Backup Exec 21 Administrator's Guide

Some characters do not appear correctly in the terminal session during the installation. What should I do?

This error occurs when the system location uses a non-English language character-set on the computer on which you install the Linux Agent. You can switch to another location setting of the same language to try to resolve this issue.

The Linux Agent installer is unable to install the Linux Agent. The following error is reported in the installralus log file. What should I do?

VxIF::Error:: Unable to compress files. Hash(0x8711e8)->({GUNZIP} not found on <hostname>

To support the uncompressing of the Linux Agent platform-specific packages, you can install the GNU data compression utility. Install this utility on the computer on which you want to install the Linux Agent.

The utility is available at the following URL:

http://www.gzip.org

The Agent for Linux is installed on a Linux server in an NIS domain. Backup Exec is unable to browse resources on the server. What should I do?

Verify if the group line and the password line in the nsswitch.conf file are set to compatibility mode. If they are, then you must configure the /etc/passwd and /etc/group files. Refer to the nsswitch.conf man pages for additional information on how to configure the nsswitch.conf to use compatibility mode.

Alternatively, change the password line and the group line to NIS files so that the Linux server validates the user through NIS. If the NIS server is unavailable or if the user is not found, the local files are used for validation.

I cannot load the Linux Agent. When I attempt to load the Linux Agent in console mode, /beremote --log-console shows the following message:

ACE_SV_Semaphore_Complex: no space left on device.

What should I do?

This issue occurs when the computer reaches its maximum limit on allowable semaphores. It can occur after an unexpected termination of the Linux Agent. When the Linux Agent unexpectedly terminates, it is unable to clean up some of the semaphore resources that it used. Other processes may have caused the use of semaphores to reach the limit. You must restart the computer to safely recover it from this condition.

If other processes are running, it may not be feasible to restart the computer. Instead, you can use the commands that let you list and then remove all semaphores that are in use by the operating system. Be careful when you select semaphores to remove. Semaphores that are in use by the Linux Agent cannot be identified. If you remove semaphores of other programs that are in use, those programs can become unstable.

To list semaphores, you can type the following command:

ipcs -a

To remove semaphores for each identifier that is listed, you can type the following command:

ipcrm -s <id>

I cannot load the Linux Agent. When I attempt to load the Linux Agent in console mode, /beremote --log-console shows the following message: Error while loading shared libraries: libstdc++.so.5: cannot open shared object file: No such file or directory.

What should I do?

This error indicates that the libstdc++.so.5 library is not in the /usr/lib directory. This library is necessary to let the Linux Agent start and function. To resolve this issue, install the libstdc++5 package.

You can install this package from the media on which your copy of Linux was provided. Or, you can run the following command from a computer that has Internet access:

apt-get install libstdc++5

For SUSE Linux Enterprise Server 11, run the following command:

zypper install libstdc++5

On an Asianux operating system, Backup Exec displays an error that the Exec logon account for backup and restore options is not a part of the beoper group.

What should I do?

Sometimes the /etc/group file is not correctly updated when a user is added to the beoper group. This leads to a failure of the getgrnam() Linux API. As a workaround, you should manually edit the /etc/group file and add the specific POSIX user name for the logon account in question to the beoper group.

For more information on editing the /etc/group file, see the Asianux operating system documentation.

On any Linux computer on which the GNOME Virtual File System (GVFS) is installed and mounted, the GVFS cannot be browsed, backed up, or restored to.

What should I do?

On computers on which GVFS is installed, the .gvfs mount point is created for every user who logs on using the GNOME user interface. The mount point is created in the user's home directory. For example, if the user's logon name is John, then the directory appears as the following:

Echo $>ls - la /home/John

dr-x------ 2 John John 0 2009-06-16 18:16 .gvfs

The output of the mount command appears as the following:

"gvfs-fuse-daemon on /home/John/.gvfs type fuse.gvfs-fuse-daemon (rw,nosuid,nodev,user=John)".

This mount point is created when the user logs on using the GNOME graphical user interface. The mount point is deleted when the user logs off.

A defect has been reported for the GVFS that even the superuser (root) does not have access to the file system. Refer to the following URL for more information:

http://bugzilla.gnome.org/show_bug.cgi?id=560658

This defect prevents the Agent for Linux from running on the GVFS. Therefore, the GVFS cannot be browsed, backed up, or restored to. Other file systems on Linux computers on which GVFS is installed are also inaccessible.

Additionally, on Linux computers on which GVFS is installed and mounted on the home folder of a particular user, browse operations, and backup and restore operations on or to the following file systems may fail:

As a resolution, try dismounting the GVFS, and retry the operation.

The Agent for Linux installer does not install the Perl Switch.pm module on 64 bit Ubuntu 14.04.

What should I do?

Using the following steps, you must manually install Switch.pm before you attempt to install the Backup Exec Agent for Windows.

For Ubuntu 14.04, you must enable manual logins and add a root user using the following steps: