Previous PageNext Page

Chapter 8 - Applying Banyan Patches

Overview

Types of Banyan Patches

Site-Specific Patch

YYNNNR - PRODUCTNAME

YY is the last two digits of the year in which the patch is issued.

NNN is a three-digit identification number that indicates that this is the nth patch released for the given year.

R is a letter code that, if present, indicates that the patch has been recut and is omitted if this is the first patch. The first recut is A, the next B, and so forth.

PRODUCTNAME is the name of the product to which you can apply the patch.

Maintenance Patch

Preserving VINES Files Modifications

Patching Multiple Servers

Creating a Patch Directory Structure

Figure 8-1. Patch Directory Structure Showing Paths to Valid Patch Subdirectories

Caution: Do not copy the files for more than one patch into a single subdirectory. Files in a FIXES subdirectory have names in the form FILEx.0, where x is a number ranging from 1 through the number of files in the patch. Because the names are not unique to a particular patch, copying another patch to the same subdirectory would overwrite the files in the FIXES directory of another patch.

Organizing Patch Directories

To Create a Patch Directory on a Fixed Disk

1. On a workstation or network fixed disk, create a top-level directory, for example, one named PATCHES.

2. Within the PATCHES directory, create first-level subdirectories to organize the types of patches you want to store on the disk.

For example, create a directory 851PATS for 8.50 (1) patches and a directory 860PATS for 8.60 patches.

3. Within the directory you use to organize types of patches, create separate directories for each individual patch.

For example, create separate directories for patches 98039 and for 98044. Their pathnames might be C:\PATCHES\851PATS\39 and C:\PATCHES\851PATS\44.

You are now ready to copy patches from diskette or download them from a Banyan server. For information on how to download patches from the Banyan Internet server, refer to "Downloading Patches from the Banyan Internet Server," which follows.

Downloading Patches from the Banyan Internet Server

To Download Compressed Patch Files

1. Create a directory structure as in Figure 8-1. For example, download patch 98039 to the directory PATCHES\851PATS\39.

2. Make the subdirectory in PATCHES where you want to store the patch files the active directory on your workstation or network fixed disk.

3. At the command prompt, enter one of the following to connect to the Banyan Internet server:

ftp ftp.banyan.com

or

ftp 131.100.110.51

You are prompted to log in.

4. Enter:

anonymous

You are prompted for a password.

5. Enter your full Internet e-mail address as the password. For example:

admin@wctus.com

You connect to the Banyan Internet server.

6. Enter:

cd pub/patches

You connect to the patches subdirectory in the directory pub on the Banyan Internet server.

7. Enter:

ls

The list of patch subdirectories appears. Each directory contains patches for a particular Banyan revision or platform.

8. Connect to the appropriate directory.

9. Enter:

ls

The list of compressed patch files appears. Each file is in the form *.zip.

10. Depending on the FTP software you use, enter one of the following:

- binary

- type image

- prog

The following message appears:

Binary transfer is complete.

You are now ready to download the compressed files.

11. When you have located the file you want, enter one of the following:

- If you are copying a single file:

get filename.zip

- If you are copying multiple files:

mget filename1.zip filename2.zip filename3.zip

where filename is the name of the compressed file you are copying. The compressed file is copied to the current directory on your workstation or network fixed disk.

The length of time it takes to download the files varies and depends on the size of the file, the speed of your connection, and the number of hops between your Internet connection and the fixed disk to which you download the files.

12. When you have copied all the files you want, enter:

bye

You break the connection with the Banyan Internet server. You are now ready to decompress the files.

To Decompress the Compressed Files

1. Make the directory containing the compressed file or files the active directory.

2. If you do not have the program PKUNZIP.EXE in your PATH statement move or copy it into the current directory.

3. For each of the compressed files, enter:

PKUNZIP -d -o filename

where filename is the name of the compressed file. The contents of the file expand from their compressed format.

The directory now contains each of the patch files, one or more documentation files in ASCII format, the compressed file and the PKUNZIP program.

4. Delete the .ZIP files and the PKUNZIP program from the current directory or move them to another directory.

5. Read the documentation files using a text editor or print them by entering:

BPRINT filename

where filename is the name of the file you want to print.

