Chapter 8 - Applying Banyan Patches
This chapter explains how to apply Banyan patches on one or more servers using the Banyan Patch Utility. If you are applying a maintenance patch, also read Chapter 9.
A Banyan patch contains one or more software fixes and the files needed to apply the patch. Each patch also includes a release notice that explains what server platforms, products, and revisions the patch applies to, and what software problems you can fix by applying the patch. You can obtain patches on physical media from your support center or you can download them electronically from Banyan Online Access (BOA) and the Internet. Patches on physical media are available as one or more diskettes or, in some cases, CD-ROM.
You apply a Banyan software fix to a server by running the Patch Utility from a DOS workstation. This procedure is known as patching a server. When patching a server, you copy the patch files from one of the following sources:
![]() |
From diskettes inserted in the diskette drive of a DOS workstation |
![]() |
From a local or network fixed disk |
If you apply a patch from a local or network fixed disk, you can apply several patches at once. You can also apply patches from a fixed disk if you need to patch multiple servers-patching from a fixed disk is faster than patching from diskettes. For information on how to prepare a patch directory on a local or network fixed disk, see "Creating a Patch Directory Structure" later in this chapter.
Two types of Banyan patches exist:
![]() |
Site-specific patch |
![]() |
Maintenance patch |
Banyan releases a site-specific patch when a site has problems running Banyan software due to issues specific to their network configuration and environment. In general, site-specific patches contain a modified version of the program or programs being patched.
You can obtain site-specific patches electronically by downloading them from BOA or the Internet.
The name of a site-specific patch consists of these elements:
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.
For example, patch 96017A-VINES is the second version of patch 96017-VINES and is applicable to a VINES server.
Site patch identification numbers are issued sequentially and are not duplicated across product lines.
Patches are released in sequence and contain cumulative modifications to the base software on the server. When applying more than one patch that modifies the same area of software, you need to apply the most recent patch last to ensure that you obtain the most recent version of all software files.
The Banyan files on your server control various operations. Files may control kernel operations, StreetTalk operations, communications drivers, and so forth. When necessary to correct a problem, Banyan releases updated versions of these files in a site patch. As you apply a site patch to the server, you overwrite and replace old versions of some file with the new versions contained in the patch.
Site patches do not all contain the same files. Unlike a full release, a site patch contains a given file only if that file has been updated to provide new functionality or improve performance. In general, a site patch contains the files that are related to a particular area of the software, such as the kernel. For this reason, in two consecutive site patches each targeting a different area of the software, there may not be any common files.
However, when two site patches do contain common files, it is important to apply them in the correct order. The files in the later patch always contain not only the changes introduced in the earlier patch, but also any new changes. The files in the earlier patch, on the other hand, do not contain the modifications made in the later patch. When a later patch contains all of the changes from and provides all of the functionality of a previous patch, it is said to supersede that patch. If you apply the patches in the wrong order, you will overwrite the most recent version of the file and lose the functionality that the later patch provides.
A maintenance patch is cumulative and contains the software modifications made since the last base release. These include most of the modifications distributed in previous site-specific patches. A Banyan maintenance patch is actually a set of three individual patches, or components. Each component of the patch contains fixes for a specific subset of the software that resides on the Banyan server. Refer to Chapter 9 for more information about Banyan maintenance patches.
Preserving VINES Files Modifications
Patches overwrite files of the same name. If you have a program in your VINES Files volume that has the same name as a program on the patch, it will be overwritten. If you have modified your VINES Files volume, and do not want the patch to overwrite the modified files, copy them to another directory (or to diskette). Copy them back after you install the patch.
When you install or patch a client program, the files are always copied into the default VINES Files directories. If you moved a VINES Files program to a different directory and now want to replace it with the new patch file, apply the patch and then copy the new file to the directory where you want it.
