND
2. Hard Drive Parameter Search - interface type, number of cylinders
3. Main Board Criteria Search - manufacturer, speed, BIOS, slots, bus slot types, CPU
4. Network Board Criteria Search - wire type, bus type, topology, transfer rate
5. I/O Card Criteria Search - video, modem, modulation protocol types
3. Installing MTL
1. run INSTALL once to install server-based files
2. run SETUP from each workstation using MTL
1. make sure users have Read and File Scan rights to the directory for Library Control Program (MH.EXE)
5. Choose which of the four tools is most likely to provide a solution to a specific type of problem

Information	NESPro	NetWire on CompuServe	NetWire on the Internet
Patches and Fixes	Yes	Yes	Yes
TIDs	Yes	Yes	Yes
Product Information	Yes	Yes	Yes
Novell Labs	Yes	Yes	Yes
Application Notes	Yes	Yes	No
Developer Notes	Yes	Yes	No
Product Manuals	Yes	No	No
Presentations	Yes	No	No
Support Forums	No	Yes	No

Section 3: Installing and Troubleshooting Cabling and Network Boards

1. Configure, install, and troubleshoot Ethernet network boards, cables, and related hardware
1. Advantages of Ethernet:
1. inexpensive, proven technology
2. works well with a large number of LAN and micro-to-mainframe applications
3. easy to install
2. Disadvantages of Ethernet:
1. performance degradation under high loads
2. difficult to isolate problems
3. How Ethernet works
1. moves messages around the network in datagrams, self-contained packets of information (CSMA/CD)
1. packets include the source address, destination address, type of data moved, and the data itself
2. to send a datagram, a network node must first listen to see if any other node is using the cable
3. when the cable appears clear, the node sends its datagram
4. if two stations both transmit at the same time, the packets may physically collide and become damaged
2. transceiver converts the signal on the wire to a form that the workstation can use
3. if the ports on the board are physically cabled together but not active, the boards have not been configured
4. Most common 802.3 standards for Ethernet:
1. 10BASE2 - thin coaxial (similar to RG-58A/U or RG-58C/U coaxial cable)
1. limited to 30 devices and 185 m; entire network cannot exceed 925 m
1. 5 segments can be connected by 4 repeaters, and 3 segments can be populated (5-4-3 rule)
2. do not use RG-58U (does not meet the IEEE specification)
2. devices must be 0.5 m apart
3. each end of the bus must have a 50-ohm terminator, one of which should be grounded
4. transceiver is normally on the network board with 10BASE2
1. turn off the Signal Quality Error (SQE) test when using repeaters
2. 10BASE5 - thick coaxial RG-11
1. limited to 100 nodes and 500 m; LAN maximum is 2.5 km
2. each transceiver on the trunk cable must be separated by 2.5 m
1. drop cable between the transceiver and network board cannot exceed 50 m (no minimum length)
3. each end of the cable must have a terminator, one of which should be grounded (impedance of 50 ohms)
4. can connect 10BASE5 networks to 10BASE2 networks
5. do not need to cut a cable segment, if using vampire taps
3. 10BASE-T - UTP
1. cannot use shielded wiring; susceptible to EMI (proper number of twists per foot prevents crosstalk)
2. up to 512 nodes on a segment
3. up to 1,024 workstations (limited by available ports on the hubs)
4. distance between a hub and a workstation must be between 0.6 m and 100 m
5. electrical characteristics may exceed the acceptable tolerances due to environmental factors
6. 5 link segments can have up to 4 concentrators, or centralized wiring hubs (5-4 rule)
1. concentrator of the physical star topology makes the network more reliable and manageable
2. single point of failure found on bus topologies is eliminated
4. 10BASE-F - fiber-optic (new standard)
1. can greatly increase the distance and bandwidth of an Ethernet network
5. Troubleshooting Ethernet
1. make sure the network parts are physically connected
2. make sure the Ethernet and manufacturer’s specifications have been met
3. check for faulty hardware by using problem isolation to divide the network into smaller parts
4. check the connecting devices and terminators
1. round BNC connector is for coaxial cable
2. square RJ-45 connector is for UTP
3. D-shaped DIX connector is for thick Ethernet or a UTP external transceiver
5. check for improperly removed T-connectors
6. check terminators for the proper resistance (25 ohms)
7. use a cable continuity testing device or COMCHECK to make sure the cable is intact
8. check for a bad transceiver if only one workstation has difficulty connecting to the network
9. configure server and workstations for correct Ethernet frame types (message "FILESERVER NOT FOUND")
10. use the diagnostics diskette to check on a problem board
11. check the jumper or port setting
12. clean the connector fingers on a problem network board
13. test for resource conflicts by removing all boards except the network board and adding one at a time
14. watch for common interrupt conflicts
2. Configure, install, and troubleshoot token ring network boards, cables, and related hardware
1. Advantages of token ring:
1. excellent throughput under high-load conditions (16 Mbps)
2. facilitates the LAN-to-mainframe connections (star is most common configuration)
3. built-in troubleshooting mechanisms, such as beaconing and autoreconfiguration
1. when a station does not receive a message from its nearest active neighbor, it sends out a beacon
2. network board may attempt to fix the problem by performing internal diagnostics (autoreconfiguration)
4. available using UTP
5. fault tolerance through automatic ring configuration (ring-wrap)
2. Disadvantages of token ring:
1. relatively expensive
2. considerable expertise to manage well
3. How token ring works
1. source station wanting to transmit waits for a free token (3-byte MAC frame) to pass by
1. source station takes the free token, adds data, and sends the token out on the ring
2. as the token is passed to each active station, each station checks the token’s destination address
3. intended recipient copies the data from the frame, and reverses two bits
2. frame continues around the ring until it reaches the source station
3. if "TOKEN-DOS-207:Installation Error=27 Ring Beaconing," token ring boards are not running at same speed
4. if the token-ring driver fails to load, the memory manager has not excluded the range of memory addresses
4. Cabling with token ring (implementation of IEEE 802.5 standard)
1. IBM permits 260 stations on an STP ring; 802.5 standard permits 250
2. MSAU (Multi-Station Access Unit) - passive concentrators that have no power plug or internal intelligence
1. can support a maximum of 8 stations
2. Ring In (RI) and Ring Out (RO) ports connect multiple MSAUs to create a larger ring
1. relays maintain the continuity of the ring when a workstation has failed or a port is not in use
2. empty positions are automatically bypassed, and positions with cables installed will be bypassed
3. if no problems, a 5-volt (phantom) signal opens the relay for the node to become part of the ring
1. can hear an audible click as the station is inserted into the ring
2. can use a set-up or initialization tool to reset the relays
3. CAU (Controlled Access Unit) - powered concentrators that take a role in physical network management
4. patch cables - connect MSAUs to each other
1. maximum length varies with the wiring type, ring speed, and MSAU make and model
