This chapter is a tutorial to walk users who are new to IBM routing products through the concepts and basic navigation of the Network Utility command-line interface. It covers:
The tutorial text makes the most sense if you follow it from beginning to end with the same Network Utility.
If you already have experience with the IBM 2216, you will find the Network Utility interface to be nearly identical. This is also true for IBM 2212 users, except for the firmware interface. IBM 2210 users will find familiar prompts and menu navigation, but differences in areas including configuring adapters, saving configurations, and rebooting the product.
If you followed one of the initial configuration procedures in Chapter 3, Performing the Initial Configuration, you have configured your Network Utility and booted it into normal operating mode. Your user console should show the asterisk (*) command prompt.
In normal operating mode, the routing function in Network Utility is running. You as the operator can use the command-line interface to look at and modify the configuration, view the active system status, look at the message log, and so on. You navigate to different parts of the command-line interface to perform these different tasks, and the * prompt is the root of the navigation tree.
Type ? from the * prompt to see the commands available from this point:
*?
CONFIGURATION (Talk 6)
CONSOLE (Talk 5)
EVENT Logging System (Talk 2)
ELS Console (Talk 7)
LOGOUT
PING <IP Address>
RELOAD
TELNET to IP-Address <this terminal type>
---------------------------------------------------
DIAGS hardware diagnostics
DIVERT output from process
FLUSH output from process
HALT output from process
INTERCEPT character is
MEMORY statistics
STATUS of process(es)
SUSPEND command completion
TALK to process
*
Although each of these commands has its purpose, you will use two of them far more than any of the others:
To use the talk command, type t n, where n (a process id) usually takes one of the following values:
To undo the talk command and move from inside any process directly back to the * prompt, type Ctrl-p.
MAS V3.3 introduced more natural commands that perform the same function of the talk command. In place of talk 6, you can simply type config. Likewise, console can substitute for talk 5 and event for talk 2.
The following three sections describe each of the major processes and explain some of the tasks you can perform inside each process. Along the way, you will become familiar with moving around between processes and menus, and entering commands.
From the * prompt, type t 6 or config to enter the command-line process for configuring the Network Utility:
* <Enter>
*t 6
Gateway user configuration
Config> <Enter>
Config>
Now that you are inside the Config process, the command prompt has changed from * to Config>. Both the Config and Console processes have unique prompts so you can tell at a glance which process you are in. The status message Gateway user configuration shows up only the first time you enter the Config process following a reboot ("gateway" is used as a synonym for "router" in various places in the system).
When you have been in a process before and reentry it using the talk command, the system gives you a blank line instead of an immediate command prompt. Press Enter and you are returned to where you were the last time you were inside that process:
Config> <Ctrl-p> <---- leave Config and go back to *
* <Enter>
*t 6 <---- go back into Config
<Enter>
Config> <---- we're back at the main Config prompt
When you are working inside the Config process, you are changing how the Network Utility is configured to operate. With a few exceptions, these changes have no effect on the running state of the router. To activate talk 6 changes you must either:
As you follow this tutorial you will see examples of both methods.
At the main Config> prompt, type ? to see an alphabetical list of the commands available to you:
Config>? ADD (device, user) BOOT and load file functions CHANGE (device, password, user) CLEAR configuration information DELETE (interface, user) DISABLE (interface, console-login, etc) ENABLE (interface, console-login, etc) EVENT logging system and messages FEATURE (non-protocol and network features) LIST (devices, configuration, patches, users) LOAD (add, delete, list) NETWORK interface configuration PATCH global configuration parameters PERFORMANCE monitor PROTOCOL configuration QCONFIG (quick configuration) SET system-wide parameters SYSTEM TIME of day parameters UNPATCH global configuration parameters UPDATE WRITE Config>
Some of these commands are for actually configuring the functions of the box, and others are for configuration management and system administration. To give you a feel for the types of things you do under talk 6, the following list groups key commands by user task:
The following examples show how to use some of these talk 6 commands to perform basic configuration tasks. As you work through the examples, you will gain experience not only with the tasks shown, but also generally with moving around through menus and issuing commands. The examples begin with a task that may already be familiar if you used the command line procedure for initial configuration.