If you are applying the patch to multiple servers, first patch all other servers on the network before you patch the following servers:
![]() |
Your StreetTalk server |
![]() |
The server where you obtain your drive Z VINES Files |
![]() |
Your routing server |
Patching your routing server causes it to become temporarily unavailable. Until the routing server becomes available, subsequent patch sessions fail because your workstation cannot access the network. Refer to "Choosing a Server to Patch" later in this chapter for information on verifying the status of a server you are patching.
Some patches reboot the server. If you are applying a patch that causes a reboot and plan to apply it to multiple servers, apply it first to servers that are the greatest number of hops away from the workstation from which the patch is being applied before applying it to servers that are fewer hops away.
If you plan to patch multiple servers or apply several patches and do not want to patch from diskettes, create a patch directory on a local or network fixed disk and copy the patches from the diskettes, as explained in the next section.
Creating a Patch Directory Structure
When you run the Patch Utility, it searches for a subdirectory named FIXES no more than two levels below the current working directory. If a FIXES subdirectory exists at the second level, the first-level subdirectory is considered a valid Patch directory.
The files for some patches are stored as diskette images in directories named PDISK1, PDISK2, and so forth. When applying a patch issued in this format, the Patch Utility skips the PDISK directory when calculating how many levels to search for a FIXES directory.
Depending on the format of the patch and the structure of your patch directory, you may need to run the Patch program from different levels to have access to certain patch files (Figure 8-1).
For example, if you run the Patch Utility from the PATCHES directory, it finds a second-level FIXES subdirectory only within the 98038-V directory. If you run the Patch Utility from the PATCHES directory, the directory 98038-V is therefore the only valid patch directory.
If you make the 851PATS directory the active directory before you run the Patch program, it finds a second-level FIXES subdirectory within the 39, 44, and 53 directories. These directories are therefore valid patch directories when you run the Patch program from the 851PATS directory.
Finally, because the format of patch 55 uses diskette images (PDISK1, PDISK2), if you run the Patch Utility from the 860PATS directory, it locates the FIXES subdirectories three levels below the active directory. 860PATS is therefore a valid patch directory when you run the Patch program from the 860PATS directory.
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.
Organize your patch directory according to how you usually apply patches. For instance, if your usual patch strategy is to apply fixes by type, consider maintaining a separate directory for patches that fix a certain part of the software. Thus, you would keep separate directories for mail service patches, STDA patches, and so on. Other possible directory structures might separate patches by:
![]() |
Server platform the patch applies to. For example, you could keep separate directories for patches that apply to different types of PC-based servers. |
![]() |
Software revision level that the patch is based on. For example, you could keep separate directories for 8.50 patches, 8.60 patches, and so on. |
![]() |
Patch range of the patch. Keep together those patches that you can apply during a single patch session. For information on patch ranges, refer to "Determining the Patch Range" later in this chapter. |
In addition, you can keep maintenance patches in a different directory than site-specific patches.
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
If you have access to the Internet, you can copy compressed patch files by using anonymous FTP to connect to a Banyan file server and download the files to your workstation or network fixed disk. You can then decompress them by using the DOS decompression utility, PKUNZIP.
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
Some patches are distributed with updated versions of the Patch Utility executable, PATCH.EXE. Check your patch documentation. If the patch requires a new version of the Patch Utility, before applying the patch, copy the new executable and its associated message file, PATCH.MSB, to the directory where you will execute the Patch Utility. That directory will be at most two levels above the FIXES directory. This ensures that you use the new executable. You must run the Patch Utility from within the appropriate directory; otherwise, it will not work properly. Do not remove the PATCH.EXE and PATCH.MSB files from this directory.
To avoid conflicts between different software revisions, each time you attempt to apply a patch, the Patch Utility automatically checks for compatibility between the server software and the patch file. You can apply a patch only if the revision level of the patch is compatible with the revision level of the server. The range of revision levels compatible with any patch is known as the patch range. During the patch process, the patch range for a patch appears on the Select Patch Names screen of the Patch program.
