![[Return to Library]](BOOKSHELF.GIF){kind=small}
![[Contents]](TOC.GIF){kind=small}
![[Previous Chapter]](http://www.tru64unix.compaq.com/docs/base_doc/DOCUMENTATION/HTML/AA-QTLMB-TE_html/REW.GIF)
![[Next Section]](http://www.tru64unix.compaq.com/docs/base_doc/DOCUMENTATION/HTML/AA-QTLMB-TE_html/NEXT.GIF)
![[Next Chapter]](FF.GIF){kind=small}
![[Index]](INDEX.GIF){kind=small}
![[Help]](HELP.GIF){kind=small}
This chapter contains notes about issues and known problems with the base operating system and, whenever possible, provides solutions or workarounds to those problems.
The following topics are discussed:
The following notes apply to restrictions on using functions that support internationalization or internationalized components.
![[Previous Section]](http://www.tru64unix.compaq.com/docs/base_doc/DOCUMENTATION/HTML/AA-QTLMB-TE_html/PREV.GIF)
The return values of the functions iswdigit() and iswalnum() for Thai digits in the range 0xF0,...,0xF9 are false, even though these values are defined as digits in the Wototo Standard, Version 2.0 relating to the Thai language.
If two or more print jobs are sent to different queues of the same printer within a very short time, it is possible that some jobs will get blocked and cannot be printed. Restart the job using the lpc command.
Depending on the composition of the line, the vi editor that supports Thai may wrap lines before the right boundary of the screen. For a normal 24x80 screen, a line wraps if more than 80 Thai or ASCII characters are entered, even when the display width of the line is fewer than 80 columns.
Digital UNIX version 4.0B contains a new curses implementation that incorporates the following sets of programming interfaces:
These interfaces support the new curses standard and also maintain source compatibility with the curses interfaces found in the previous version of Digital UNIX. However, they do not maintain binary compatibility. Therefore, additional library files are also provided to maintain binary compatibility with existing programs
The new curses implementation can be found in the following:
| Directory | Contents |
| /usr/include | curses.h, curshdr.h, term.h, unctrl.h |
| /usr/lib | libcurses.a |
| /usr/shlib | libcurses.so |
Libraries supporting the previous curses implementation will be found in the following directories:
| Directory | Contents |
| /usr/opt/lib | libcurses.a |
| /usr/shlib/osf.1 | libcurses.so |
There are no header files provided for the old implementation. All programs that call curses directly will use the new header files when compiling and linking with the new libraries.
Existing binary files that were built with the old version of the curses shared library, libcurses.so, will continue to run without modification. Through the technique called versioning old binary files will select the correct shared library.
When rebuilding applications that do not call curses directly, but which link with libraries that interface with the old curses implementation, set LD_LIBRARY_PATH to the proper directory as indicated in the preceding table, or else specify the complete pathname on the compile command.
Many functions declared in curses.h and term.h are implemented both as routines and macros. In these cases, the default is to compile them as macros which execute more quickly. If you wish to execute the corresponding routines instead of using the macros, add the _NOMACROS compile-time switch.
The new curses library provides source compatibility with the old version with respect to function calls, data names, and macro names. However, the implementation of types, variables, structures and macros has changed considerably in some cases. If the existing user code takes advantage of the assumed implementation details of such elements, that program may need to be changed to work with the new curses. Elements whose implementation have changed include:
In previous implementations, the ACS definitions in curses.h were statically defined constant values. In the new implementation, the definitions are dynamically defined at run time based on the current setting of the TERM environment variable. This implementation allows terminals with extra ACS capabilities to make them available to the user while providing a set of default ACS definitions for terminals with lesser capabilities.
The implementation changes may cause compile time failures for some programs that depend on the static definitions. For example, the following declaration will not compile when it occurs at the global level:
char A = ACS_ULCORNER;
The new implementation of the curses library defines the ACS definitions at run time and requires that all assignments be made after the initscr() function has been called.
The X/Open Curses standard requires that the cbreak() function disable the ICRNL input processing flag. In the previous Digital UNIX implementation, cbreak() did not disable this flag. In applications that relied on this default behavior to advance to new lines, subsequent output lines may now overwrite the last line addressed. Those applications should now set the ICRNL flag explicitly after the call to cbreak(). Here is a sample code fragment that sets the ICRNL flag:
#include <termios.h>
struct termios tty;
tcgetattr(0, &tty); tty.c_iflag |= ICRNL; tcsetattr(0, TCSANOW, &tty);
The following notes apply to commands and utilities.
A conflict exists between the Dataless Management Services (DMS) and the System V Environment (SVE) which causes a failure when attempting to add a client on a server. This note only applies when the server is running SVE.
The tar utility which is shipped as part of the System V Environment, does not support all of the features supplied by the Digital UNIX tar utility. The SVE tar utility is installed into the /usr/sbin directory. As a result of the directory path set in the dmu utility, the dmu utility attempts to use the SVE tar utility, instead of the Digital UNIX tar utility.
To resolve this problem, you must edit lines 719, 724, 1121, 1128, and 1131 in the /usr/sbin/dmu script to include the full path name of the Digital UNIX tar utility. Change the occurance of tar to /sbin/tar.
Copying files to stdout using the dump command is not supported in this release.
The make command will not recogize escaped comment symbols as literal characters in a Makefile. Comment lines that begin with a # (number sign) and all text following this symbol up to the end of the line are considered part of a comment. This is also true even if the symbol is preceeded with a backslash (\).
The following notes describe problems that may occur when using commands and utilities under certain security settings.