5. adapter cables - connect token ring boards in workstations to the MSAU, either directly or via a wall jack
1. maximum distance is 100 m, but varies with MSAU make and model
6. token ring cable types:
1. Type 1 - braided cables shielded around two copper twisted pairs (noisy industrial environments)
1. new IEEE specification for UTP/TR at 16 Mbps replaces the 4 Mbps specification (less attenuation)
2. Type 2 - Type 1 with four twisted pairs of telephone conductors
3. Type 3 - solid copper UTP with two twists
4. Type 5 - fiber-optic
5. Type 6 - STP with a higher loss per unit length
6. Type 8 - cable for data communications under carpeted floors
7. Type 9 - fire-retardant version of Type 6
7. Active Monitor and Standby Monitor stations attempt to deal with minor errors in the background
1. first station turned on is the Active Monitor; all other stations are Standby Monitors
2. every seven seconds, the Active Monitor sends out a frame to the next active device in the ring
1. frame requests that the device introduce itself to the next active device
2. process continues until each active device is acquainted with the device that has contacted it
3. information identifies the fault domain, the portion of the ring where a problem may occur
5. Troubleshooting token ring
1. malfunctioning token ring board may report internal errors in its custom statistics
2. token ring networks can be complicated combinations of MSAUs and their related patch and adapter cables
1. up-to-date documentation of the site’s physical layout and special cable testers help isolate problems
3. on small networks, remove all drop cable connectors, reset all the ports, and reconnect to the MSAU
4. be sure that a new address falls into the legal range and does not duplicate that of another board
5. resolve resource conflicts between boards
6. compare the network’s data rate (4 or 16 Mbps) to the board settings
7. configure parameters with the device drivers: board addresses, shared RAM locations, Early Token Release
8. mixing MSAUs from different vendors can cause problems
3. Configure, install, and troubleshoot ARCnet network boards, cables, and related hardware
1. Advantages of ARCnet:
1. inexpensive and extremely reliable (oldest LAN option)
2. easy to install and troubleshoot (called token-bus network)
3. excellent interoperability
4. variety of cables
2. Disadvantages of ARCnet:
1. slow (2.5 Mbps)
2. not designed for interconnectivity
3. How ARCnet works
1. distributed in a cluster of hubs (active or passive) on one or more buses
1. active hubs are powered and normally have 8 ports, and regenerate the signal before redistributing
2. passive hubs are not powered and simply distribute the signal to their 4 ports
2. eliminates contention for the cable by circulating a token
3. Thomas-Conrad Network System (TCNS), derivative of ACRnet, can achieve higher speeds, up to 100 Mbps
4. Cabling with ARCnet
1. up to 255 stations, with a maximum cable distance of 20,000 ft
2. maximum length between an active hub and a node is 2,000 ft (coaxial cable often used)
3. maximum length between a passive hub and a node (or active hub) is 100 ft
4. terminate unused ports on active hubs with 93-ohm resistors (optional)
1. must terminate unused ports on passive hubs
5. must use special boards and hubs with twisted-pair ARCnet cable
1. lengths can be up 400 ft with a maximum of 32 stations
6. linear bus ARCnet cable requires special boards that permit multiple nodes (maximum of 8 boards)
1. trunk cable can be up to 1,000 ft; termination is required
7. fiber-optic cable and boards can greatly increase the distance and bandwidth of an ARCnet network
5. Troubleshooting ARCnet
1. some ARCnet hardware does not work with passive hubs
2. common mistake is to accidentally connect two active hubs in a ring
3. ARCnet board I/O settings (IRQ, DMA, base address) must be the same as those set in NetWare
4. node addresses are individually set on ARCnet boards (be careful not to duplicate)
5. failure to terminate unused ports on a passive hub will cause signal reflections; test the terminator
6. LEDs are used on active hubs to monitor their status; learn about the LEDs
7. ARCnet drivers may indicate a cable disconnect in the case of constant autoreconfigurations
8. ACRnet is not compatible with SFT III
4. Describe FDDI
1. Advantages of FDDI:
1. can run at 100 Mbps, and support long distances
2. built-in network management
3. fair access to the medium and increased reliability (eliminates cable breaks and EMI)
4. can maintain ground isolation
5. cost comparable to that of UTP
2. Disadvantages of FDDI:
1. concentrators can cost over $1,000 per port; network boards are also relatively expensive
2. substantial expertise is needed to install and maintain an FDDI network
3. How FDDI works
1. one ring moves data and the other ring performs backups and other services (dual counter-rotating rings)
1. timed token rotation, rather than multiple tokens, achieves high data rates
2. all stations negotiate for level of service, or how soon they need to have possession of the token
3. FDDI uses frames of up to 9,000 symbols (symbol is 4 bits)
2. station must capture the token before it can transmit frames
1. Class A stations connect to both rings; Class B stations can connect to either ring but not both
3. stations downstream from the originating station read those messages marked with their address
1. when a fault occurs, the stations immediately before and after the problem reroute the data (wrapping)
4. FDDI concentrators function liken token ring MSAUs, but have a higher level of management intelligence
4. Cabling with FDDI
1. fiber-optic cables come in a variety of types and wavelengths (some use toxic coatings)
2. maximum of 1,000 connections with a maximum length of 200 km
3. TP-PMD (Twisted Pair-Physical Medium Dependent) is used in LAN backbones
1. much less expensive, but does not provide the security or resistance to EMI that FDDI provides
5. Troubleshooting FDDI
1. type of fiber-optic cable must match the intended distance between nodes
1. multimode fiber is acceptable up to 2 km; single-mode fiber is required above 2 km
2. can use an Optical Time Domain Reflectometer (OTDR) to test fiber-optic cables
1. to test for a complete break on a cable segment, can use a flashlight
3. loss of optical power greater than 11 decibels is serious
1. look for bad connectors, bad connections, or an open condition on the cable
2. do not look into a cable being used with a laser light source; eye damage could result
4. dirty connectors can cause problems; clean the connectors with a lint-free cloth dipped in alcohol
5. plastic fiber-optic cable is available, but is much less robust than glass cable for runs greater than 50 m
6. FDDI has an unavoidable delay factor of as much as 4 milliseconds (use Packet Burst)
7. source routing is the preferred choice when using FDDI with NetWare
5. Describe ATM
1. ATM (Asynchronous Transfer Mode)
1. packet-switching technology with rates from 51 to 622 Mbps
2. supports complex multimedia applications
3. brings switched, dedicated bandwidth directly to the desktop over existing UTP or fiber-optic cabling
4. well-suited to backbone implementations, providing scalability and seamless integration

• Demonstrate skills prerequisite to installing network boards, such as identifying the appropriate board type; setting the IRQs, memory addresses, and port addresses; setting jumpers and DIP switches; and ensuring network board and slot compatibility