In the running example used throughout this tutorial, the user first booted a Network Utility with the following configuration:
If you want to follow this example, use clear dev to erase your own device configuration 6 and then use add dev and del int to enter the ESCON/TR device configuration, as shown below.
From the Config> prompt, type list device (or li dev, abbreviated) to see the adapters and ports that are defined in the current configuration. If you have no configuration, or no adapter ports defined, li dev gives no output but simply reissues the user prompt. Because you have cleared all devices, you can add one. Type add dev ? to see a list of all the adapter types you can add:
Config>clear dev You are about to clear all Device configuration information. Are you sure you want to do this? ? [No]: yes Device configuration cleared Config>li dev Config>add dev ? ATM 1-port 155 Mbps ATM adapter EIA-232E 8-port EIA-232E/V.24 adapter ESCON Channel 1-port ESCON Channel adapter ETHERNET 2-port Ethernet adapter ETH100 1-port 10/100 Mb Ethernet adapter FDDI 1-port FDDI adapter HSSI 1-port HSSI adapter PCA 1-port Parallel Channel adapter TOKEN-RING 2-port Token-Ring adapter V35/V36 6-port V.35/V.36 adapter X21 8-port X.21 adapter Config>
Use the add dev command to configure a single port on a single adapter. For a multi-port adapter, you must specify which port you are adding to the configuration, and reissue the command for each port you want to have active. Here we add a single-port ESCON adapter, and both ports of a 2-port Token-Ring adapter:
Config>add dev esc Device Slot #(1-2) [1]? 1 Adding ESCON Channel device in slot 1 port 1 as interface #0 Use "net 0" to configure ESCON Channel parameters Config>add dev tok Device Slot #(1-2) [1]? 2 Device Port #(1-2) [1]? 1 Adding Token-Ring device in slot 2 port 1 as interface #1 Use "net 1" to configure Token-Ring parameters Config>add dev tok Device Slot #(1-2) [1]? 2 Device Port #(1-2) [2]? 2 Adding Token-Ring device in slot 2 port 2 as interface #2 Use "net 2" to configure Token-Ring parameters Config>li dev Ifc 0 ESCON Channel Slot: 1 Port: 1 Ifc 1 Token-Ring Slot: 2 Port: 1 Ifc 2 Token-Ring Slot: 2 Port: 2 Config>
To specify the adapter type, you type on the same line as add dev the first few characters of the words in the left column of the add dev ? output list (enough characters to distinguish the adapter type you want). When prompted, you must supply the slot and (for multi-port adapters only) the port number. Slot and port numbering is fixed as follows:
The add dev command makes sure that you do not try to add two different adapters in the same slot, add an adapter to a slot that does not exist, or specify a port number that does not exist on a given adapter. It does not validate the device type you select against the adapters that are physically installed in your Network Utility. This allows you to configure adapters you have not yet installed, or produce a configuration for a different Network Utility. The system validates device configuration only when you boot up with a particular configuration or try to activate an interface dynamically. The system reports mismatches through the LEDs on the front of your adapter, as well as from a locally viewable event log. You can also type commands to see adapter status, as you will see later in this tutorial.
In response to your add dev command, the Network Utility assigns a logical interface number or net number to the port you have just added. This is the key number by which you refer to this interface on every other command in the system. Only the add dev command uses physical slot and port numbers; all other commands use the logical interface number. When you subdivide a physical ("base") port such as ESCON into multiple virtual interfaces, each virtual interface also has an interface number. As shown above, you can use the li dev command to see the interface number for every physical and virtual interface.
If you make a mistake and want to undo the add dev command, or want to delete adapter/port configuration for any reason, use the delete interface command. (It is not named "delete device", because it deals with logical interface numbers and not adapter slot/port numbers.) To continue the example, assume that you want to use only port 2 of the Token-Ring adapter. Delete port 1 (which happens to be interface 1) as follows:
Config>li dev
Ifc 0 ESCON Channel Slot: 1 Port: 1
Ifc 1 Token-Ring Slot: 2 Port: 1
Ifc 2 Token-Ring Slot: 2 Port: 2
Config>del int
Interface number? 1
Interface being deleted... please be patient.