Programs cannot reliably inspect the permission bits in the stat structure and determine the access that will be granted to a particular user. On local filesystems, read-only mounts and ACLs can both modify the access that will be allowed, On remote filesystems, in addition to read-only mounts and ACLs, there may also be additional controls that can alter the permitted access such as:
Programs which copy files to update them, rather than updating them in place, will often not preserve ACLs. Some programs that have this problem are gzip, compress, and emacs.
The best solution for programs that need to make access decisions is for the program to use the access() call to determine what access will be granted. Note that even this may not work as the access protections of the file could be changed between the access() call and the read, write, or execute operation.
For programs that copy files, the following command will copy a file preserving ACLs and any other extended attribute (property list ):
#
cp -p
See the acl(4), and proplist(4) reference pages for more information.
This note is to clarify the interactions between the archive tools pax, tar, and cpio, and files containing property lists or Access Control Lists (ACLs). or ACLs alone.
When you extract files with the above utilities without using the -p option, and the following conditions apply:
Then, when files that do not have an associated ACL them are extracted, they will inherit the default ACL that is in force on the directory in which they are being created. This behavior was selected to allow file extractions to work as expected in as many cases as possible.
For times when the above behavior is not appropriate, an alternative behavior has been associated with the -p option. If you use the -p option, any file that is extracted from the archive will be given only the property list, ACL, and file permissions that were stored in the archive with that file. For tar archives you can use either the -p option to the tar command or the -p p option to the pax command. For cpio archives you can use the -p p option to the cpio command
See the tar(1), pax(1), cpio(1), and tar(4) reference pages.
By default, emacs will rename the original file, and save the new file as a copy, under the original name. If the original file had an Access Control List (ACL) it will now apply to the backup file. If the directory had a default ACL, the new file (original filename) will now have the default ACL instead of the original ACL. If the directory did not have a default ACL, the new file will be protected only by the file permission bits.
The emacs utility has some user-preference variables which can be set to control which file will retain the original ACL. The relevant EMACS variables are:
The gendisk utility is used to create product media. There is a problem in using it on FDI diskette devices, the diskette drive found on all non-Turbochannel bus Alpha platforms.
The solution involves making some hard links to the diskette device special files using the name of the device that gendisk will use.
#
cd /dev
#
ln rfd0c rfl0c
#
ln rfd0a rfl0a
#
ln fd0a fl0a
#
ln fd0c fl0c
#
fddisk -fmt /dev/rfd0c
You will see the following messages:
NOTE: Setting interleave factor to ``-i2:4''. Use ``-i<nnn>[:<ccc>]'' option to override. Disk type: 3.50 inch, HD (1.44MB) Number of sectors per track: 18 Number of surfaces: 2 Number of cylinders: 80 Sector size: 512 interleave factor: 2:4 Formatting disk... Percentage complete: Format complete, checking... Quick check of disk passes OK.
#
disklabel -wr fd0 rx23
Note
When running the gendisk utility on the diskette using these instructions, do not respond yes to the question asking to clean the disk.
The following is an example of a gendisk command session:
#
gendisk -d MYPRODUCT410 /dev/rfd0c
Generating MYPRODUCT410 Kit from <system address> on /dev/fl0c
WARNING: this will remove any information stored in /dev/fl0c. Are you sure you want to do this? (y/n): y
Do you want to clean the entire disk first? Note: This will replace your current disk label with a default one. (y/n) [n]: n
Preparing /dev/fl0c (floppy) done.
Checking /dev/fl0c /sbin/ufs_fsck /dev/rfl0c ** /dev/rfl0c File system unmounted cleanly - no fsck needed
Mounting /dev/fl0c on /usr/tmp/cd_mnt8344
Writing Images (dd=/).
Image instctrl...done. Image SVGASTATIC100...done.
Verifying Images (dd=/).
Image instctrl...done. Image SVGASTATIC100...done.
Kit MYPRODUCT410 done.
Cleaning up working directories. Unmounting /dev/fl0c
Digital ships the Emacs software as it is received from the source. The following command line options do not work as documented in the emacs(1) reference page:
| -geometry | -iconic |
In some cases a solution is available using an appropriate X resource.
The /usr/bin/od command has the following restrictions:
The following notes apply to restrictions on using the SysMan applications.
The following notes apply to the account manager, dxaccounts.
When copying user accounts via cut and paste or drag and drop, the Allow Duplicate UIDs option in the General Preferences dialog box will be honored. For example, when making a copy of user account that has a UID of 200, if the Allow Duplicate UID's check box is off (the default), the resulting copy will have a unique UID automatically generated. If the Allow Duplicate UID's check box is on, then the copy will have an indentical UID. The same rules apply to copying groups.
The account manager has the following restrictions on both base security and enhanced security (C2) systems:
Workaround: Delete the original icon after the copy has been completed.
Workaround: Set a starting value within the range using the usermod or groupmod commands:
usermod -D -x next_uid=xxx
usermod -D -x next_gid=xxx
For example, if the Minimum UID is 100 and the Maximum UID is 10000 then:
usermod -D -x next_uid=5000
causes account manager to start generating UIDs from 5000.
Workaround: Use the chown command to change the directory and files, if applicable.
Workaround: Use the copy/paste feature to copy users, groups, or templates from account manager A to B.
Workarounds: Account manager correctly allows two or more system administrators to work on the same password files simultaneously. The proper file locking will occur and new accounts can be added or modified. However, the local groups file, /etc/group, and the NIS groups file, /var/yp/src/group, are written out after each group modification. Therefore, the last system administrator to make a change in a groups view window would overwite the any prior changes from a differnt system administrator. For this reason, running multiple, concurrent account manager instances is not recommended.
Warning: DtComboBoxWidget: Unable to find item to select
Workaround: None. These messages can be safely ignored.