1. Choosing the Correct Board (based on network topology and PC bus)
1. ISA (Industry Standard Architecture) - 16-bit
1. requires every 32-bit word to be split into two pieces for transfer across the bus
2. EISA (Enhanced Industry Standard Architecture) - 32-bit
1. theoretically downward compatible with 16- and 8-bit ISA boards
3. Micro Channel Architecture - incompatible with non-Micro Channel machines
1. may have both 16-bit and 32-bit slots (basic slot plus slot extension connector)
4. VL-Bus (VESA Local Bus) - makes use of the full speed of the newest CPUs
1. implemented by adding a second slot extension connector to a 16-bit ISA slot
2. downward compatible with both 8-bit and 16-bit ISA boards
5. PCI (Peripheral Component Interconnect) - supports 10 devices and 32-bit bus mastering
1. places the network and disk boards on separate buses for improved throughput
1. 32 bits wide, which yields a peak of 133 MB/second at 33 MHz
2. scalable to 64-bit
3. buffer between the CPU and peripherals
6. PCMCIA (Personal Computer Memory Card International Association) - useful for notebook computing
1. new architecture for connections packaged in a credit-card size
2. boards configure themselves for their host, eliminating the need for jumpers and DIP switches
2. Identifying Configuration Information
1. IRQs (Interrupt Request Channels) - hardware interrupt
1. when an interrupt is invoked, the CPU puts other work on hold and services the interrupting device
2. newer PCs have sixteen IRQs, existing in two banks of eight each
2. DMA (Direct Memory Access) channel - saves time
1. allows certain devices to write data straight to system memory without the intervention of the CPU
2. AT machines have eight DMA channels; XTs have four channels:
3. I/O address - base I/O or port address in memory reserved by the CPU (mail stop)
1. set on the board using jumpers, DIP switches, or configuration software
2. mismatch between the device and the software configuration will cause an error
4. Memory address - base memory or shared memory
1. some boards use ROM, which needs to have control of some of the address space in the main memory
2. if two resources try to use the same memory addresses, conflicts can occur
3. address range CA000 to DFFFF is usually a safe choice for configuring a network board
3. Configuring the Network Board with Jumpers and DIP Switches
1. Jumpers - tiny metal clips covered with plastic that stick out from the circuit board
1. if defective, the jumper may not complete the circuit even when properly installed over the two pins
2. DIP (DIPolar or Dual In-Line Package) switches - resemble small light switches (either ON or OFF)
1. organized in banks of two, four, or more (meaning is set by the vendor)
2. jumpers and DIP switches are often used to provide settings such as IRQs, DMA, and I/O addresses
4. Physically Installing the Board
1. Properly seating the board
1. use firm, even pressure across the entire length of the board to push it into the slot
2. do not rock the board back and forth (can damage the board and slot)
3. use the mounting screw to connect the board to the computer and ground the board
5. Configuring the Board with Software
1. CMOS - special memory chip that maintains the system configuration
1. contents can be maintained with a small lithium battery with a long life span
2. SETUP program that specifies the configuration in CMOS is often accessed by <Ctrl><Alt><Ins>
2. IBM Reference Diskette - contains a copy of the system programs for PS/2 computers
1. configuration takes place automatically when these programs run
2. programs accommodate network boards installed in the system using the Adapter Definition Files (ADFs)
3. Compaq EISA Configuration Utility - automatically configures EISA boards and resolve resource conflicts
1. can allocate system resources to ISA boards and provide instructions on how set switches and jumpers
4. COMCHECK utility - isolates physical communication problems
1. does not require a network operating system to be running
2. files available in the COMCHK.EXE compressed file on NSEPro or NetWire:
1. COMCHECK.EXE
2. COMCHECK.HLP
3. SYS$MSG.DAT
4. SYS$ERR.DAT
5. IBM$RUN.OVL
6. $RUN.OVL