The router must be restarted
Interface 1 deleted successfully
Config>li dev
Ifc 0 ESCON Channel Slot: 1 Port: 1
Ifc 1 Token-Ring Slot: 2 Port: 2
Config>
Note that Token-Ring port 2 has now become logical interface 1. If there had been other interfaces with numbers higher than 1, these numbers also would have been decremented by 1. If you want to delete every interface in a configuration, just delete interface 0 repeatedly until there are no more interfaces.
In addition to device configuration itself, it is normal to have protocol configuration that is associated with a particular interface. When you delete an interface using the del int command, the system also deletes all protocol configuration associated with that interface, and renumbers all protocol configuration associated with renumbered interfaces7. You need to reboot Network Utility for a del int operation to take effect in the running system.
To look more closely at how to issue commands in general, try something simple, like using the set command to set up a name ("host name") for this Network Utility.
| Note: | This example assumes you are running with Command Completion disabled. See Automatic Command Completion to understand how the Network Utility can provide automatic command completion. |
Config>set
Command not fully specified
This error message reports that the set command is backed by a menu of additional keywords, and you need to type more keywords until you form a complete command that will perform an action. Anytime you are at a menu (as you have seen already), you can type ? to see the available commands or keywords to type. If you are just trying to remember command keywords, it is usually much faster to move around typing ?, than it is to look up the command in a manual. In this case, the options are:
Config>set ? CONTACT-PERSON DATA-LINK DOWN-NOTIFY GLOBAL-BUFFERS HOSTNAME INACTIVITY-TIMER INPUT-LOW-WATER LOCATION PACKET-SIZE PROMPT RECEIVE-BUFFERS SPARE-INTERFACES
As you can see, the set menu includes a mix of data items: some for system administration, some for node tuning, and so on. In Network Utility, node tuning options are defaulted for you and you should not have to change them.
Back to the task, the keyword you want is clearly "hostname". You can abbreviate any menu item (command name or keyword) to the number of characters needed to make it unique, so shorten "hostname" a bit:
Config>set host
Host name for this node []? rtp01
Host name updated successfully
rtp01 Config>
By default, the system inserts the new host name in front of all command prompts. Many users like this because it enables them to Telnet into a number of routers from a single work station and easily distinguish one router console from another. If you want to choose a different prompt prefix, you can use the set prompt command to do so. To reset either the host or the prompt to a null value, use the clear host or clear prompt command and reboot the Network Utility. To look at the current values, use list config.
Note that set host is an exception to the normal talk 6 rule in that it took effect immediately and did not require you to issue some sort of "activate" command, or to reboot the Network Utility. Very few talk 6 commands behave this way, but this one is useful because you can immediately see its effect on the user prompt.
Suppose you do not like the new prompt and want to change the host name from "rtp01" to "RTP01". You can do this in a single command, as follows:
rtp01 Config>set host RTP01
Host name updated successfully
RTP01 Config>
The system did not prompt you for the host name because you typed it on the original command line. This illustrates another general rule: when a full command prompts you for input parameters, you have the option of typing them on the original command line and skipping the prompts. If you choose to skip prompts, you need to be careful to type parameters in the right order.