Leading and trailing white space is not stripped from text entry areas. This could lead to confusion, for example, if a field on the Find dialog contains a space character before the desired search string. The search string would not match becuase of the spurious space character.
The following problems apply to the account manager when running on enhanced security systems.
Workaround: Change the template lock setting on the Create/Modify Template dialog screen after selecting the template by double clicking on the template icon in the Template view icon box.
Workaround: Set passwords through /usr/tcb/bin/dxchpwd or the /usr/bin/passwd command when the C1Crypt Encryption type is chosen.
Workaround: Set the C1Crypt Encryption type for the user from the Create/Modify User dialog.
Workaround: Set passwords through /usr/tcb/bin/dxchpwd or the /usr/bin/passwd command if the Minimum/Maximum password length limitation is necessary.
Workaround: To delete a user account you must do the following:
#
/usr/tcb/bin/edauth -r <user name>
Workaround: Use the following command to remove the dangling protected password database entry:
#
/usr/tcb/bin/edauth -r <user name>
Workaround: Restart the account manager to restore the former template icon. Delete the undesired template using the Delete Toolbar icon or the Edit->Delete... option from the Template view.
Workaround: Modify the copied user and change his template from default to the desired template. Note that the template reference is maintained if the user is dropped within the same view.
Workaround: Only the drag and drop method of template assignment has this problem. You can use the Create/Modify dialog box to change a single user's template or use the Modify Selected dialog box to change templates for several selected users. Both methods will correctly propagate the template's lock field.
Workaround: None.
Workaround: Please restart account manager and then delete the template.
Workaround: Manually remake the NIS maps or perform an account manager function (eg. Account Modification) that will trigger the maps to be remade. To manually remake the maps do the following:
#
cd /var/yp
#
make all
The dxshutdown application does not create the /etc/nologin files as described in the documentation. This means that users will be able to login to a machine that is being shutdown up until the actual time of the shutdown.
Note, this behavior differs from that of the shutdown command which creates the /etc/nologin file at 5 minutes prior to the shutdown.
The Print Configuration Manager may have some problems with /etc/printcap files from DEC OSF/1 Version 3.2 or earlier, as follows
Using /etc/printcap files in the current version of Digital UNIX, the system assigns printers names lp[0-9]*, [0-9]*, and for the default printer, lp. For example, the default printer may have a name field such as lp0|0|lp|default|declaser3500:.... Another printer may be named lp7|7|some_alias|another alias:.... Therefore, the system has difficulty with printers that have less than two names or that use these reserved names as aliases.
Some of the attribute value checking is different between earlier versions and the current version. For example, some fields that were not required now are, and some attributes values that were legal no longer are.
The Print Configuration Manager requires that all comments be associated with a printer. As a result, comments appearing after the last printer are truncated.
To avoid these problems, invoke the printconfig utility with the menu interface (printconfig -ui menu). This brings up the lprsetup utility which is fully compatible with earlier printcap files.
The following notes apply to system administration.
Unattended installation and modem console will be supported only on the following platforms for Digital UNIX Version 4.0B.
| Platform | Required Firmware Revision |
| ALPHASERVER 1000A | 4.7 (from 3.8 firmware cd) |
| ALPHASERVER 4100 | 3.0 (field test only) |
| ALPHASERVER 2100 | 4.6-201 (from 3.8 firmware cd) |
For this release, bootable tape will not work with the LSM product. Not all platforms and tape drives support bootable tape. The following processor platforms are supported:
The following tape devices are supported:
In order to use the btcreate utility, your system must have at least 156,000 512-byte blocks of free space in the /usr directory.
You will not have enough space if your system uses an RZ26 or smaller disk with the default partitions and you have installed all of the subsets and kernel options.
To overcome this limitation, you can refain the required space by removing some subsets or by rebuilding the kernel with only the necessary options. You can also gain additional space, while keeping all subsets and kernel options, by creating and mounting new partitions.
The following steps show you how to create and mount new partitions for a UNIX file system (UFS). If you use AdvFS, use the mkfdmn and mkfset commands.
#
newfs /dev/rz1d
#
cd /usr/sys
#
mkdir FLAWLESS.BOOTABLE
#
mount /dev/rz1d /usr/sys/FLAWLESS.BOOTABLE
This device should have at least 75,000 512-blocks available.
#
newfs /dev/rz1b
#
mount /dev/rz1b /mnt
#
cp * /mnt
#
umount /mnt
#
mount /dev/rz1b /usr/sys/bin
After completing these steps, your system should have the necessary space to run btcreate.
If you are using AdvFS, the /usr/sys/bin file system must be copied during btcreate in order to copy the entire contents of the /usr file system.
Ensure that the kernel has been built with the tape drive connected to your system. If the drive was not connected when the kernel was built, you will see dump errors and the system will not be able to boot from the tape drive.
Bootable tape will not function with the -m mfs option on systems with 32MB memory configurations. After booting the kernel from tape, commands that use shared libraries will core dump.
Use the -m ufs option while creating the tape on systems with 32MB memory configurations.
Bootable tape does not support the bootable kernel built with the /usr/sys/conf/GENERIC kernel configuration file.
Be sure to use a custom kernel.
Using a bootable tape on a platform other than the one on which it was created is not supported. For example, you cannot create a tape on a 4100 system and boot from it on a 1000A system.
Using an inproperly configured QIC tape drive to create a bootable tape will result in an I/O error, a write error, or permission denied error. Therefore, you must take one of the following actions:
If creating a bootable tape with a UFS file system extends to multiple tapes, the /sbin/dump command will display a message indicating that the tape must be changed. If the tape is not changed promptly, warning messages will repeat periodically until the tape is changed.