When you choose multiple patches during a single patch session, the Patch Utility checks that the patch range of each patch is consistent with the ranges of other patch names you have selected so far. If the Patch Utility detects an incompatibility between patches, an error message is displayed and you are prompted to choose another patch instead.
If you attempt to apply a patch to a server that is outside the appropriate patch range, the patch is not applied and a message is displayed informing you that the server is not in the patch range. For example, if a patch has a patch range of 6.00 (0) through 6.20 (0), you cannot apply it to a server that is at a revision level of 7.00 (0).
You can patch all servers on the network from a single location. Applying patches involves the following procedures:
![]() |
Specifying where the patch files exist |
![]() |
Choosing patches |
![]() |
Choosing servers |
![]() |
Deciding when to apply the patch |
![]() |
Copying the patch source files to the selected servers |
You can specify some options from the menus and screens of the Patch Utility. You can specify the following options on the command line: the server name, the path to the patch files directory, and the level of logging.
Once the files containing the software fixes are copied to a server, the patch process runs as a scheduled task.
Patching an Existing Patch Utility Executable
Occasionally, the documentation for a patch indicates that the patch contains an updated version of the Patch utility executable, PATCH.EXE. Before applying the patch, be sure to copy the file and its associated message file, PATCH.MSB, to the directory where you will execute PATCH. That directory will be two levels above the FIXES directory. This ensures that you use the new version of the executable.
Applying Patches from Diskettes
If you are applying the patch from diskettes, you can apply only one patch at a time. For patches consisting of multiple diskettes, the entire set of diskettes is copied to one server at a time before being copied to the next server selected. You insert the diskettes one at a time as prompted. After all files for the server's platform type are copied from the first diskette, you are prompted to insert the second diskette, and so on, until you have used all of the diskettes required for that server platform in numerical order. You must then re-insert the diskettes, in sequence for each subsequent server being patched. The diskette label contains the sequence number.
You may not be prompted to insert each diskette in a set. In a set of patch diskettes, some diskettes may contain files that do not apply to the platform type of the server being patched. The Patch Utility copies only the files that apply to a server's platform type.
A patch session involves applying one or more patches at a time to one or more servers.
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.
If you apply the patch from a workstation that runs Microsoft Windows, do not apply the patch from within a reduced DOS window. You can apply patches from the DOS prompt of a full-context DOS window.
![]() |
Decide which patch or patches to apply and which servers you want to patch. If you are applying patches from diskettes, you can apply only one patch at a time. If you are applying patches from diskette and would like to apply multiple patches, first copy the patches to a workstation or network fixed disk. Refer to "To Copy Patch Files from Diskettes to a Fixed Disk" earlier in this chapter. |
![]() |
Verify that you are on the AdminList@servername@Servers for each server that you plan to patch. Unless you are a member of the Admin List for a server, you will be unable to patch that server. |
![]() |
If you are applying multiple patches to a server, copy the patches to a local or network fixed disk. You may also want to copy patches to a fixed disk if you are patching multiple servers. Refer to "To Copy Patch Files from Diskettes to a Fixed Disk" earlier in this chapter to copy patch files to a patch directory on a network or local fixed disk. |
![]() |
If a recent complete backup does not exist for the server, create one now. Refer to Chapter 3 for information on how to back up a server. |
![]() |
Verify that the server has sufficient disk
space 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 type of server platform. The Patch
Utility copies only the version that applies to a server's platform
type. The amount of disk space available on the server varies as files are written to it during server activity. When you select a server to patch, the Patch Utility checks whether sufficient disk space to apply the patch is available at that moment. However, 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. |
You begin a patch session by entering the PATCH command from a DOS workstation. This section explains how to specify options on the PATCH command line and how to start the Patch Utility.
Entering Options on the Command Line
When you are ready to begin a patch session, you can specify certain information on the command line. The following command line options are available:
/P:drive:pathname - Specify a local or network drive and the pathname to the directory containing the patch source files. If the directory containing the patches is not on the active drive, you must specify the drive letter in addition to the pathname.