Now that you have configured your host name, try something a bit more complex. Suppose you noticed when you rebooted from Config-only mode, that your newly configured Token-Ring adapter port 2 did not come up. You can see what ring speed it is configured for, and change that value. This sort of low-level device-specific configuration parameter is what you use the net command for, as shown here:
RTP01 Config>li dev <------ what were those i/f numbers again? Ifc 0 ESCON Channel Slot: 1 Port: 1 Ifc 1 Token-Ring Slot: 2 Port: 2 RTP01 Config> <Enter> RTP01 Config>net 1 <------ I configure interface 1 Token-Ring interface configuration RTP01 TKR config> <Enter> <------ note the new subprocess prompt RTP01 TKR config>? <------ what are the commands here? EXIT FRAME LIST LLC MEDIA SET PACKET-SIZE bytes SOURCE-ROUTING SPEED Mb/sec RTP01 TKR config>li <------ show me what I have now Token-Ring configuration: Packet size (INFO field): 2052 Speed: 4 Mb/sec <----- It should be 16Mb/sec Media: Shielded RIF Aging Timer: 120 Source Routing: Enabled MAC Address: 000000000000 RTP01 TKR config>speed Speed (4 or 16) [4]? 16 <------ change the speed here RTP01 TKR config>li <------ verify the new value Token-Ring configuration: Packet size (INFO field): 2052 Speed: 16 Mb/sec <----- looks good now Media: Shielded RIF Aging Timer: 120 Source Routing: Enabled MAC Address: 000000000000 RTP01 TKR config>ex <------ exit the subprocess RTP01 Config> <------ you are back at the main T 6 menu
This change to the ring speed does not take effect immediately, but requires a talk 5 command or reboot to activate it. Dynamic Reconfiguration covers the basics of activating configuration changes without a reboot. Normally, you use the net command immediately after add dev to look at the default settings for the new interface and make any necessary changes before activating the port for the first time.
In this example, when you typed net 1, you moved into a subprocess for configuring Token-Ring interfaces. The base menu changed and the prompt also changed to let you know you were no longer at the main Config> menu, but one level deeper. In order to leave any subprocess menu and return to the next higher one, type exit. Remember also that Ctrl-p immediately takes you all the way out to the * prompt, and when you return to that process you reenter where you were last:
RTP01 Config> <Enter> <------ start here
RTP01 Config>net 1 <------ enter a Config subprocess
Token-Ring interface configuration
RTP01 TKR config> <Ctrl-p> <------ jump out
RTP01 * <Enter>
RTP01 *t 6 <------ go back to Config
<Enter>
RTP01 TKR config> <Enter> <------ you are back in the subprocess
RTP01 TKR config>ex <------ exit the subprocess
RTP01 Config> <------ You are back where you started
Now try two more examples in the Config process, and then move on to the Console process. The first example shows how to reduce the time to reload the box, and the second shows how to change parameters associated with a box protocol.
From the Config> prompt, type boot to reach the subsystem for managing configurations, code loads, and boot options. Chapter 7, Handling Configuration Files gives the full background on this subsystem, so you do not need to look at all the commands here. Look under the enable command and try out the "fastboot" option:
RTP01 Config>boot <----- enter subprocess
Boot configuration
RTP01 Boot config> <Enter> <----- note new prompt
RTP01 Boot config>en ? <----- list "enable" options
AUTO-BOOT-- set Unattended mode
FAST-BOOT-- bypass diags
RTP01 Boot config>en fast <----- try out "fast-boot"
FastBoot mode is now enabled.
Operation completed successfully.
RTP01 Boot config>ex <----- exit the boot subprocess
RTP01 Config>
If you watched the console bootup messages when you powered on your Network Utility or typed the reload command, you may have noticed that the system runs through a number of power-on diagnostics when it is booting. While this is desirable for a production router that is rebooted infrequently and whose hardware should be validated, it does lengthen the boot time. If you are actively configuring and repeatedly rebooting a given router, you may wish to reduce the boot time by skipping these diagnostics. You have just done this with the enable fast-boot command. The next time you do a reload, it will proceed more quickly. You should undo this change using disable fast-boot before placing the Network Utility into production.
Note that the fast-boot mode can be controlled only by the command line and not from the Configuration Program. The system's boot mode is stored in nonvolatile memory on the box, and is not part of the configuration file.
The final Config process example uses the menus and commands of the IP protocol subprocess to change an interface IP address. As noted on page ***, this example started with a Network Utility that had an IP address configured on Interface 1 (port 2 on the Token-Ring adapter in slot 2).
RTP01 Config>li dev <----- what are the intfcs again?