When the tape has changed, the warning messages will stop.
When selecting disk partitions while restoring file systems from tape, add 5 percent to the needed file size displayed on the console.
A QIC tape created with the btcreate utility may fail with the following error when booted:
failed to send Read to mka...
Be sure that the tape is write protected before booting.
The behavior of the open call to a tape device has changed. It is no longer possible to open a write protected tape using write mode. The attempt to open the tape will fail, returning the following message:
EACCES (permission denied).
If an application is written so that it attempts to open the tape device with O_RDWR when the intention is to only read the tape, the open attempt will fail. Applications should be changed to open the device with O_RDONLY. For applications that cannot be changed, use the following command to obtain the previous behaviour of the open call:
#
sysconfig -r cam_tape open_behaviour=0
For Digital UNIX Version 4.0B and its software supplements, the supported version of ECU is V1.9 or higher. If your system is configured with an EISA bus, you should update the ECU to this supported version.
Mixing RZxx-VA and RZxx-VW drives in a wide storage enclosure may cause a problem. If you do have problems with this configuration, contact Digital technical support.
Consult the Open3D Software Product Description (SPD) before installing Open3D to ensure that this Digital layered product is supported on your system.
Installing Open3D on systems not supported by the Open3D layered product can leave your system in an unusable state.
The following notes apply to the use of enhanced security features.
This note covers problems that may occur when distributing of enhanced security profiles via NIS.
The more NIS slave servers a given NIS domain has, the slower this update process will be. Also, simultaneous login attempts will be processed sequentially by the NIS master, each waiting on the yppush for the previous to succeed. Thus, with several simultaneous attempts, some may time out and need to be retried by the users. This can be alleviated to some extent by using the -p option to yppush. One way to do this is to modify the file /var/yp/Makefile and change the YPPUSH= line. For example:
YPPUSH=$(YPDIR)/yppush -p 6
This example allows up to six simultaneous transfers to NIS slave servers.
This can be addressed by ensuring that all the NIS slaves have these maps by a procedure like the following (to be executed on each slave server which does not yet have these maps):
#
/var/yp/ypxfr -d <domainname> -h NISMASTER -c prpasswd
#
/var/yp/ypxfr -d <domainname> -h NISMASTER -c prpasswd_nonsecure
In the above, substitute the name of the local NIS master server for the NISMASTER token. This will transfer initial copies of those maps for those slave servers.
#
make passwd prpasswd
Internal testing has demonstrated that 4000 profiles and infrequent logins (three at a time) can work, but even fewer profiles can be accommodated if bursts of nearly-simultaneous logins are frequent.
Because the user profiles and ttys information are now stored in database files, the previous recovery method of editing the files while in single-user mode is no longer available. However, as long as the /usr (and, if separate, /var) filesystems are mounted, the edauth utility can be used in single-user mode to edit extended profiles and ttys database entries.
If the /etc/passwd file is somehow lost, but the extended profiles are still available, then a command sequence like the following can be used to recover some of the missing data:
# bcheckrc # /tcb/bin/convuser -dn | /usr/bin/xargs /tcb/bin/edauth -g | \ sed '/:u_id#/!d;s/.*:u_name=//;s/:u_id#/:*:/;s/:u_.*$/:/' \ >psw.missing
This will create a psw.missing file containing entries like the following:
root:*:0:
Primary group information, finger information, home directory, and login shell are not recorded in the extended profile. The data for those fields must be recovered by other means.
Enhanced security will not allow user names longer than the documented maximum of 8 characters.
The initial release of Multimedia Services for Digital UNIX Version 2.0 required several modifications to fully support Version 4.0 and Version 4.0 components. As a result, Multimedia Services for Digital UNIX Version 2.0A is shipped on the Digital UNIX Associated Products Volume 1 CD-ROM. The Multimedia Services Version 2.0A runtime should be installed instead of the Version 2.0 runtime. The Multimedia Services Version 2.0 development kit, not distributed with the Digital UNIX Version 4.0 distribution, can be used with the Multimedia Services Version 2.0A runtime kit. See the release notes for Multimedia Services for Digital UNIX Version 2.0A for more details. The release notes can be found on the CD -ROM as
DOCUMENTATION/HTML/MME201_RELNOTES*
Replaying the console messages that are generated during system startup is not supported until an updated version of DECevent is available.
If a printer is connected to multiple queues through a LAT or a local tty port and different jobs are submitted to different queues within a short period, some of the jobs may be lost. If this happens, resubmit the print request.
The following notes apply to network and communications software.
The following notes and restrictions apply to the PCI Token Ring support:
#
ifconfig tra0 speed 4
When using netconfig while CDE is running, avoid restarting network services after reconfiguring the primary network interface. This action can result in error dialog boxes and may even cause CDE to hang.
The problems may not be observed until you have set up BIND using bindconfig. In particular, do not use the following netconfig features while running a CDE session:
restarting network services
In order for the configuration changes to take effect, reboot your machine from the command line using /sbin/reboot or /sbin/shutdown -r now.
When using bind() on a UNIX domain (AF_UNIX) socket, the default modes of the socket have changed in this release. Previously, the mode of a newly created socket was always 0777, regardless of the value of the creating process's umask. In this release, this behavior has changed so the mode of a newly created socket is as follows:
(0777 &~ umask)
The previous behavior (0777, regardless of umask) may be restored by setting the kernel configuration flag insecure_bind to a value of 1. This can be done by either or both of the following two methods:
generic: insecure_bind = 1
Then you must reboot the kernel.