You are now ready to apply the patch.

To Copy Patch Files from Diskettes to a Fixed Disk

1. Create a patch directory structure. Refer to "Creating a Patch Directory Structure" earlier in this chapter.

2. Insert the patch diskette in the diskette drive of the workstation.

3. Enter:

XCOPY A:*.* drive:\pathname /E /S /W

The switches (/E/S/W) tell XCOPY to copy directories and subdirectories even if they are empty and to verify each file as it is written.

For example, the command line might be:

xcopy A:\ *.* C:\PATCHES\700PATS\patchname /e /s /w

where patchname is the name of the directory where you want to store the patch files. Substitute the drive letter or path you are using for those shown in the preceding statement.

XCOPY copies the contents of the first diskette into the 700PATS subdirectory on drive C of your workstation. A FIXES directory is automatically created within the patchname subdirectory. The Patch program looks for the FIXES directory below the first-level subdirectory to indicate that the directory is a valid patch directory.

Note: Do not use COPY for this step. You must use XCOPY because it copies zero-length files. If you use COPY and the patch diskette has a zero-length file, that file is not copied, and the Patch Utility will not work properly.

The pathname for the FIXES directory in the 700PATS subdirectory is:

C:\PATCHES\700PATS\patchname\FIXES

4. Remove the diskette from the drive.

5. Repeat steps 2 through 4 for each diskette in the set.

6. Store the diskettes properly, in case you need them again. You are now ready to patch a server.

Updating a Patch Utility Executable

Determining the Patch Range

Applying Patches

Patching an Existing Patch Utility Executable

Applying Patches from Diskettes

Preparing for a Patch Session

Caution: When you apply multiple patches, the patches are applied in the order in which you choose them. Because patches may contain different versions of a given file, it is important to apply patches in the appropriate sequence. If a file on the patch has the same name as a file that already exists on the server, the version of the file on the patch overwrites the version on the server. The version of a file that exists on the last patch applied is the version that exists after all patches have been applied. Site-specific patches are generally applied in numerical order according to the year and three-digit identification number. For specific information about the sequence to use in applying a patch, read the documentation for each patch you apply.

Beginning a Patch Session

Entering Options on the Command Line

patch Server1 /p:c:\patches\700PATS

Logging disabled, cannot open C:PATCHLG

Caution: After you apply a patch, you cannot easily revert back to the server configuration that existed prior to applying the patch. Patches include modified versions of the executables provided on your release software. Executables are not backed up by a complete system backup. You must reinstall software and re-apply any patches to return the server to the state it was in prior to applying a patch.

To Begin a Patch Session

1. On the DOS workstation from which you apply the patch, make the drive that contains the patch files the active drive. For example, if the patch files are on your local fixed disk, enter:

C:

Note: If the patch supplies a new version of the patch executable (PATCH.EXE), you need to start the Patch Utility from the directory that contains the new file. Refer to "Updating a Patch Utility Executable" earlier in this chapter.

2. Use the DOS CD command to change to the appropriate patch directory on the drive.

For example, to change directories to one of the patch directories created earlier in this chapter, enter:

cd \patches\700PATS

3. Enter one of the following:

PATCH - To use the menus to indicate options.

PATCH [servername] [/P:drive:pathname] [/L:loglevel] - To indicate options on the command line.

Note: You cannot enter the name of more than one server on the command line.

The Patch Utility menu appears.

Figure 8-2. Patch Utility Menu Screen

4. Depending on whether you want the PATCHLG file written to the default directory on the root of drive C, do one of the following:

- Press F5 if you want the log file written to a different directory. The Enter Pathname screen appears. Continue with step 5.

- To use the default directory for the log file (root of drive C), refer to "To Choose Patches to Apply," which follows.

Figure 8-3. Enter Path Name Screen

5. Enter the pathname to the directory where you would like to maintain the PATCHLG file. The path may be to a directory on a workstation or network drive.

To Choose Patches to Apply

1. From Patch Utility menu, choose Install Patches. One of the following occurs:

- If the current directory contains patches, the Select Patch Names screen appears, listing the available patches. Skip to step 3.

- If the current directory does not contain any patches, the Enter Pathname screen appears and displays the pathname to the current directory.

- If you specified a servername on the DOS command line and it cannot be reached, the following message appears:

STK1037 : Appropriate StreetTalk is unavailable.

The server may not be in operation or a problem may exist on the network. Check the server and try again later, or select another server to patch.

2. Enter the pathname to the directory that contains the patch source files. The path may be to a directory on a workstation fixed disk or diskette drive or to a network fixed disk. For example:

C:\PATCHES\600PATS

3. One of the following occurs:

- The Select Patch Names screen appears, listing the available patches in the current directory. Continue with step 4.

- If you do not enter the path to a valid patch directory, the following message appears:

Patch directory(ies) not found.

Verify that you entered the correct pathname. (Refer to "Creating a Patch Directory Structure" earlier in this chapter for information on identifying a valid patch directory.)

Figure 8-4. Select Patch Name Screen

4. The Select Patch Names screen lists the available patches in the current directory. Do one of the following:

- If the patches do not appear on the list, press F5. The Enter Pathname screen appears. Enter a new pathname (see step 2).

- If the patches appear on the list, choose one or more patches to apply.

5. One of the following occurs:

- If you select a single patch and it is within the patch range of a server selected in step 1, an asterisk (*) appears beside the name of the patch or patches and the patch range appears on the bottom of the screen. Continue with step 6.

- If you select multiple patches and each patch you select possesses an equivalent patch range consistent with the revision level of a server selected in step 1, an asterisk (*) appears beside the name of the patch or patches and the patch range appears on the bottom of the screen. Continue with step 6.

- If you select multiple patches and the patch range for any one patch differs from the range of a previously selected patch, the following message appears:

Patch is out of the selected patch ranges.

You can apply multiple patches at one time only if all the patches fall within the same patch range. Choose a different set of patches or cancel the patch operation.

- If you selected a server in step 1 and the server revision does not fall within the patch range, a message similar to the following appears:

Server Admin1 at revision 7.00 (0) not in patch range.

You can patch a server only if the server's revision level is within the patch range of the patch. Choose a different set of patches or cancel the patch operation.

6. Press F4 to see a brief description of the patch. This information includes special factors to consider in applying the patch, a list of the fixes contained in the patch, and the amount of space required to apply the patch.

Note: This information is not available for all patches.

7. Press F10 after you have finished choosing the patches to apply. One of the following occurs:

- If you entered the name of the server on the command line, the Current Configuration screen appears. To continue the patch application, refer to "Setting the Time and Notifying Users When You Apply the Patch" later in this chapter.

- If you did not enter the name of the server on the command line, the Available Servers screen appears. You are now ready to choose a server to patch. To continue the patch application, see "Choosing a Server to Patch," which follows.

Choosing a Server to Patch

Table 8-1. Types of Server Platforms

Designation Type
AT ISA, EISA, and PCI bus servers
MC Micro Channel bus

..Z..R..AT ... 6 - $Admin1

To Choose a Server to Patch

1. From the Available Servers screen, use the arrow keys or the PAGE UP and PAGE DOWN keys to scroll through the list of servers.

Figure 8-5. Available Servers Screen

2. Press ENTER to select a server. You can select a maximum of 819 servers. Files are copied to each server in the order that you select them.

Figure 8-6. Selecting a Server from the Available Servers Screen

3. After you select a server, one of the following occurs:

- A dollar sign ($) and the server's platform type appear beside the server's name. Pressing ENTER again de-selects the server. Continue with
step 4.

- If you select a server and are not an administrator for that server, the following message appears:

You are not listed on AdminList@Servername@servers

Contact your system administrator to acquire the appropriate administrative privileges.

- If you select a server that is outside the appropriate patch range, a message similar to the following appears:

Server Servername at revision 6.00 (0) not in patch range.

Verify that you are applying the correct patch or select another server.

- If you select a server that does not have enough space on disk1 to apply the patch, the following message appears:

This patch requires at least n bytes of server disk space.

where n is the total number of bytes required to apply the patch. Record this number and refer to the procedure, "To Increase the Disk Space Available for the Patch," which follows.