Section 4: Installing and Troubleshooting Network Storage Devices

1. Describe basic hard disk principles, including the purpose of hard disks and the various interface types
1. What a hard disk does
1. provides data storage and data retrieval:
1. with total reliability
2. at the highest possible speed
3. in the smallest possible space
4. at a reasonable cost
2. Disk interface types:
1. ST-506 - originally built by Seagate for 5 MB hard disks
1. used for bigger disks via Modified Frequency Modulation (MFM) and Run Length Limited (RLL) encoding
2. Integrated Device Electronics (IDE) - uses RLL encoding, which offers increased data density over MFM
1. controller hardware is placed on the disk itself for improved performance
3. Enhanced IDE - enhancement of IDE technology that adds:
1. high capacity addressing of AT Attachment (ATA) hard disks
2. fast data transfer rates of up to 13.3 MB/second
3. dual ATA host bus adapters that allow support of up to 4 devices
4. support for non-hard disk peripherals
4. Enhanced Small Device Interface (ESDI) - more powerful version of ST-506
1. uses a bus, controller, and disk combination
2. popular for high-performance, large capacity disks (rapidly being replaced by SCSI)
5. Small Computer Systems Interface (SCSI) - high-speed parallel interface
1. allows up to 7 devices to be daisy-chained on a single cable
2. faster throughput facilitates larger disks
6. SCSI II - increased speed through reduced overhead
1. Fast SCSI-II and Wide SCSI-II are new standards that further increase speed
3. Tips for hard disks
1. hard disks are delicate; bumping or shaking can cause permanent damage (use protective packaging)
2. 95% of all SCSI hard disk problems come from improper termination or ID settings
3. adding another hard disk to an existing controller board can be extremely cost-effective
4. use ESDI with older 386 system boards that do not correctly support 16-bit memory transfers
5. SCSI controllers must renew connections with all SCSI devices before activating during power on or reset
6. route cables with care; rolling up, running alongside metal for long distances, or routing past noise generators can affect the impedance and cause signal reflections
7. most SCSI disks are shipped with the PARITY ENABLED jumper setting
8. adapter hardware allows a SCSI bus to be connected through the parallel printer port (used for tape backup)
2. Configure, install, prepare, and troubleshoot SCSI disks on a workstation and a NetWare server
1. Planning the SCSI bus
1. termination - must terminate both ends of the SCSI bus
1. active termination provides termination when required, eliminating the need to terminate physically
2. exact placement of the terminating resistors is different for internal and external disk configurations
2. SCSI addresses - HBA has a SCSI ID of 7
1. verify that SCSI IDs are set properly through hardware or software
2. if HBA and hard disk do not have different SCSI IDs, "WARNING:POSSIBLE SCSI BUS CONTENTION"
3. cabling - pay attention to what types of connectors each system component requires
1. floppy disk drives and hard disk cables look similar, but have different cable twists; do not interchange
2. Pin 1 is identified with a colored stripe; match to the receptor marked with a "1" or a square dot
3. can use 25- or 50-pin cables
4. maximum distance between connectors on a SCSI cable is 0.5 m (18 in)
5. use impedance-matched cables with external devices
6. do not roll the cable back on itself, or run the cable next to sources of EMI
2. Configuring and installing the host bus adapter (HBA)
1. set the proper SCSI address
2. set the proper termination
3. configure the controller interrupt, base memory, base I/O address, and DMA channel
3. Configuring and installing the SCSI disk
1. set the proper SCSI address
2. set the proper termination
4. Attaching disk cables
1. attach the cable connecting the hard disk to the HBA and the power cable
2. use impedance-matched cables with external disk subsystems
5. Setting the CMOS disk type
1. set the disk type to 0 or NOT INSTALLED unless this setting is controlled by the HBA
6. Completing the hard disk configuration and termination
1. run software to complete the configuration process, which may set termination and SCSI addresses
2. if jumpers not set for BIOS/Enabled, "DRIVE NOT READY—ERROR" or "COMPUTER CANNOT START"
3. Configure, install, prepare, and troubleshoot IDE disks on a workstation and a NetWare server
1. Configuring and installing the controller board
1. base I/O address - leave at the factory setting (must match address on controller board)
2. base memory - must be set for controller boards that have ROM BIOS (avoid addresses over E000h)
3. DMA channel - avoid letting two devices share a DMA channel when using NetWare
4. controller interrupt - leave at the factory defaults
2. Configuring and installing the IDE disk
1. single drive - only one IDE disk
2. master - boot disk when installing two IDE disks
1. if a single IDE disk is not set as master, "1782-Disk Controller Failure"
3. slave - not the boot disk when installing two IDE disks
3. Attaching disk cables
1. floppy disk drives and hard disk cables look similar, but have different cable twists; do not interchange
2. Pin 1 is identified with a colored stripe; match to the receptor marked with a "1" or a square dot
3. use a 40-pin cable, and make sure that the cable is not more than 18 in long
4. Setting the CMOS disk type
1. use the CMOS list
1. do not choose a disk type that has more cylinders or more heads than the hard disk
1. pick the disk type whose key characteristics match those of the disk to install
2. characteristics include number of cylinders, number of heads, capacity in MB, write precompensation
1. write precompensation - cylinders closer to center of disk have shorter tracks than those away
2. choose the type with the highest capacity that does not exceed the physical disk’s capacity
2. define a type
1. choose the User Defined disk type for the system (often Type 47)
2. insert the parameters, and record them on a label on the hard disk (data inaccessible if battery fails)
5. Completing the controller board and IDE disk configuration
1. run software to complete the configuration process, which may set controller board configuration parameters
4. Establish NetWare file storage by creating NetWare partitions and volumes, spanning hard disks, and mirroring hard disks
1. Preparing the hard disk for use
1. perform low-level formatting
1. tests the disk heads by running a surface analysis
2. creates the disk tracks in sectors and assigns sector IDs
3. marks off bad sectors to prevent data from being written to them
4. sets the interleave ratio and temporarily fills each sector
2. create DOS partitions (20 MB in size)
1. divides the disk into logical units called partitions (FDISK)
3. perform a high-level format
1. scans the disk and identifies bad sectors
2. creates the DOS boot sector
3. creates the DOS FAT
4. creates a blank root directory
5. copies system files (format /s)
2. Preparing NetWare partitions and volumes
1. create a NetWare partition
1. NetWare recognizes primary and extended DOS partitions and NetWare partitions
1. can only have one NetWare partition per disk for each version of NetWare
2. partitions for other operating systems are displayed as non-NetWare partitions
3. non-NetWare partitions cannot be mirrored under NetWare
2. DOS partition and the entire NetWare partition are considered physical partitions
1. NetWare identifies DOS partition as physical partition 0 and NetWare partition as physical partition 1
2. NetWare establishes the Hot Fix Redirection Area within the NetWare physical partition
3. part of the NetWare physical partition from the beginning to the end of the data area is a logical partition
2. create NetWare volumes (up to 64 volumes and 32 TB)
1. can be a part of the NetWare partition
2. can be the entire NetWare partition (Volume Definition Table keeps track of volumes)
3. can span volumes over multiple hard disks
1. spanning can be risky; losing one disk and its volume segment will lose access to the entire volume
3. mirror or duplex the partitions (set up using INSTALL)
1. mirroring and duplexing effectively reduce the chance of data loss and minimize network downtime
1. mirroring - same data is maintained on two or more logical partitions connected to same disk channel
1. protects against failure of hard disk, but not against failures in other disk channel components
2. accomplished using the operating system software or by hardware
3. best to mirror partitions that are the same size
2. duplexing - protects data against failure in any component along the disk channel
1. disk channel is the entire electronic pathway involved in hard disk storage and retrieval
3. NetWare switches from a failed partition to its duplicate without any interruption of service
2. tips for working with mirrored and duplexed disks:
1. keep a record of the device codes of the hard disks
2. always load disk drivers in the same order
3. periodically check the status of the mirrored or duplexed groups with INSTALL
4. perform system backups on a regular basis
5. perform additional backups after cleaning or changing the tape drive unit
3. mirroring might be inadequate for IDE disks
1. when running IDE disks as a pair, the controller of the second disk is disabled
1. second disk runs from the first disk’s built-in controller
2. failure in the primary IDE disk causes failure in the controller and the second disk
2. for true fault tolerance, operate IDE disks in duplexed configurations
1. use entirely separate IDE interfaces, using a second board and a second port address
4. RAID combines two or more standard physical disks in a single logical disk to achieve data redundancy
1. disk arrays (Redundant Array of Inexpensive Disks) have advantages over individual disks:
1. improve I/O performance by balancing the I/O load
2. improve data reliability by replicating data
3. simplify storage management
2. six levels of protection:
1. level 0: data striping and block interleave
2. level 1: disk mirroring or duplexing
3. level 2: data striping with bit interleave (does not work for personal computers)
4. level 3: bit-interleave data striping with parity checking
5. level 4: block interleave data striping with parity checking
6. level 5: block interleave data striping with distributed check-data (good compromise)
5. limitations of IDE disks under NetWare:
1. neither IDE nor DOS supports overlapped, multitasking I/O
2. IDE does not support bus mastering, which off-loads work from the CPU processor
3. older IDE products do not support tape or optical disks
4. IDE imposes a disk capacity limitation of 528 MB per disk without a special nonstandard BIOS
5. new installations of IDE disks may need to use the IDE.DSK driver in place of ISADISK.DSK
5. Install and configure a CD-ROM as a NetWare volume
1. Mounting a CD-ROM as a NetWare volume
1. physically install the CD-ROM drive
2. start the server
3. load the driver for the HBA attached to the CD-ROM drive (can load automatically in STARTUP.NCF)
4. for NetWare 3.12, load the AFTER311 and NPA312 drivers
5. load any special CD-ROM drivers, such as ASPICD
6. load the latest CDROM.NLM
7. enter CD DEVICE LIST
8. if the volume name does not appear, unload and reload CDROM.NLM
9. configure NetWare by entering SET RESERVED BUFFERS BELOW 16 MEG=200 in STARTUP.NCF
10. to mount, enter CD MOUNT volume_name or CD MOUNT media_manager_object_number
2. Networking implications of CD-ROM
1. if the CD-ROM drive shares a SCSI bus with a disk subsystem that contains volumes to which NetWare installation files are copied, the keyboard may lock up while copying
1. reconfigure the CD-ROM drive to use the parallel-to-SCSI adapter instead of the shared SCSI bus
2. obtain a separate SCSI board for the CD-ROM drive
3. re-create the SYS: volume on the internal hard disk of the server, rather than on the subsystem disk