#
sysconfig -r generic insecure_bind=1
The following changes and restrictions apply to ATM
The command syntax for the atmarp command has changed due to Multiple LIS support. The atmconfig command has added many new options mainly due to CBR (Constant Bit Rate) support. Lastly a new command, atmsig, is now required in all ATM startup scripts. Therefore all ATM startup scripts including /etc/atm.conf will need to be modified. See the reference pages associated with atmconfig, atmarp, and atmsig for further details.
When the ATM end system is connected to a Digital Gigaswitch, that is running Version 1.3 (or less) of the Gigaswitch software, the following line specifying the useesi keyword is required:
atmconfig up driver=lta0 useesi=1-4 wait
The lockd daemon will need to be restarted after ATM is brought up. This is because the lockd daemon takes a one time look at the IP interface list. ATM interfaces such as LIS0 are dynamically added.
The following restriction applies only when LANE is active.
If UNI signaling is disabled with the atmsig down command, it will not be correctly restarted by an atmsig up command. Following a command such as:
atmsig down driver=ltaX
You must do the following to successfully restart signalling:
atmconfig down driver=ltaX
atmconfig up driver=ltaX
atmsig up driver=ltaX
The IFNET paradigm allows the bridging of streams device drivers to sockets. This release supports SVR4 streams, but the IFNET paradigm is not fully supported. IFNET is only supported over the ln ethernet interface and the number of ln devices supported is limited to two devices.
This release does not support Orderly Release in XPG4 XTI (default XTI interface). It is still available for users of XPG3 XTI. See the Network Programmer's Guide for information on using XPG3 XTI.
When you restart the network using the following command:
#
/usr/sbin/rcinet restart
The ifconfig command is run by the /usr/sbin/rcinet script. This will clear and reset the primary network interface address.
Network interfaces with configured interface aliases use the alias address as a source address for outgoing packets. Resetting the primary network interface address can cause a problem for systems with a firewall or proxy-access configuration based on the primary address. Generally, alias addresses are not in the access control lists in such systems.
To avoid this problem, you can use one of the following solutions:
#
ifconfig <if_w_aliases> down delete
When restarting the network using netsetup, an error message similar to the following will be displayed:
kill: 204: no such process
This problem also exists when running the following commands:
#
rcinet stop
#
rcinet restart
The message is incorrect and has no effect on your system.
The Common Desktop Environment (CDE) provides facilities and features for applications to communicate in a networked environment. After the network is configured and enabled, these features become available each time a new desktop session is started. After a desktop session has started, the current session has a static dependency on the state of the network configuration. Network and system administrators should be very cautious about dynamic changes to the network configuration while in a network aware desktop session.
Prior to making any dynamic network changes, such as changing the state of your network adapter to off or changing your primary network address, add the following entry to the /.dtprofile file:
export DTNONETWORK=true
The system administrator must then log out and back in as root for the change to take effect. This change removes the dependency on the state of the network. Failure to do this may result in a session hanging after clicking on a CDE icon, such as the screen lock or Exit icons.
After all network changes are completed, remove the export DTNONETWORK=true entry from the /.dtprofile file.
The following notes apply to Local Area Transport (LAT).
The latsetup utility sometimes creates devices with duplicate minor numbers. If you manually create LAT BSD devices that do not match the valid BSD tty name space convention, latsetup can create devices with duplicate minor numbers. For example, creating device tty0 with a minor number 2 instead of 1 can cause this problem.
When a CTRL/A character is typed during a LAT tty session, all lowercase characters are converted to uppercase. Another CTRL/A will change the mode back to normal.
When doing a number of simultaneous llogin connections it is recommended to use llogin with the -p option. To speed up an llogin connection, it is also recommended to add the target host name as a reserved service.
It is no longer necessary to build LAT into the kernel. LAT is not made a mandatory kernel option upon selecting the LAT subset and will not appear in the kernel configuration file. As LAT requires the Data Link Bridge (DLB), it is still necessary to build DLB into the kernel when using LAT.
The default behavior upon booting to multi-user mode is for LAT to be dynamically loaded into the running kernel. If LAT is not started at boot-time via the /sbin/rc3.d/S58lat script, the recommended method for starting and stopping LAT is to verify that LATSETUP is enabled in /etc/rc.config and execute the /usr/sbin/init.d/lat program, using the start or stop options.
In the slave-only Version 2.0 LAT implementation, service group codes are used in solicitation messages for host-initiated connections,
Many users rely on the ability to control access to server ports by changing the group codes of the locally offered services. Although it is contrary to the recommendations of the LAT protocol, this behavior is once again supported. Outgoing port group codes, normally used for this purpose, continue to be used in all other cases where they are required by the protocol.
If the shutdown -r command is executed when there are LAT login sessions with active background processes, the shutdown program appears to stall. The workaround for this problem is to halt LAT (using the latcp -h command) either before executing the shutdown command or after it has stalled.
The notes in this section apply to file systems.
The mount command was changed in Digital UNIX Version 3.2D-1 and all versions thereafter. It now checks all of the sizes of the volumes in the file domain of the fileset it is mounting. It verifies that the volumes are at least the size stated in the file domain internal structures.
If all of the volumes pass the size check, a mount update is issued on the fileset. In this case, the fileset is left in a read-only state only if the user originally issued the mount command with the read-only option.
If one or more volumes fail the size check, each volume is flagged and the fileset is left as read-only. To correct this, make a complete backup copy of all volumes in the file domain. This will assure that you can recover your data in the event of a catastrophic failure during the volume label correction process.