Note: If you schedule a patch to begin at a later time, files written to the disk after checking the disk space may leave insufficient space to apply the patch.

4. Press F10 when you are ready to continue. The Current Configuration screen appears. To continue with the patch application, refer to "Setting the Time and Notifying Users When You Apply the Patch" later in this chapter.

To Increase the Disk Space Available for the Patch

1. Use StreetTalk Explorer or OPERATE to check the amount of disk space available on the server.

2. Do one of the following:

- If you do not know how much space the patches require, from the DOS command line enter:

PATCH

The Patch Utility menu appears.

- If you already know how much space the patches require, skip to step 7.

3. Choose Install Patches. The Select Patch Names screen appears.

4. Use the arrow keys to highlight the patch you want to apply.

5. Press F10. The Available Servers screen appears.

6. Choose a server from the list.

7. Press F4. The screen displays the amount of server disk space required to apply the patch. The amount of space required for the patch may vary among different types of servers. Some files in the patch are for specific types of servers and exist in several versions, one version for each of the four types of server platform. The Patch Utility copies only the version that applies to a server's platform type.

8. Press ESC four times to exit the Patch Utility.

9. Examine the files services residing on disk1 of the server and identify files that may be removed.

10. Delete or move to a different disk enough files to free the amount of space required.

11. Run the Patch Utility again.

Setting the Time and Notifying Users When You Apply the Patch

Note: If the server is running a revision of Banyan software prior to version 6.00 the option to disable or enable 25th-line messages is not available. If you applied a previous patch that disables the server's 25-line messages, an N appears on the Current Configuration screen.

Figure 8-7. Current Configuration Screen

Caution: Avoid scheduling a patch so that it executes when the mail janitor runs or STDA is rebuilding.

To Set the Time and Notify Users

1. Examine the list of servers on the Current Configuration screen.

2. Do one of the following:

- To schedule each server to be patched at a specific time, use the arrow keys to select a server. You change the schedule individually for each server. Continue with step 3.

- To apply the patches as soon as possible to all servers, skip to step 9.

3. Press F4. The Current Time screen appears and displays the current time. You are prompted to enter a time and to decide whether to enable 25th line messages.

Figure 8-8. Changing the Current Configuration Screen

4. Enter the time when you want to patch the selected server. Enter the new time in HH:MM format.

For example, if you want a patch to be installed at 7PM, enter 19:00. If you make an error when entering the time, use the BACKSPACE key to delete what you have entered.

5. Press ENTER. The cursor moves to the next line.

6. To send a 25th line message to users, warning them that the server will be shutting down, do one of the following:

- Enter Y to send a 25th line message.

- Enter N if you do not want to send a 25th line message.

7. Press ENTER. The Current Configuration screen appears again. The time that you entered appears beside the name of the server. The patch application will begin at the time shown.

8. Repeat steps 1 through 6 for each server that you want to schedule. Servers that you do not schedule are patched as soon as the patches are copied to the server.

9. To change the setting for sending a 25th line message that warns users that the server is shutting down, do one of the following:

- Press F7 to enable 25th line messages. As each server shuts down, users receive a 25th line message.

- Press F6 to disable 25th line messages. For each server on the list, the letter N appears below the column "25th Line Enabled." As each server shuts down, users do not receive a warning message.

10. After configuring the options for each server, do one of the following:

- To continue with the patch process, press F10. If servers will reboot after the patch has been applied, a message appears, warning you that the patch will reboot the server. You are prompted to continue.

- To cancel the patches, press ESC. The Select Patch Names screen appears. Select a new patch or press ESC to view the Patch Utility menu.

11. Do one of the following:

- Press F10 to continue. See "Copying Patches to the Server," which follows.

- Press ESC to cancel the patch. The Patch Utility menu appears.

You can cancel a patch up until the time that it is actually applied to the server. Refer to "To Cancel Scheduled Patches" later in this chapter for instructions on how to cancel a patch.

Copying Patches to the Server

Copying 247 files to server Servername
Copying file #18

Copying system file to server Servername.

Note: You cannot use the Patch Utility while it is copying files to a server. The copy procedure preempts any other patch activity on the server until the copying is finished.