If you are copying the patches from the top-level directory of a diskette in drive A of your workstation and you are working in drive A, you do not have to specify the location of the patch source files. The Patch Utility searches within the current working directory and its subdirectories for the patch software.
You do not have to specify the path to the FIXES subdirectory. For example, if the patch files are stored in F:\PATCHES\700PATS\FIXES, you would enter:
patch Server1 /p:c:\patches\700PATS
/L:loglevel - Specify the level of logging that occurs in the PATCHLG file on drive C. The log for all patch activity is written to the file PATCHLG in the root directory of your workstation's drive C. If you apply patches from a workstation that has no local fixed disk, when the Patch Utility menu appears, it displays the following message:
Logging disabled, cannot open C:PATCHLG
You can change the location of the PATCHLG file.
Refer to "Setting Log Levels" later in this chapter for information about log levels.
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.
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.
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.
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.)
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.
If you do not enter the name of the server to patch at the command line, you can choose one or more servers from the Available Servers screen. You can only choose a server if you have administrative privileges for that server and the server's s revision level falls within the patch range for the selected patch or patches.
Selecting a server from the list of available servers, provides information about the server's platform (Table 8-1). You also learn whether the server has the Z drive or is the routing server for your workstation.
Designation | Type |
AT | ISA, EISA, and PCI bus servers |
MC | Micro Channel bus |
If you choose the server from which your workstation obtains its VINES Files (drive Z), a Z appears to the left of the platform type. If the server is your routing server, an R appears to the left of the platform type. For example, if server Admin1 were your VINES Files server and your routing server, the following would appear:
..Z..R..AT ... 6 - $Admin1
If you are applying the patch to multiple servers, patch all other servers on the network before you patch the following servers:
![]() |
Your StreetTalk server |
![]() |
The server where you obtain your drive Z VINES Files |
![]() |
Your routing server |
Patching your routing server causes it to become temporarily unavailable. Until the routing server becomes available, subsequent patch sessions fail because your workstation cannot access the network.
To Choose a Server to Patch
1. From the Available Servers screen, use the arrow keys or the PAGE UP and PAGE DOWN keys to scroll through the list of servers.
2. Press ENTER to select a server. You can select a maximum of 819 servers. Files are copied to each server in the order that you select them.
3. After you select a server, one of the following occurs:
- A dollar sign ($) 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
You can specify when to apply a patch and whether you want to notify users that the server is shutting down. After choosing which servers to patch, the Current Configuration screen appears (Figure 8-7). This screen contains the following information:
![]() |
The name of each server being patched. |
![]() |
The time that the patch is scheduled to run. By default all servers are scheduled to be patched ASAP. |
![]() |
The status of the 25th line warning messages sent to users when you patch the server. By default, warning messages are enabled. |
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.
After you disable 25th-line messages, they remain disabled until you enable them. When you apply subsequent patches to the same server, the Current Configuration screen indicates that messages are turned off.
To apply the patch at a later time, take into account the length of time required to copy the patch files to the server, especially if you are applying a large patch to several servers. The Patch Utility needs several minutes to copy the patch source files to the server. The actual time depends on the size of the patch, the number of servers being patched, and whether the files are being copied from diskettes or a fixed disk.
Set the time far enough ahead to avoid conflict between the copy operations and applying the patch. If you set a time and the patch files are not yet copied to the server when that time arrives, the patch is applied as soon as possible after the files are copied.
For example, if you schedule the patch to be applied in 15 minutes, but it takes 20 minutes to copy one patch to one server, the installation takes place as soon as possible after all files have been copied to the server.
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.
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.
After you press F10 to confirm the patch schedules, the Patch Utility automatically copies patch files to the first server selected. Patch files are copied to each server in the order in which you choose the servers from the Available Servers screen. (Refer to "Choosing a Server to Patch" earlier in this chapter).
The screen displays the total number of files being copied and incrementally displays the number of each file as it is copied:
Copying 247 files to server Servername
Copying file #18
If a patch requires a system file, the following message appears as it is copied to the server:
Copying system file to server Servername.