Make sure there is adequate free space in the fileset. There must be enough free space to hold all the data with any one of the incorrectly labeled volumes removed. If there is not enough free space, you must add additional volumes to the fileset. Once this is done, perform the following steps on each of the failed volumes:
Step 4 is very important. If the domain is not balanced after adding the corrected volume, other incorrectly labeled volumes may fill and cause an I/O error.
After completing the corrective procedures, recheck the volumes by issuing a mount -u command on the fileset. This will cause the mount command to recheck all of the volumes.
The following is an example of the error message that the mount command will issue if a volume in a file domain is incorrectly labeled.
#
mount dmn#fs1 /fs1
Domain dmn, Volume /dev/rz9c has an incorrect size! Domain dmn, Volume /dev/rz10c has an incorrect size! Domain dmn, Volume /dev/rz11c has an incorrect size! Domain dmn, Volume /dev/rz12c has an incorrect size! Domain dmn, Volume /dev/rz13c has an incorrect size! Domain dmn, Volume /dev/rz14c has an incorrect size! The file system dmn#fs1 has been mounted READ ONLY! Please check the disklabel of all volume(s) listed above!!
Under certain hardware failure scenarios, an LSM volume configured with a sparse plex may erroneously return success to the file system or application when in fact the I/O failed. Digital recommends that you do not configure volumes with sparse plexes until this problem is fixed.
Root, primary swap, and secondary swap volumes configured under LSM have the following restrictions:
When using the UFS file system, there is a problem when setting properties. Setting a property on a FAST symbolic link, a block special file, or a character special file will cause fsck to erroneously detect contradictory block counts and produce inconsistent file system activity. There is no solution for this problem and it will be fixed in a future release.
The following AdvFS variables are not documented in the Guide to File System Administration for the POLYCENTER Advanced File System and Utilities for Digital UNIX.
This configurable parameter controls the behavior of both the sync system call and the update daemon. The sync system call asynchronously writes dirty in-memory file data to disk. There are two ways a file can have dirty data in memory. One way is via the write system call, the other way is from a memory write reference after an mmap system call. The update daemon runs every 30 seconds and issues a sync call for every read/write file system mounted.
AdvFS provides the ability to select the behavior of the sync system call. When AdvfsSyncMmapPages has the default value of 1, the sync system call asynchronously commits dirty memory mapped file data to disk. If AdvfsSyncMmapPages is reconfigured to a value of 0 then only dirty data from the write system call will be asynchronously written to disk.
For each device, AdvFS maintains a number of queues that control when a given I/O request is handed off to a disk driver (or pseudo driver). I/O requests for file data are placed on the AdvFS consolidation queue, where small logically contiguous blocks requests are consolidated into larger consolidated I/O requests. The file data consolidation queue contains both read-ahead file data and asynchronously written file data.
The AdvfsMaxDevQLen variable controls how many I/O requests to move off the consolidation queue and onto the device queue. All I/O requests that are put on the device queue are sent to the device driver. By default, AdvfsMaxDevQLen is set to 80. This means that the device queue must contain fewer than 80 I/O requests before AdvFS will take a request from the consolidation queue.
Limiting the length of the device queue directly affects the amount of time a synchronous (blocking) I/O will take before it complete. AdvFS issues several types of blocking I/O requests, including AdvFS metadata and log data.
The optimum value of AdvfsMaxDevQLen can vary from system to system. The default value of 80 should be appropriate for most configurations. It is possible there may be a need to change this value, particularly for very fast of very slow devices and controllers. A good rule of thumb is to set the AdvfsMaxDevQLen parameter no higher than the average number of I/O requests that be performed in 1/2 second. If no limit is desired, AdvfsMaxDevQLen can be set to zero.
This configurable parameter is used to control whether or not a defined number of I/O requests should be moved from the consolidation queue to the device queue when a flush of the blocking queue is requested. AdvfsFavorBlockingQueue should not be changed from its default value of 1.
The fsync system call is used to synchronously write dirty file data to disk. There are two ways a file can have dirty data in memory. One way is via the write system call. The other is from a memory write reference after an mmap system call. For AdvFS files, the fsync() system call will only write out dirty data from the write system call. If dirty data from an mmap() also needs to be written then the msync system call must also be used.
For an NFS client to make direct use of ACLs or extended attributes (property lists) over NFS the proplistd daemon must be enabled on an NFS server. The proplist mount option must be used when mounting on the client. Access checks will be enforced by the server in any case, although NFSv2 client caching could sometimes cause inappropriate read access to be granted. Correctly implemented NFSv3 clients will make the necessary access checks.
Start the proplistd daemon by selecting the number of proplist daemons to run when you use the nfssetup utility. You can also start the daemon manually with the proplistd command. For example:
#
/usr/sbin/proplistd 4
On the client, the filesystem must be mounted with the proplist option by either of the following methods:
sware1:/advfs /nfs_advfs nfs rw,proplist 0 0
#
mount -o proplist sware1:/advfs /nfs_advfs
See the acl(4), fstab(4), proplist(4), mount(8), nfssetup(8), and proplistd(8) reference pages for more information. Note that the proplist option is not documented in mount(8).
On AdvFS file systems there is a hard limit of 1560 bytes for a property list entry. Since Access Control Lists (ACLs) are stored in property list entries, this equates to 62 ACL entries in addition to the three required ACL entries. The error EINVAL will be returned if you attempt to exceed this limit.
To facilitate interoperation of the UFS and AdvFS ACLs, a configurable limit has been imposed on UFS ACLs. The default value of the UFS limit is 1548 bytes, equivalent to the 65 entry limit on AdvFS. The UFS configurable limit on ACLs has been added to the sec subsystem and has been given the attribute name ufs-sec-proplist-max-entry. The attribute can be dynamically configured using the sysconfig utility or by setting the attribute in the file sysconfigtab.