Note: You cannot use your DOS workstation for other tasks until all files are copied to the servers selected. If you are running Windows and switch to another task, the copying continues.

To Copy File Status Information to a File

1. After all files have been copied, the File Copy Status screen appears and displays the status of the patch operation.

Figure 8-9. File Copy Status Screen

2. Press F4 to write the status information to a disk file. You are prompted to enter a filename to contain the status information.

3. Enter the full pathname for the file in which you want to store the status information. The information from the File Copy Status screen is copied to the specified file.

4. Press ESC. The Patch Utility menu appears.

To Verify the Patch Application

Viewing the Patch Log

Logging disabled, cannot open C:PATCHLG

Note: Each time you begin a patch session, log information is appended to the PATCHLG file. Periodically delete or edit the PATCHLG file to prevent it from using too much space on your hard drive.

Setting Log Levels

Level 1 - No logging. Specify this level if you do not want logging to occur. Use Level 1 if you do not have a C drive.

Level 2 - This is the default level. It includes standard display information with some additional error reporting. Refer to the example that follows this section.

Level 3 - Full logging with function calls and their arguments, and additional error reporting.

Level 4 - Same as Level 3 but in a more secure mode. Messages written to the patch log file buffer are immediately written to disk.

Level 5 - Same as Level 4 but more verbose.

Informational - Contains data that may be displayed on the screen, the patch type being processed, and other information applicable to the patch session.

Notice - Contains information relative to an operation that will not be performed, for example, the particular patch is not applicable to the server type.

Error - Contains a full description of an encountered problem.

Warning - Contains information on unexpected results that may occur if you continue to process the patch after a displayed condition. For example, the server does not have enough disk space or the server is your routing server and will be patched before other servers on the list.

Example PATCHLG Showing Default Level Logging (Level 2)

Client patch utility L2 log session. December 30 1994 9:57:28

Routing Server: Server7 Drive Z Server: Server5.

Processing Name History command...
Name History completed

Client patch utility L2 log session. December 30 1994 10:10:34

Routing Server: Server7 Drive Z Server: Server5.

Starting Patch Application Process

User: Gene Perry@Mktg@WCTUS

Searching for patches

Searching for patches

Searching for patches

NOTICE: Did not or Could not find a patch directory

Patch application incomplete.

Client patch utility L1 log session. December 30 1993 10:15:41

Routing Server: Server7 Drive Z Server: Server5.

Starting Patch Application Process

User: Gene Perry@Mktg@WCTUS

Searching for patches

Searching for patches

5 Patch(es) found:

--- #1 of 5: 4.11(5) Site-Specific Patch 4.11-GP-1

--- #2 of 5: 5.00(5) Site-Specific Patch 5.00-CN-1

--- #3 of 5: 5.52(5) Site-Specific Patch 5.52(5)-VNS-GM-1

--- #4 of 5: CATEGORY=ANS Vines Assistant Ver 2.04A

--- #5 of 5: History 4.10(5) Site-Specific Patch 4.10-GY-1

Patch #3: 5.52(5) Site-Specific Patch 5.52(5)-VNS-GM-1 Was Selected.

Applicable revision range: 5.50(1)-5.57(4)
Patch Location: O:\PATS\GM-1

Processing Server Selections

Validating selection of server: Server7.

WARNING: Server7 is your Routing server

Applying 1 of 1 selected patches on server: Server7

Applying patch 5.52(5) Site-Specific Patch 5.52(5)-VNS-GM-1 to server ...Server7.. Path: O:\PATS\GM-1

Patch application complete.

Processing Name History command...Validating selection of server: Server7.

Name History completedProcessing Full History command...

Validating selection of server: Server7.

Note: The patch log contains entries for all patch activity that occurs from a given workstation. You can also use PHIST to display the full patch history of a server. Refer to the Command Reference for more information.

To Set the Log Level

Viewing the Patch History

To View the Patch History of a Server

1. From the DOS prompt of your workstation, do one of the following:

- To view the patch history for a particular server, enter:

PATCH servername

where servername is the name of the server for which you want to view the patch history. Continue with step 2.

- To view the patch history of multiple servers, enter:

PATCH

The Patch Utility menu appears.