Section 5: Troubleshooting the DOS Workstation

1. Troubleshoot malfunctioning installations of IPX- and ODI-based DOS workstations
1. Troubleshooting tips for DOS workstations running IPX and NETX
1. make sure the network board matches the IPX configuration (can check with IPX -i)
2. check for IRQ, DMA, and memory conflicts
3. check for Ethernet frame type conflicts
4. determine if IPX is properly bound to the LAN driver at the file server
5. load each LAN driver individually
6. run the diagnostics utilities for the network board to verify the settings
7. run COMCHECK to circumvent the NETX and determine whether a good physical connection exists
8. make sure the network board is properly seated
9. make sure the cables are properly connected
10. can add the name of a preferred file server to the LOGIN command or in NET.CFG
2. Open Data-Link Interface (ODI)
1. supports multiple frame types and protocols
2. simplifies support of multiple protocols on a single network
3. facilitates the integration of new protocols
4. allows software drivers to be upgraded without regenerating the shells
5. allows up to 4 active network boards in the client station
6. supports LANalyzer for Windows
7. provides improved memory management
3. Troubleshooting the DOS ODI workstation
1. to unload ODI drivers from memory, unload them in reverse order:
1. NETX u
2. IPXODI u
3. MLID u
4. LSL u
2. if other TSRs are loaded after NETX, unload those TSRs first
4. Troubleshooting tips for the NetWare DOS Requester (replaces NETX)
1. VLMs will load from the current directory
2. load order of VLMs is important (load child VLMs before multiplexer VLMs)
3. rename an unwanted VLM with a different extension, or do not include in NET.CFG
4. do not load NETX.COM or NETX.EXE with VLMs (use NETX.VLM instead)
1. if an .EXE file and a .COM file with the same name are both available, the .COM file will run
5. IPXNCP.VLM is the transport protocol implementation using IPX (requires IPXODI.COM)
2. Install and troubleshoot a remote boot workstation
1. How the remote boot process works
1. code in the boot PROM is executed during the boot sequence on the workstation
1. boot PROM allows a diskless workstation to boot from a remote boot disk image file on the server
2. when a diskless workstation starts, it requires system sectors, FATs, and DETs
2. use the DOSGEN program to create a boot disk image file (equal to the total RAM on the server)
3. use the BOOTCONF.SYS file to manage multiple boot disk image files
4. advantages of using remote boot workstations include cost, convenience, security, and speed
2. Remote Program Load NLM (RPL.NLM)
1. acts as a protocol stack with diskless workstations
2. install RPL in SYS:SYSTEM by typing LOAD RPL
3. RPLFIX allows workstations to load remotely with DOS 5.x and above
1. if using ODI drivers with the boot PROMs, load RPLODI.COM after LSL.COM
3. Troubleshooting the remote boot workstation
1. if "Error opening boot disk image file," the file server does not contain the remote boot disk image file
2. if "Batch file missing," make sure the AUTOEXEC.BAT is in either SYS:LOGIN or the default directory
3. if only one user can log in, flag .SYS files as Shareable and assign users the Modify right to SYS:SYSTEM
4. if using a Remote Reset PROM on a token ring network board and cannot boot, load the TOKENRPL.NLM
5. to determine if the boot PROM on the workstation is sending packets, use TRACK ON to watch for requests
6. load MONITOR to determine which file the workstation opens
7. boot the second workstation with the boot diskette created for the first workstation
3. Troubleshoot workstation connection and logon problems with TRACK ON
1. Workstation loads the client and hangs, or displays "A file server could not be found"
1. communication problem has occurred on the LAN
2. make sure IPX is properly bound to the network board driver
3. activate TRACK ON for the file servers and routers
4. load the workstation shell, and TRACK ON should report these messages:
1. Get Nearest Server
2. Give Nearest Server server_name
3. Route Request
5. if the messages do not appear, suspect a bad cable, hub, or connection
6. if only "Get Nearest Server" appears repeatedly, might have a network transmission problem
7. if "Get Nearest Server" and "Give Nearest Server" appear repeatedly, might have a physical problem
2. File server is running, but is not listed when a workstation initiates SLIST or NLIST
1. file server might not be receiving information on other servers or networks
2. activate TRACK ON on all the file servers
3. incorrectly configured network number will cause "WARNING!!! MULTIPLE ROUTER WITH SAME INTERNET ADDRESS" or "ROUTER CONFIGURATION ERROR!!! Router X claims LAN A is yyy!"
4. Diagnose and repair workstation conflicts based on resources such as memory, IRQ, and DMA
1. Processor modes:
1. real mode - computer emulates the 8086 chip
2. protected mode - allows access to enhanced features of the newer chip (multitasking)
2. Types of memory:
1. conventional - first 640 KB of main memory, available to DOS and user applications
2. upper - next 384 KB of memory, reserved for video, ROM BIOS, and add-on boards
3. expanded - uses expanded memory chips installed on an expansion board
4. high - first 64 KB of extended memory (1024 to 1088 KB)
5. extended - above the 1 MB address space, used by a disk cache program, device drivers, and TSRs
3. Memory optimization on the DOS workstation
1. with Load High, memory manager software moves the NetWare drivers and TSRs into upper memory
1. can leave over 600 KB of base memory for DOS applications
2. VLMs will attempt to load automatically
2. NOEMS parameter with EMM386.EXE eliminates the expanded memory page frame