Ifc 0 ESCON Channel Slot: 1 Port: 1
Ifc 1 Token-Ring Slot: 2 Port: 2
RTP01 Config>p ip <----- short for "protocol ip"
Internet protocol user configuration
RTP01 IP config> <enter> <----- now in IP Config subprocess
RTP01 IP config>li addr <----- list configured IP addresses
IP addresses for each interface:
intf 0 IP disabled on this interface
intf 1 192.1.1.8 255.255.255.0 Local wire broadcast, fill 1
RTP01 IP config>change addr
Enter the address to be changed []? 192.1.1.8
New address [192.1.1.8]? 192.7.7.7
Address mask [255.255.255.0]? <enter>
RTP01 IP config>li addr <----- verify the change
IP addresses for each interface:
intf 0 IP disabled on this interface
intf 1 192.7.7.7 255.255.255.0 Local wire broadcast, fill 1
RTP01 IP config>ex <----- exit IP config
RTP01 Config>
This is the first example of using the protocol command to enter the subprocess for an individual protocol. IP is just one of many protocols you could have selected, and there is a similar list of features you can access using the feature command. Type list config from Config> for a full list of the protocols and features you can configure, or just p ? or f ? for a quick reminder. All protocols and features work the same way: you enter the subprocess for a protocol or feature, configure it using commands specific to that protocol or feature, then exit to the main Config> prompt.
For detailed command reference material on configuring any given protocol, refer to the chapter relating to that protocol in one of the two volumes of MAS Protocol Configuration and Monitoring Reference. Each of these chapters provides introductory material about the protocol, and a description of each configuration and monitoring console command for that protocol. For the same information about MAS features, refer to MAS Using and Configuring Features.
You have now completed the overview of the Config process and its commands. You can move on to talk 5, the Console process. Remember, to leave any process you type Ctrl-p to reach the * prompt and then you are ready to use the talk command to enter another process:
RTP01 Config> <Ctrl-p>
RTP01 *
From the * prompt, type t 5 or console to enter the command line process for monitoring and controlling the Network Utility's active state:
RTP01 * <Enter>
RTP01 *t 5
CGW Operator Console
RTP01 + <Enter>
RTP01 +
Now that you are in the Console process, the command prompt has changed from * to +. The Config and Console processes and their subprocesses have unique prompts that indicate your position at a glance. The status message CGW Operator Console shows up only the first time you enter the Console process following a reboot. As explained with talk 6, if the system gives you a blank line when you type t 5, that means you have been in talk 5 before and need to press Enter to resume wherever you were last.
When you are working in the Console process, you type commands to view and modify the active running state of the Network Utility. You cannot modify the Network Utility's configuration files from this process. Some talk 5 commands allow you to dynamically modify configuration parameters, but these changes are lost when you reboot the Network Utility. If you have made configuration changes under talk 6, however, you can dynamically activate some of them from talk 5 without rebooting the Network Utility.
At the main + prompt, type ? to see an alphabetical list of the commands available to you:
RTP01 +?
ACTIVATE interface
BUFFER statistics
CLEAR statistics
CONFIGURATION of router
DISABLE interface or slot
ENABLE slot
ERROR counts
EVENT logging
FEATURE commands
INTERFACE statistics
MEMORY statistics
NETWORK commands
PERFORMANCE monitor
PROTOCOL commands
QUEUE lengths
RESET interface
STATISTICS of network
TEST network
UPTIME
RTP01 +
Some of these commands are for viewing the status of the box and some are operator commands for actively changing that status. In addition, under each protocol and feature there is a Console subprocess containing a mixture of these two command types. The following list groups key talk 5 commands by user task:
As you did from talk 6, try some of these talk 5 commands. Those for viewing box status are all quite simple; you simply type the one-word command and look at the output:
RTP01 +mem
Physical installed memory: 256 MB
Total routing (heap) memory: 228 MB
Routing memory in use: 3 %
Total Reserve Never Perm Temp Prev
Alloc Alloc Alloc Alloc
Heap memory 239390720 26616 232309212 7029792 49828 1888
Number of global buffers: Total = 1000, Free = 1000, Fair = 194, Low = 200
Global buff size: Data = 4478, Hdr = 82, Wrap = 72, Trail = 7, Total = 4644
RTP01 + <Enter>
RTP01 +buff
Input Buffers Buffer sizes Bytes
Net Interface Req Alloc Low Curr Hdr Wrap Data Trail Total Alloc
0 ESCON/0 255 255 20 0 86 72 4478 0 4636 1182180
1 TKR/0 250 250 7 0 85 72 2052 7 2216 554000
As you can see, mem shows box-level status, while buff gives interface-level information. For all the commands that give per-interface information (buff, config, error, int, queue, stat), you can specify a list or range of interface numbers you are interested in:
RTP01 +int 0-1
Self-Test Self-Test Maintenance
Net Net' Interface Slot-Port Passed Failed Failed
0 0 ESCON/0 Slot: 1 Port: 1 0 0 0
1 1 TKR/0 Slot: 2 Port: 2 0 0 0
RTP01 +stat 1
Net Interface Unicast Multicast Bytes Packets Bytes
Pkts Rcv Pkts Rcv Received Trans Trans
1 TKR/0 0 0 0 0 0
Refer to the MAS Software User's Guide chapter "The Operating/Monitoring Process" for a description of the fields in each command's output.