If a patch contains many files, copying the files may take several minutes.
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.
The Patch Utility copies one patch at a time to one server at a time. Thus, if you are patching multiple servers, all patches are copied, in order, to each server in turn, beginning with the first server selected. In other words, after the first patch is copied to the first server, the second patch is copied to the first server, and so on, until the first server has received all of the files for each patch.
Files are copied to a temporary directory on the server immediately after you confirm that you want to continue applying the patch. However the patch is not applied until the server clock reaches the time that you specified. If you chose to apply the patch ASAP, it is applied as soon as the files are copied.
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.
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
Do one of the following:
- Examine the PATCHLG file. If you did not change the default location of this file, it is automatically generated in the root directory of your workstation's drive C. For each patch that was successfully copied to the server, the log provides the name of the patch and the server. Refer to "Viewing the Patch Log" later in this chapter.
- Examine the patch history for each server. For each patch that was successfully applied, the patch history lists the name of the patch and the date and time that it was applied. Refer to "To View the Patch History of a Server" later in this chapter.
- Use PHIST. The PHIST command displays the patch history of a server. You can execute the PHIST command from the command line or from within a batch file. Refer to the Command Reference for further information on using PHIST.
Each time you initiate a new patch session, patch information is written to the PATCHLG file. The PATCHLG file contains information about all patch activities carried out from a particular workstation. Information is either appended to an existing file or, if no file exists, a new PATCHLG file is created.
By default, the file exists on drive C of your workstation. If you apply patches from a workstation that has no local fixed disk, when the Patch Utility menu appears, it displays the following message:
Logging disabled, cannot open C:PATCHLG
You can choose to have the log written to a network drive.
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.
You can set the level of logging that occurs during a patch session. When you set a log level, it is in effect for one complete patch session.
The following log levels are available:
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.
Log levels 2 through 5 contain the following types of messages:
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)
This example shows a patch log with entries made at the default log level (Level 2). The log entry for each patch session begins with the log level, the date and time, and the names of the routing server and drive Z server.
An explanation of the log contents follows the example.
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 completedClient 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-1Processing 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.
In the first session, the user selected Display patch history.
In the second session, the user (Gene Perry@Mktg@WCTUS) selected Install patches, but the directory specified did not contain any patches, that is, there was no FIXES directory.
In the third session, the user entered a pathname that contained a FIXES directory, and the Patch Utility found five patches. The Patch Utility displays the platforms to which the patches are applicable. The user selects patch 3, and the Patch Utility displays the revision range to which the patch is applicable and the location of the patch on drive O.
Next, the user selects server Server07 to apply the patch to. A warning appears, stating that this is the user's routing server. The patch is then applied.
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.
When entering PATCH, use the /L:n switch, where n is the log level you want to set (refer to "To Choose Patches to Apply" earlier in this chapter). Each log level provides more verbose comments than the previous level. Unless a support person requests you to use the higher levels for debugging purposes, Banyan recommends that you use Level 2.
You can find out which patches have been applied to a server by 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.
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.
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.
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
You can check a server to determine whether it is currently scheduled for patching. When you patch a server, the patches are first copied to the server and then applied. If you schedule the patch to occur ASAP, patching begins as soon as the files are copied to the server. If you schedule the patch to occur at some later time, patching begins at the time specified on the Set Time screen (refer to "Setting the Time and Notifying Users When You Apply the Patch" earlier in this chapter).
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.
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.
You can cancel patches that you scheduled to be applied at a later time. You cancel patches on one or more servers at a time. You cannot cancel individual patches from a set of patches being applied; you can only cancel all patches pending on a server.
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.
5. Do one of the following:
- To cancel the patches, press F5.
- To exit without canceling, press ESC.
To view the status of the patches for a server, look at the patch history for that server, which tells you that the patches pending for that server have been canceled. You can also display pending patches for that server to make sure that no patches are pending.