Section 6: Troubleshooting the Server and the Network

1. Locate and install the most current software for a NetWare 3.12 or 4.1 server, including patches and NLMs
1. Patches to the operating system - on NetWire and NSEPro (load all patches in a kit)
1. dynamic patches - implemented as .NLM files that can be loaded and unloaded while the server is running
1. unloading a dynamic patch will restore the operating system to its original unpatched state
2. to load dynamic patches:
1. rename any old patches
2. copy the Patch Manager NLM and new patches into a subdirectory on the server’s C: drive
3. place the commands to load the Patch Manager and patches into STARTUP.NCF
2. static patches - DOS executables that modify the SERVER.EXE file
1. applied once and permanent (do not patch the SERVER.EXE file on the original distribution diskette)
2. to use static patches:
1. copy SERVER.EXE to SERVER.OLD
2. copy static patches to the same directory as SERVER.EXE
3. type patch_name SERVER.EXE
3. Patch Manager - NLM which tracks and manages all official patches for NetWare
1. available in the same file as the patches for the system (be sure to use the correct version)
2. must be loaded before any dynamic patches can be loaded
3. to view loaded patches, type PATCHES at the server console
2. Device drivers - form the interface between NetWare and devices (NLMs enable communication)
1. disk drivers - allow controller boards to communicate with hard disks
2. NPA drivers - called Host Adapter Modules (HAMs) and Custom Device Modules (CDMs)
1. help the NetWare Media Manager keep track of and communicate with storage devices and media
3. NLMs - latest versions on NetWire in the Novell Library area
1. self-extracting archives that can be downloaded to the hard disk (copy to the SYS:SYSTEM directory)
2. to unload the NLM being replaced, enter UNLOAD NLM_name at the server console
2. Diagnose and troubleshoot server-related abends and lockups
1. Understanding abends (Abnormal END of program)
1. CPU-detected errors - processor can interrupt program execution by issuing an interrupt or an exception
1. interrupt is generated by an external device that needs attention
2. exception is caused by the processor responding to a condition detected while executing instruction
3. nonmaskable interrupt (NMI) is hardware-specific and in no way related to NetWare
4. example: "Page Fault Processor Exception"
2. consistency check errors - internal tests in the NetWare operating system code (code-detected errors)
1. ensure the stability and integrity of internal operating system data
2. failed consistency check indicates some degree of memory corruption (serious error)
3. caused by corrupted operating system files, outdated drivers and NLMs, bad packets, or hardware failure
4. associated with defective memory chips, ESD, faulty power supplies, or power fluctuations
5. example: "SubAllocFreeSectors given invalid FAT chain end that was already free"
2. Analyzing abend messages
1. line 1: date and time - date and time the system halted
2. line 2: abend message string - text helps determine whether the abend is CPU-detected or code-detected
3. line 3: operating system version - version of the NetWare operating system on the server
4. line 4: current running process - thread or path of execution running in the operating system
5. line 5: current stack dump - part of the CPU’s stack for the current running process
3. Server lockups
1. full server lockup - no processes are allowed to run
2. partial server lockup - users might still be able to log in to the server and accomplish work
3. possible causes are a server or NLM thread caught in a tight loop, or a process that locks up resources
4. to diagnose the cause, generate a memory image file
4. Steps for troubleshooting server abends and lockups:
1. gather information about the problem (CONFIG.NLM)
1. error messages
2. hardware configuration
3. disk and LAN drivers
4. NLMs and .NCF files on the server
5. recent changes to the system
6. events occurring prior to the crash
7. known problems
2. understand the problem and identify probable causes
1. can better understand the problem by answering questions about the information and facts gathered
3. test possible solutions
1. apply current patches, drivers, and NLMs
2. replace components
3. simplify the configuration (one at a time)
4. discuss the problem with others
4. use debugging tools
1. MONITOR
2. network analyzers
3. memory image file
5. resolve the problem
1. can resolve software problems with patches, workarounds, and new drivers
2. must repair or replace any hardware causing the problem
6. document the problem
1. documenting the problem will greatly reduce the time needed to solve the same problem again
3. Diagnose and troubleshoot server-related performance problems
1. Disk I/O problems - CPU and bus I/O are functioning at peak performance
1. load MONITOR and check dirty cache buffers and current disk requests
2. using smaller disks to replace one large disk improves the speed of the disk system
1. more than one set of read/write heads can be active
3. upgrade to a faster disk controller
1. Fast SCSI II controllers transfer data at 10 MB/second
2. Fast/Wide SCSI II controllers transfer data at 20 MB/second
2. Network I/O problems - usually occur in applications that have a high network utilization rate
1. manifest themselves through the LAN/WAN Information or LAN information screen in MONITOR
1. may have a network I/O problem if the "Send packet too big," "Receive packet overflow count," "Receive packet too big count," "Send packet miscellaneous errors," "No ECB available," "Packet Receive Buffers," or "Receive packet miscellaneous errors" variables are constantly high or growing
2. upgrade to faster network boards, or divide overloaded network segments with a bridge or router
3. CPU problems - processor performance is rarely an issue
1. if using CPU-intensive applications, check with the software vendor for the proper configuration
2. off-load some processing requirements by using bus master LAN and disk controller boards
4. Bus I/O problems - difficult to distinguish from CPU problems
1. I/O devices contend with cache systems for access to main memory, but run much slower than the CPU
1. servers overcome this problem by allowing I/O devices and the CPU to access memory simultaneously
2. bus master boards help the CPU increase performance by handling data transfer
4. Use a protocol analyzer to diagnose performance problems and to alleviate overloaded networks or servers
1. Protocol analyzers
1. monitor network performance (trends, network traffic)
2. troubleshoot network errors
3. optimize the LAN
4. plan for growth
2. LANalyzer for Windows (LZFW)
1. works directly with Ethernet cabling (LZFW runs in the background)
1. network boards run in promiscuous mode, observing all packets and not just the ones addressed to it
2. for acceptable performance, run LZFW on a 386DX CPU with a 16-bit network board (32-bit is best)
2. capture trend data for up to 6 months to establish a baseline to set alarm thresholds (minimum of 1 month)
3. set the default values for alarm thresholds high enough that normal peak activity will not trigger an alarm
1. Packets/second - at 5-10% over normal
2. Utilization% - at 5% above normal (bandwidth)
3. Broadcasts/second - at 10
4. Fragments/second - at 15
5. CRC Errors/second - at 5 (more than one general error per second is a problem)
6. Server Overloads/minute - at 5
4. when activity exceeds the alarm thresholds:
1. Network Alarm indicator changes from green to red
2. LZFW sound the alarm
3. Alarm Clock appears in the lower corner
4. scrolling message appears along the bottom of the screen
5. when a network alarm activates:
1. read the error message
2. open the error log
3. ask the NetWare Expert for help (database that indicates the type of error and recommends a solution)
6. typical errors reported by LZFW:
1. CRC/Alignment errors - bad packet has been sent
1. packet has a faulty Frame Check Sequence (FCS) or does not divide evenly by eight
2. often caused by cable problems
2. fragment errors - packets contain less than 64 bytes and a faulty FCS are being sent
1. cause by collisions and are normal on Ethernet networks
2. if more than 2-3% of the total packets and the network load is high, install bridges or routers
3. if the network load is low, a network board or transceiver has failed
3. less common errors - oversize packets, undersized packets, and jabber errors
1. typically caused by Data-Link protocol violations
7. overloaded network identifies a network cabling system that has slowed down
1. symptoms:
1. error messages such as "Error receiving/sending on network xxxx"
2. slow response when opening or launching applications
2. causes:
1. increased number of network devices
2. increase in applications launched from the server
3. unusually large file transfers
4. increased Internet traffic
8. when a NetWare Client transmits data on the network, the IPX Retry Count is set to 0 and the IPX Receive Timeout Counter begins to increment
1. if the server does not respond within the Receive Timeout maximum, the workstation:
1. resets the Receive Timeout Counter to 0
2. increments the Retry Count to 1
3. retransmits the data (maximum retries is 20)
9. if the cabling segment is the bottleneck on the network, segment the network with routers
1. balance the load to ensure that the heavy cable users are located on different network segments
2. when an overloaded server cannot handle requests, it sends a "Request Being Processed" message
5. Create a data repair and recovery plan that uses NetWare and utilities to recover from hardware failure
1. Use VREPAIR to repair a damaged volume (run repeatedly until no errors are reported)
1. hardware failure either prevented a volume from mounting or caused a disk read error
2. power failure corrupted a volume
3. server displays memory errors and cannot mount a volume after a name space is added
4. need to remove a name space (to see added name spaces, type VOLUMES)
5. volume has bad blocks
2. Use DSREPAIR to repair damage to NDS information
1. can repair records, schema, bindery objects, and external references
2. -U parameter instructs DSREPAIR to run, exit, and then unload from memory
3. Restore data from a backup
1. most important element in a successful disaster recovery plan
2. test the backups before beginning repairs
4. Use third-party software to recover data (Ontrack Data Recovery for NetWare)
1. NetFile4 - saves files to diskettes, a DOS partition on a device, or to another server
2. NetScan4 - nondestructively repairs structural errors, and examines defective sectors
3. NetDisk4 - examines and modifies the data in any sector of a device, and accesses data from a file server
5. Use professional data recovery services
1. if the data is irreplaceable and valid backups are unavailable, do not attempt repairs
2. in the event of a head crash, data may be recoverable from the data areas not directly affected