The config command is particularly important, because at the end of the output is the status of all specified interfaces (this example output is edited to remove blank lines):
RTP01 +c
Multiprotocol Access Services
NetU-TX1 Feature 1001 V3.1 Mod 0 PTF 1 RPQ 0 MAS.DE1 netu_38PB
Num Name Protocol
0 IP DOD-IP
3 ARP Address Resolution
11 SNMP Simple Network Management Protocol
29 NHRP Next Hop Resolution Protocol
Num Name Feature
2 MCF MAC Filtering
7 CMPRS Data Compression Subsystem
8 NDR Network Dispatching Router
10 AUTH Authentication
2 Total Networks:
Net Interface MAC/Data-Link Hardware State
0 ESCON/0 ESCON ESCON Channel Not present
1 TKR/0 Token-Ring/802.5 Token-Ring HW Mismatch
RTP01 +
The Network Utility from which this example output was captured in fact has an empty slot 1 and an Ethernet adapter in slot 2. In talk 6, it does not matter if what you configure does not match the installed adapters, but when you reboot with that configuration, talk 5 will show you that your configured interfaces have not come up.
If you had configured correctly, the interface state would start with "Testing" then move to "Up", and you would be able to use the net command to enter an adapter-specific Console subprocess to get more detailed status information. As it is now, you get the following:
RTP01 +net 0
Network interface is not available.
RTP01 +
To view and control what's currently going on with any given protocol, use the protocol command to enter the Console subprocess for that protocol. As explained previously, p ? will generate a quick list of the protocols supported in a given software load. For example, select Data Link Switching (DLSw):
RTP01 +p dls <----- short for "protocol dlsw"
Protocol DLSW is available but not configured
RTP01 +
DLSw is available because it is supported by this software load8, but it is not configured because you never went into talk 6 and entered the commands to enable DLSw. Now that you have booted the box without DLSw in the configuration, it is not running and there is no DLSw status to view or modify from talk 5.
As noted on page ***, this example started on a Network Utility already booted with an IP configuration. IP is therefore actively running, so you can enter its Console subprocess and see what commands are available:
RTP01 +p ip <----- short for "protocol ip"
RTP01 IP>?
ACCESS controls
CACHE
COUNTERS
DUMP routing tables
INTERFACE addresses
PACKET-FILTER summary
PARAMETERS
PING dest_addr [src_addr size ttl rate]
REDUNDANT Default Gateways
RESET
RIP
ROUTE given address
ROUTE-TABLE-FILTERING
SIZES
STATIC routes
TRACEROUTE dest_addr [src_addr size probes wait ttl]
UDP-FORWARDING
VRID
VRRP
EXIT
RTP01 IP>
If you compare this command list with that generated in talk 6 by typing ? at the IP config> prompt, you see that the talk 5 and talk 6 commands are quite different. In talk 5, for example, you can initiate a ping to see if you can reach a given IP address from the Network Utility. Because this is an active command immediately operating on an active network interface, it does not belong in talk 6. Other commands to view active status likewise are talk 5 commands and not talk 6 commands.