A configurable property list element size for UFS has also been added to the sec subsystem and has been given the attribute name ufs-proplist-max-entry. The value of ufs-proplist-max-entry must be larger than ufs-sec-proplist-max-entry by enough space to hold a property list element header. Adjustment of ufs-proplist-max-entry to achieve this is done automatically by the sysconfig utility. The default value of ufs-proplist-max-entry is 8192 bytes.
See the cfgmgr(8), seconfig(8), seconfigdb(8), and sysconfigtab(4) reference pages for more information.
The following notes discuss features, problems, and restrictions of the POLYCENTER Advanced File System (AdvFS).
Note
This is an important note for users of vdump and vrestore
Backups made using vdump on Digital UNIX Version 4.0 cannot be restored using vrestore on earlier versions of Digital UNIX. Patches will be made available for earlier versions of vrestore to correct this problem.
Backups made using vdump on earlier versions of Digital UNIX can be restored using vrestore under Digital UNIX Version 4.0 without problems.
Under some circumstances, AdvFS can panic with the following message:
log half full
This can occur when a big percentage of a very large file is truncated and the fileset containing the file has a clone fileset. Truncation occurs when an existing file is overlaid by another file, and explicitly by the truncate system call.
The same will happen if very large, very fragmented files are migrated. Migration occurs when the balance, rmvol, and migrate AdvFS utilities are run. Files with greater than 40000 extents are at risk, unless the transaction log size is increased as follows:
The following command will indicate how many extents a file is using:
#
showfile -x <filename> | grep extentCnt
Backup and restore will help to defragment files as will copying (not moving with mv) a file to another name. However, if the fileset has a clone and if a large file is truncated as a result of the copy, the truncation panic could occur.
On AdvFS file systems there is a hard limit of 1560 bytes for a property list entry. Since Access Control Lists (ACLs) are stored in property list entries, this equates to 62 ACL entries in addition to the three required ACL entries. The error EINVAL will be returned if you attempt to exceed this limit.
To facilitate interoperation of the UFS and AdvFS ACLs, a configurable limit has been imposed on UFS ACLs. The default value of the UFS limit is 1548 bytes, equivalent to the 65 entry limit on AdvFS. The UFS configurable limit on ACLs has been added to the sec subsystem and has been given the attribute name ufs-sec-proplist-max-entry. The attribute can be dynamically configured using the sysconfig utility or by setting the attribute in the file sysconfigtab.
A configurable property list element size for UFS has also been added to the sec subsystem and has been given the attribute name ufs-proplist-max-entry. The value of ufs-proplist-max-entry must be larger than ufs-sec-proplist-max-entry by enough space to hold a property list element header. Adjustment of ufs-proplist-max-entry to achieve this is done automatically by the sysconfig utility. The default value of ufs-proplist-max-entry is 8192 bytes.
See the cfgmgr(8), seconfig(8), seconfigdb(8), and sysconfigtab(4) reference pages for more information.
Running the verify command with the -F flag causes some recovery to be done on the domain before the attempt to mount it.
Avoid using the rmfset utility on busy domains. If you attempt to remove a fileset using the rmfset command and the target domain is experiencing a lot of I/O, the rmfset operation may hang.
The vdump and vrestore utilities correctly save and restore AdvFS sparse files. In previous versions, the holes in the sparse files were allocated disk space and filled with zeros. Note that sparse files that are striped are still handled as in previous versions.
AdvFS will now verify at mount time that all of the data in all of the volumes in a domain can be accessed. It does this by attempting to read the last block in each volume as specified by the disk label that was in use at the time that the volume was added to the domain. If it cannot read that block, it attempts to read the last block that AdvFS has marked as being currently used to hold data. If AdvFS cannot read the last in-use block for any volume in the domain, the mount will fail. If it can read the last in-use block but cannot read the last block as specified by the disk label, the mount will succeed but in read-only mode.
One reason that the last block may not be able to be read is that a disk may be mislabeled on a RAID array. The user should check the labels of the flagged volumes in the error message. If the disk label is incorrect, the user can repair the domain in one of the following ways.
Before attempting corrective action, you should back up all filesets in the domain. The corrective action depends on the state of the domain. If the domain consist of multiple volumes and has enough free space to hold two entire volumes, it is possible to remove the offending volumes one at a time, fix the disk label and add them back to the domain. Perform the following operation on each of the failed volumes:
Step 4 is important. If the domain is not balanced after adding the corrected volume, the user runs the risk of filling up one of the incorrectly labeled volumes and inducing an I/O error.
If the domain's free space is less than two volumes, you should back up all the filesets in that domain, remove the domain, fix the disk labels of the volumes, and rebuild the domain. Then restore the filesets from the backups.
Another example of why a mount might fail in this way would be that an LSM volume upon which an AdvFS domain resides has been shrunk from its original size.
When log or metadata write errors occur (for example, due to a disk failure or media error) AdvFS initiates a domain panic rather than a system panic on any non root file domains. A domain panic prevents further access to the domain, but allows the filesets in the domain to be unmounted.
When a domain panic occurs, a message is displayed in the following format:
AdvFS Domain Panic; Domain <name> Id <domain_Id>
For example:
AdvFS Domain Panic; Domain cybase_domain Id 2dad7c28.0000dfbb
After a domain panic, use the mount command to list all mounted filesets then use umount to dismount all filesets in the domain specified in the error message. You can then take the necessary steps to correct the hardware problem. After you have corrected the hardware problem, it is recommended that you run the verify command (the domain structure checker) on the domain before remounting it. This will determine if the write error compromised the domain.