Section 7: Troubleshooting Network Printing

• Diagnose and resolve NetWare printing problems related to physical connections, print queues, print servers, remote printing, configuration, print utilities, and PostScript files
• Prevent, diagnose, and resolve common problems with printers

1. Initial Steps in Troubleshooting Network Printing (first find out if printing ever worked)
1. Quick-fix techniques:
1. ask, "What changed?"
2. check cabling
3. turn the printer off and on
4. check the printer cover and paper feed
5. check the workstation printer redirection
6. check printing forms
7. look for console messages and printer error conditions
2. Determining if the problem occurs before or after the print queue:
1. if the print job never arrives at the print queue or if the print job status indicates Adding or Hold and never moves to Ready, then the problem is probably at the workstation
2. otherwise, the printing problem probably occurs in the queue or after the print job has left the queue
2. Troubleshooting Printer Problems at the Workstation
1. Symptoms of network printing problems occurring at the workstation:
1. printing problem is specific to one workstation only
2. print job does not arrive at the print queue or arrives corrupted
3. print job arrives at the print queue, but remains in the Adding or Hold mode
4. print job arrives at the printer, but merges with another print job
2. Basic troubleshooting steps:
1. determine if the application is network aware, and:
1. make sure the correct print queue or printer is selected
2. make sure you have the most current printer driver
3. verify that the application can print to other network printers
2. if the application is not network aware:
1. check the print job redirection parameters
2. check the printing setup inside the application
3. to see if local printing works, attach the printer directly to the workstation and print
4. check for conflicts with other workstation components
3. Advanced troubleshooting techniques and tips:
1. use PRINTDEF and PRINTCON to manage advanced printers and create the printer and job definitions
2. increase the buffer size up to a combined maximum of 255 bytes using PRINT HEADER and PRINT TAIL
3. if using graphics-based software, provide a menu-driven configuration file the users can access
4. if printing large graphic files, lengthen or disable the Timeout Count
5. to share PRINTCON job configurations, place the PRINTCON.DAT file in SYS:PUBLIC
6. if a user requires more than 37 print job configurations, issue multiple login IDs to the user
7. TAB parameter normally exchanges the ASCII tab character (09) for eight spaces
1. use the /NT switch with CAPTURE to turn off the tabs for laser printers
8. overnight print jobs generated through CAPTURE can be lost if a backup clears connections at a set hour
1. use the /KEEP switch with CAPTURE to preserve the part of the job already captured
9. print job cannot complete if the volume containing the queue or the user does not have enough disk space
1. increase the amount of disk space available on the volume or to the individual user
10. plotters do not have direct interaction with the application
1. save a print job to a file and then use NPRINT or PCONSOLE to print the job
3. Troubleshooting Printer Problems at the Print Queue
1. Symptoms of network printing problems occurring at the print queue:
1. print job was sent uncorrupted but is corrupted in the print queue
2. print server abends when accessing the print queue
3. printing occurs sporadically
2. Troubleshooting techniques and tips:
1. use the latest versions of workstation files, printer drivers, and utilities
2. use short queue names
3. make sure CAPTURE (or NPRINT) and the applications send output to the same LPT port
4. if not enough disk space for a new print job, "WARNING—CANNOT CREATE SPOOL FILE" appears
1. use VOLINFO to verify the space problem, and then clear or add the needed disk space
5. to fix a corrupted print queue, delete the print queue definition, redefine the print queue, and reassign the print queue to a printer (may need to run VREPAIR or BINDFIX)
6. non-alphanumeric characters in the queue name may cause problems
4. Troubleshooting Printing Problems at the Print Server
1. Symptoms of network printing problems occurring at the print server:
1. print job status goes to Active, but the print job never prints
2. print job leaves the print queue, but never prints
2. Basic troubleshooting steps:
1. most problems with the print server occur during initialization
2. print job arrives at the print queue, but never starts or completes printing
1. check the Queue Status field to make sure the print server is attached to the print queue
2. make sure the print server is running
3. bring down the print server and re-initialize it
4. if using a third-party print server, consult with the print server vendor
5. if the problem persists, contact Novell Technical Support
3. Advanced troubleshooting techniques and tips for PSERVER.NLM:
1. use the latest version of PSERVER.NLM
2. use caution when bringing down the print server with UNLOAD PSERVER
1. allow the current print job to complete before bringing down the print server
2. use NetWare Administrator or PCONSOLE
4. Advanced troubleshooting techniques and tips for PSERVER.EXE:
1. to automatically reboot the print server after an interruption, use NETERR.ZIP from NetWire
2. PSERVER.EXE, combined with DOS and NetWare shell, requires 512 KB of conventional memory
3. if PSERVER.EXE hangs the workstation, reboot the computer
1. caused by network board conflicts or by older versions of IPX or the NetWare shell
4. set SPX CONNECTIONS=60 when working with PSERVER.EXE
5. Advanced troubleshooting techniques and tips for both PSERVER.EXE and PSERVER.NLM
1. use the most recent versions of PSERVER.EXE and PSERVER.NLM
2. plotters can be confused by software-generated print servers (applications expect interactive dialogue)
3. difference in performance between a file server-based print server and a dedicated print server is negligible
4. file server performance can slow down when printing large graphics files (add RAM)
5. slow performance may indicate an overloaded print server
6. "not enough free buffers" or "unable to create display portal" indicates a lack of server RAM
7. if prompted for a password, suspect either a misspelled print server name or print server configuration error
8. corrupted print server definition can cause slow or erratic printing (recreate and re-initialize the print server)
9. PCONSOLE uses an SPX connection when the Status/Control feature under Print Server Status/Control runs
5. Troubleshooting Printing Problems at the Remote Printer Workstation
1. Symptoms of network printing problems occurring at a remote printer workstation:
1. print job passed uncorrupted through the print queue, but arrived corrupted at a remote printer
2. print job contains dropped characters or random errors
3. print job status goes to Active or leaves the print queue, but never prints
2. Basic troubleshooting steps:
1. confirm that the printing problem occurs after the print job leaves the print queue
2. redirect the print job in the print queue to a different network printer
3. attach a printer operating correctly to the same remote printer workstation and try printing to this printer
3. Advanced troubleshooting techniques and tips for initialization:
1. "Not Connected" indicates that RPRINTER or NPRINTER has not yet been run to activate a remote printer
2. remote serial printers must be configured consistently with their print server (through DIP switches)
1. configured via PCONSOLE and NetWare Administrator
2. both must agree on data speed, data bits, stop bits, handshaking, and parity
3. if a remote printer completes a self-test but does not work as a network device, check the configurations
3. use the latest versions of RPRINTER.EXE, NPRINTER.EXE, and NPRINTER.NLM
4. copy the remote printer files to the local drive, or put them in the network LOGIN directory
5. trouble getting RPRINTER.EXE or NPRINTER.EXE to initialize at the workstation often has a lack of memory, incompatible clone, or obsolete versions of software
6. RPRINTER and NPRINTER can be installed successfully only after PSERVER is up and running
7. RPRINTER or NPRINTER may conflict with the network board settings and hang remote printer workstation
8. use RPRINTER and NPRINTER in polled mode when working with Windows or experiencing port conflicts
9. print job on a remote printer will be ruined if the user on the workstation to which the printer is physically attached does not have CAPTURE invoked and attempts to print locally (PSC PRI)
10. when re-establishing a lost connection, might get a message that the remote printer is still in use
1. previous SPX connection has not timed out and torn down the session information
4. Advanced troubleshooting techniques and tips on the printing process:
1. printing at a remote workstation might be slower than using the same printer locally from DOS because:
1. IRQ conflict exists
2. workstation may not support hardware interrupts
3. some printers require nonstandard parallel cables
2. be careful when using RPRINTER or NPRINTER with programs that will compete for memory
3. router between the remote printer workstation and the print server can cause problems
4. if remote printers hang periodically and at random, increase SPX ABORT TIMEOUT=larger_number and IPX RETRY COUNT=larger_number in the NET.CFG
5. RPRINTER and NPRINTER can potentially conflict with other TSRs (unload one at a time)
6. non-certified hardware or older drivers can cause RPRINTER and NPRINTER to behave erratically
7. older PCs acting as print servers can be slow if they are directly attached to printers
8. SPX connection made by RPRINTER or NPRINTER can be impeded by physical problems
6. Troubleshooting Printer Problems at the Printer
1. Symptoms of network printing problems occurring at the printer:
1. print job passes uncorrupted through the print queue, but never prints or is corrupted when it prints
2. print job prints properly when a different printer is attached to the same printer port
2. Troubleshooting techniques and tips:
1. if print jobs arrive at the print queue but do not print, look for physical problems at the printer
2. network printer going offline but working fine when re-initialized might have a static problem
1. carbonless forms can generate large amounts of static (attach a ground to a metal part of the printer)
3. parallel printing is faster than serial printing
4. some word processor drivers are substantially faster than earlier versions (use latest versions)
5. network printer cable contains several wires, each of which connects to a particular pin on the connector
6. if the printing setup does not provide adequate throughput:
1. add more queues
2. add multiple printers to a queue that has user notification
3. use high-speed printers
4. change the physical design of the printing workgroups
7. if the printer has insufficient buffer size in memory, the print jobs might be missing characters or words
3. Parallel printers:
1. 4 to 6 times faster than serial
2. 10 feet maximum standard distance
3. limited error checking but relatively error free
4. interrupt only must be set by the installer for parallel port
5. universally compatible
4. Serial printers:
1. slower than parallel
2. 50 feet maximum standard distance
3. uses parity, reducing speed by 10%
4. installer sets interrupt, XON/XOFF, parity, baud rate, data bits, and stop bits
5. compatibility can be a problem
5. Troubleshooting tips for working with PostScript printers
1. update the PostScript drivers for applications
2. make sure that the PostScript cartridge is properly installed in the bay
1. if not properly installed, the PostScript code from the application will be unintelligible
3. some printers require that the SYS SWITCH be ON to enable PostScript
4. use No Banner (/NB) and No Tabs (/NT) for all PostScript print jobs
5. use No Form Feed (/NFF) in NPRINT and CAPTURE
6. when PRINTDEF is used to create PostScript Printer Definition Files (PDFs) that include modes larger than the default shell header buffer size of 64 bytes, a problem can occur