In talk 6 you changed the IP address of Token-Ring port 2 from 192.1.1.8 to 192.7.7.7. Now see what value appears under talk 5:
RTP01 IP>int <----- short for "interface"
Interface IP Address(es) Mask(s)
TKR/0 192.1.1.8 255.255.255.0
The talk 6 change had no effect on the operational state of the Network Utility, because you have not yet activated it either by explicit command or by rebooting. Use the command reset ip to reread the current talk 6 IP configuration and dynamically activate it in the running system:
RTP01 IP>reset ip
RTP01 IP>int
Interface IP Address(es) Mask(s)
TKR/0 192.7.7.7 255.255.255.0
RTP01 IP>ex
RTP01 +
As you can see, the IP address change (and any other IP changes you made under talk 6) are now active. Most protocols have some mechanism for dynamic reconfiguration, but not every protocol has a reset command under talk 5. See Dynamic Reconfiguration for more background on ways to do dynamic reconfiguration.
You have now seen how to issue talk 5 commands to actively query the status of the system. There is another, more passive mechanism available: viewing event messages that the Network Utility generates. To do this you use talk 2. As always, type Ctrl-p to leave the current process:
RTP01 + <ctrl-p>
RTP01 *
From the * prompt, type t 2 or event to attach your console to the process for viewing the Network Utility's local message log:
RTP01 * <Enter>
RTP01 *t 2
00:00:50 GW.001:
Copyright 1984 Massachusetts Institute of Technology,
Copyright 1989 The Regents of the University of California
00:00:50 GW.002: Portable CGW RTP01 Rel NetU-TX1 Feature 1001 V3.1 Mod 0 PTF 1
RPQ 0 MAS.DE1 netu_38PB
strtd
00:00:50 GW.005: Bffrs: 1000 avail 1000 idle fair 194 low 200
00:00:50 DOLOG: .....Remote Logging Facility is now available.....
In this example, only four messages have been logged since the Network Utility was last booted. Each message has the format:
All 4 of the above messages were logged in the same second, 50 seconds after the clock started.
GW.001, GW.002, and GW.005 are ELS messages in the GW (GateWay) subsystem. DOLOG is a nonstandard, unconditional type of message you will see from time to time.
The body of GW.001 is the two copyright statements. The body of GW.002 is the software version statement. To look up the meaning of any particular ELS message, see the Event Logging System Messages Guide either on the Web or in CD-ROM format.
Unlike the talk 6 and talk 5 processes, the talk 2 process has no user command prompt. That is because you do not type commands when you are in talk 2; you simply watch messages roll by as the Network Utility generates them. You control what messages appear by enabling or disabling individual or groups of messages under the event subprocess of either talk 6 or talk 5. See Monitoring Event Messages for an introduction to ELS concepts and controlling ELS messages.
Under talk 2, then, the only thing you would normally type is Ctrl-p, to return to the * prompt and move to talk 5 or talk 6. If messages are scrolling by too quickly to read, you can use Ctrl-s to pause scrolling, and Ctrl-q to resume it. Other options for capturing fast-moving event messages include:
These options are described in detail in the MAS Software User's Guide chapter "Using the Event Logging System (ELS)."
When you enter talk 2, the system displays all the messages that have been buffered up since the last time you left talk 2. If the message buffer has been overrun or the system is currently generating messages faster than it can display them, you will see lines about "messages flushed" interspersed within the talk 2 scrolling output.
If you are about to enter talk 2 and you know that there is a backlog of old messages to be displayed before you can see the current messages you are interested in, use the command flush 2 from the * prompt before typing talk 2. The system discards the entire backlog and talk 2 displays only messages generated after you entered the flush command.
Type Ctrl-p to exit talk 2 and return to the * prompt.
If you followed through the examples in this tutorial, you have made the following talk 6 configuration changes since you began:
Note: You also enabled the "fast-boot" option, but this change is stored in NVRAM and is not relevant here.
On a Network Utility, talk 6 changes are actually made in a RAM copy of the configuration. If you want these changes to become permanent and be used with the next reboot of the Network Utility, you need to write them to the hard disk. Two different command sequences can accomplish this task:
RTP01 *t 6
<Enter>
RTP01 Config>write
Config Save: Using bank A and config number 3
<boot messages start to appear>
RTP01 Config> <Ctrl-p>
RTP01 *reload
Are you sure you want to reload the gateway? (Yes or [No]): yes
<boot messages start to appear>
..... or .....