If a disk has a partition erroneously labeled AdvFS that overlaps a UFS partition, a file system check and repair operation (fsck and ufs_fsck) will fail on a partition that overlaps the AdvFS partition. The solution is to relabel the AdvFS partition on the disk.
Conversely, if a disk partition that overlaps AdvFS is erroneously labeled UFS, an AdvFS file system check and repair operation, verify, will fail on a partition that overlaps the UFS partition. To correct the problem, relabel the UFS partition.
You can reuse a partition that was previously part of an AdvFS domain. However, before you reuse the partition, you must remove the domain on the partition you want to reuse. Remove the entire domain by using the rmfdmn command. After the unused domain is removed, you can create a new domain on the partition.
If you attempt to mount a UFS file system while in single-user mode on a system that is configured with AdvFS as root, the following error will occur:
Error checking for overlapping partitions: Invalid MSFS fileset name (root_device) in mounttab.
To manually mount a UFS file system while in single-user mode on a system with an AdvFS root, you must perform a mount update on the root file system. Use the following command:
#
mount -u /
You can then mount any UNIX file systems.
Under certain conditions, the disk usage information on an AdvFS file system may become corrupted. To correct this, turn on quotas in the /etc/fstab file for the affected file system, and then run the vquotacheck command on the file system. This should correct the disk usage information.
Both the fstat(2) reference page and the /usr/include/sys/stat.h file inaccurately state that the combination of the st_dev and the st_ino fields of the stat structure create a unique file identifier through time. Because the st_ino value can be reused, the only way to create a unique file identifier is the combination of the st_ino, st_dev, and st_gen fields. While this is true in UFS, it is even more important in AdvFS, which recycles the st_ino values rapidly.
If an NFS server is exporting an AdvFS directory and the client to which it is exporting crashes, upon reboot of the client, the following error message may appear on the console of the server:
lockd : can't clear lock after crash of client client_name : invalid argument.
This message does not disrupt NFS operations.
When a large percentage of a very large file is truncated and the fileset containing the file has a clone fileset, the system will panic with a log half full error message. Truncation happens when an existing file is overlaid by another file or when the truncate system call is made.
The panic also occurs if very large, very fragmented files are migrated. Migration occurs when you run the AdvFS balance, rmvol, and migrate utilities. Files with greater than 40,000 extents are at risk. To determine how many extents a file is using, enter the following command:
#
showfile -x
filename | grep extentCnt
The backup, restore, and copy commands tend to defragment files. The mv command does not. However, if the fileset has a clone and a large file is truncated as a result of the copy, the log half full panic again could result.
Logical Volume Manager (LVM) support was retired in Digital UNIX Version 4.0 and there will be no further support of LVM.
All volume management functions are provided by the Logical Storage Manager (LSM). For information on how to migrate to LSM, see the Logical Storage Manager manual.
The following notes describe problems and restrictions of the Logical Storage Manager (LSM).
Physical block 0 on Digital disks is typically write-protected by default. If a disk is added to LSM by using the voldiskadd utility, physical block 0 is skipped. However, if a partition that includes physical block 0 is encapsulated into LSM by using the volencap, vollvmencap, or voladvdomencap utility, physical block 0 is not skipped. This is not a problem because the file system already skips block 0 and does not write to it.
A problem can occur when an LSM volume that contains a write-protected block 0 is dissolved and its disk space is reused for a new purpose. Neither the new application nor LSM know about the write-protected physical disk block 0 and a write failure can occur.
To fix this problem, use the following steps to remove the write-protected physical disk block 0 from the LSM disk before it can be assigned to the new volume:
When an LSM mirror is created using a disk that is configured as Just-a-Bunch-of-Disks (JBOD) off of either the SWXCR-P or SWXCR-E RAID controllers, a disk failure requires that you reconfigure the disk on the controller. The disk is in an unusable state once it is set to off line by the controller and cannot be used by LSM until it is reconfigured. Refer to the StorageWorks RAID Array 200 Subystem Family Installation and Configuration Guide.
If you install LSM by using the setld utility after you originally install Digital UNIX, you must rebuild the system kernel to enable LSM.
To rebuild the kernel, run the doconfig utility with no command flags. Note that the doconfig menu display does not include LSM. However, the doconfig utility will build a kernel that includes LSM. Refer to the LSM Installation documentation for more information.
Only LUN 0 is supported as a boot device by the console. Hence, the LSM rootvol and swapvol volumes can be mirrored only to LUN 0 in an HSZ. Therefore, when you use the volrootmir script to mirror rootvol and swapvol, use only an LUN 0 on an HSZ as an argument to the volrootmir script.
If you use the LSM rootvol volume for the root file system and the swapvol volume is in use as a primary swap volume, LSM adds the following entries to the /etc/sysconfigtab file to enable rootability:
lsm: lsm_rootvol_is_dev=1 lsm_swapvol_is_dev=1
If these entries are deleted or if the /etc/sysconfigtab file is deleted, the system will not boot. If this happens, you can boot the system interactively as follows:
>>>
boot -fl i
......... .........Enter kernel_name option_1 ... option_n: vmunix
After the system boots, edit the /etc/sysconfigtab file and add the LSM entries as shown above. Reboot the system for the changes to take effect.
LSM Volumes enabled with Block Change Logging (BCL) requires two or more log subdisks that are at a minimum one sector long. If you intend to use BCL on any volumes that do not already have logging subdisks, Digital recommends that you allocate at least two sectors to each log subdisk.
Any volumes that currently use single sector logging subdisks will continue to work correctly. However, Digital recommends that you reconfigure as soon as convenient to avoid being forced to do so at a later date.
Implementing these recommendations now will make the transition to new requirements in future releases easier.