Section 8: Introduction to Network Management

1. Describe the role of network management
1. Categories of network management:
1. administrative tasks - day-to-day configuration and troubleshooting (reactive)
1. network or LAN administrator configures new workstations, adds new users, and assigns rights
2. management tasks - planning, installing, and maintaining (proactive)
1. network manager or escalation engineer solves the most difficult network system problems
2. Common tasks of a network manager:
1. evaluate proposed business solutions and determine a course of action
2. remotely access, manage, troubleshoot, and monitor network devices
3. install an integrated, network-wide virus protection system
4. inventory, re-engineer, and optimize the existing network system
5. perform network analysis, including setting up early warning systems
6. gather information for evaluating system performance
7. gather information about real-time problem notification, diagnosis, and solution information
8. identify performance bottlenecks, bad equipment or configurations, and solutions to network problems
9. provide network environment management
3. Tool-based versus business-based approach
1. primary purpose of network management is to increase the organization’s profitability
2. tool-based approach causes network managers to consider only the tasks that their tools perform well
3. business-based approach makes optimum business decisions and picks the appropriate tools to do the job
2. Explain the network management life cycle
1. Analysis and specification phase
1. recognize and state the underlying business need
2. gather general information about the project
2. Design phase
1. identify alternative generic solutions
2. evaluate alternatives
3. design a specific solution
3. Implementation phase
1. define implementation-specific milestones
2. develop and test a working solution
3. document the solution and train the users
4. test the final solution and obtain customer acceptance
4. Maintenance phase
1. monitor the network system in real time
2. establish ongoing checks and system performance reports
3. correct new problems
3. Describe the major features of ManageWise (server component and workstation console portion)
1. Asset inventory and management
1. automatic discovery
2. integrated graphical maps
3. hardware and software inventory
4. address management
2. NetWare server monitoring and management
1. multiple-server management
2. unattended monitoring
3. server optimization
4. print queue monitoring
3. Desktop management (LANDesk Manager)
1. manage workstation configuration
2. access and manage the desktop remotely
3. activate chat mode
4. perform file transfer
4. Network analysis (NetWare LANalyzer)
1. traffic analysis
2. network troubleshooting (supports multiple networks and multiple protocols)
5. Network-wide virus protection
1. comprehensive scanning
2. continuous real-time protection
3. mobile user protection
4. automatic self-maintenance