RTP01 *reload
Are you sure you want to reload the gateway? (Yes or [No]): yes
The configuration has been changed, save it? (Yes or [No] or Abort): yes
Config Save: Using bank A and config number 3
<boot messages start to appear>
In the first sequence, the user uses the write command to commit changes to disk before reload. In the second sequence, the user does not use the write command and the system asks whether to save changes to disk before proceeding with the reload.
Which method you use is completely up to you. Many users prefer the second method because it involves less thinking and typing, but it may also be easier to forget what talk 6 changes you have made if you do not issue a write shortly after making them.
Until now, the examples have always booted the Network Utility all the way up to the operational software, at either the Config (only)> or * prompt. There is one other major console user interface that you have not yet visited, that of the firmware. You may not need to interact much with the firmware, but you should be aware of it because it provides an alternative way to load code and configuration files onto the hard disk, and may give you a way out of a difficult problem.
Network Utility firmware is low-level software that drives the power-on and boot logic of the system. It resides in flash memory rather than on the hard disk, so in the event of a failure such as corruption of your operational software load on disk, you can retrieve new software or configuration files and get back up and running.
To reach the firmware user interface, your user console must be through local or dialed-in ASCII terminal emulation. You cannot telnet to the firmware user interface. To reach the main firmware menu, do a reload from the * prompt and look for these messages:
Starting Boot Sequence...
Strike F1 key now to prematurely terminate Boot
Look closely because these only appear for a few seconds each. Press F1 when prompted, or hold down Ctrl-c before and during the messages to interrupt the normal boot sequence and move into the firmware.
After you interrupt the boot sequence, the system may prompt you for a supervisory password before you can see the firmware main menu. This password controls access to sensitive low-level firmware functions. Its initial value from the factory is "2216". You can change it only from the firmware itself, under the Utilities menu.
If you dial into the Network Utility via modem to get your console and lose your connection on reload, you may not be able to connect back in time to press F1. In this case, go to the boot subsystem of the Config process and issue the disable auto-boot command:
*t 6 Gateway user configuration Config>boot Boot configuration Boot config>dis auto <----- short for "disable auto-boot" Select the duration to disable autoboot: (once, always) [always] once AutoBoot mode is now disabled once. Operation completed successfully. Boot config> <Ctrl-p> *rel y <----- short for "reload, yes" <boot messages appear>
With AutoBoot mode disabled, the system will stop the reload process at the firmware, without your having to press F1. Then when you connect back in, you will be at the main menu or the request for the supervisory password.
If you disable auto-boot always in talk 6 to reach the firmware or if you were not given the duration (once/always) prompt, remember to re-enable it when you reach the operational code, or you will stop in the firmware for every reload.
When you reach the firmware, you see a text menu like the one below at your user console:
Nways System Firmware
Version 3.00 built on 04/21/98 at 22:18:42 in cc3:paws_netu6e:cc3_6e
(C)Copyright IBM Corporation, 1996, 1998. All rights reserved.
System Management Services
Select one:
1. Manage Configuration
2. Boot Sequence Selection
3. Select Device to Test
4. Utilities
Enter - Esc=Quit - F1=Help - F3=Reboot - F9=Start OS
------------- ------------- ------------- ------------- ------------
The firmware menu structure and its options are described in the 2216 and Network Utility Service and Maintenance Manual in the chapter "Using 2216 Firmware." You do not type any commands, but move through a sequence of menus by selecting options. The key tasks you may need to perform from the firmware are:
These functions are equivalent to the boot subsystem functions under talk 6. You find them in the firmware menus under "Utilities," then "Change Management."
To do this, start with "Utilities" on the main menu, then "Update System Firmware."
You may want to move around the menus a little to get familiar with them. When you have completed any firmware task, press Esc to return to the main menu. Use one of the following options to continue:
F3=Reboot - starts the boot process all over. If you have auto-boot disabled, you will just stop in the firmware again. If you are dialed in, you will lose your connection again.
F9=Start OS - continues the boot process past the firmware up into the operational code.
You have reached the end of this Network Utility user interface tutorial. The following chapters cover a number of other important Network Utility concepts and methods, and assume you have the background provided in this chapter.