Figure 8-10. Patch Utility Menu Screen

2. Choose Display Patch History. Depending on whether you specified a server name in step 1, one of the following occurs:

- If you specified a name, the Abbreviated Patch History screen appears. The abbreviated patch history provides the name of each patch applied to the specified server, and indicates when it was applied. Skip to step 5.

- If you did not specify a server name, the Available Servers screen appears.

Figure 8-11. Abbreviated Patch History Screen

3. Select one or more servers from the list of available servers. A dollar sign ($) appears beside the name of each server you choose.

4. Press F10 to view the patch history of the first server selected. The Abbreviated Patch History appears. The abbreviated history lists each patch applied to the server and reports when it was applied.

5. Press F5 to view the Full Patch History. The Full Patch History screen appears. Pressing F5 toggles the view between the full history and abbreviated history.

Figure 8-12. Full Patch History Screen

The full history provides details on the contents of each patch applied to the specified server, including the information from the Abbreviated Patch History. The Full Patch History may not differ from the Abbreviated Patch History.

If the screen cannot display the entire history file, the following message appears:

History file too large to display entirely

If the full history is too large to view on-screen, you can view it by writing it to a file (see step 6).

6. Press F4 to write the patch history to a file. You are prompted to enter the name of a file to write the data to.

7. Enter the pathname to your workstation or a network file. The currently displayed patch history (full or abbreviated) is written to the file you specify.

8. Press F10 if you selected multiple servers in step 3 and want to view the patch history of another server. The history for each server appears in the order in which you selected the servers.

9. After viewing a history, press F10 again to view the next history. After you have viewed the history of the last server selected, the history of the first server appears again.

10. Press ESC to return to the Display Patch History menu.

Viewing the Patch Status of a Server

To Find Out if a Server Is Scheduled for Patching

1. From the DOS prompt of your workstation, do one of the following:

- To view the patch history for a particular server, enter:

PATCH servername

where servername is the name of the server for which you want to view the patch history.

- To view the patch history of multiple servers, enter:

PATCH

The Patch Utility menu appears.

2. Choose Display patch status. One of the following occurs:

- If you specified a server name in step 1, the Patch Status screen of the specified server appears. Skip to Step 5.

- If you did not specify a server name in step 1, the Available Servers screen appears.

3. Press ENTER to choose a name from the list of available servers. You can choose more than one server. A dollar sign ($) appears beside the name of each server you choose.

4. Press F10 to view the patch status of the first server selected. The Patch Status screen appears and displays information about patches that are waiting to be applied to the server.

Figure 8-13. Patch Status Screen

5. Do one of the following:

- If you selected multiple servers in step 3, press F10 to view the patch status of the next server.

- Press ESC to return to the Patch Utility menu.

Canceling Pending Patches

To Cancel Scheduled Patches

1. Depending on whether you want to cancel patches on one or more servers, enter one of the following:

PATCH servername

where servername is the name of a single server where you scheduled one or more patches.

PATCH

To cancel patches on several servers.

The Patch Utility menu appears.

2. Choose Cancel Pending Patches. One of the following occurs:

- If you specified a server name, the Patch Status screen appears. Skip to step 5.

- If you did not specify the server name, the Available Servers screen appears.

3. Choose the names of any servers on which you want to cancel a scheduled patch. You may select more than one server. After you select a server, dollar sign ($) appears next to your selection.

4. Press F10 after choosing servers from the list. The Patch Status screen appears. Depending on the patch status of the server, one of the following appears:

- If no patches are scheduled, the following message appears:

No patches are pending on this server.

- If patches are scheduled, a message similar to following appears:

Patches to be applied on Fri Feb 17 05:01:08 1995
6.00(0)-VNS-AA-1

- If the server is currently being patched, a message similar to following appears:

Patches currently being applied
6.00(0)-VNS-AA-1

- If the server cannot be found, the following message appears:

STK1037: Appropriate StreetTalk is unavailable.

Note: StreetTalk may be unavailable for reasons other than a patch being currently applied.

Figure 8-14. Canceling a Pending Patch Screen

5. Do one of the following:

- To cancel the patches, press F5.

- To exit without canceling, press ESC.

Previous PageTop Of PageNext Page