Physical and Virtual Tape Access
- W.Erber
Adding tape devices to the systems emulated by Charon-PAR is a process consisting of several steps:
- Link the tape container files and/or physical tape devices with the corresponding configuration paths in the Charon configuration file and load the devices.
- Enable the tape for use by the guest operating system.
The storage on the host system must provide enough capacity to store the tape containers used as virtual tapes by the emulated system.
The sections below describe the process in more detail.
Adding Tape Devices to the Configuration File
The new emulated tape device must be added to the configuration file of the Charon-PAR instance. This requires two commands:
Step | Configuration file entry | |
|---|---|---|
| 1 | Load the SCSI device. | load MKXnnn |
| 2 | Link the device with tape container file
physical tape device |
The path can point to an existing tape container file, but this is not mandatory. The emulator can create the container file if required.
You can identify the Linux tape device name using the command |
| 3 | Enable loading the virtual tape automatically (for virtual tapes based on container file). |
Please note:
|
The parameter MKXnnn encodes the device type and the device path of the emulated SCSI device presented to the guest operating system:
A tape device name has the following components:
- Value MK at the beginning of the device name: specifies that device type is tape.
- X: value consists of an uppercase letter encoding the device path. Please refer to Emulated Model Hardware Configuration Details for system specific device path information.
Example (rp4000):
A = device path 0/0/1/0
B = device path 0/0/1/1
C = device path 0/0/2/0
D = device path 0/0/2/1 - nnn: value encodes the SCSI device connected to the SCSI controller.
Formula: (SCSI target ID*100)+LUN
The LUN is always 0 for tape devices.
The following example adds a SCSI tape device with device path 0/0/1/0.6.0 (controller A, SCSI target ID 6, LUN 0) to the configuration:
|
Using Tapes on MPE/iX Guest Systems
Displaying Tape Characteristics on MPE/iX Systems
You can use the SYSGEN > IO to display tape characteristics as shown in the following example for LDev 7:
Using :DEVCTRL to Load and Eject a Tape on MPE/iX Systems
:DEVCTRL is an script that allows several aspects of physical tape devices to be controlled: in-drive data compression, tape loading, and automatic tape eject. This script is also used with Charon-PAR to load virtual tape image files. When :DEVCTRL is used to load a virtual tape file, it causes Charon-PAR to create (if necessary) and open the virtual tape file configured in the Charon-PAR configuration file.
The :DEVCTRL syntax is show below:
|
For example, to load an emulated type device on LDev 7, use the following command:
:DEVCTRL 7; LOAD=ONLINE |
Reading and Writing Tapes in MPE/iX
To read and write tapes under MPE/iX, the :STORE and :RESTORE commands are used. Please contact your Stromasys representative or partner if you need support with data backup and restore under MPE/iX.
Using Tapes on HP-UX Guest Systems
Displaying the Tape Configuration in HP-UX
After booting HP-UX with the virtual tape device configuration, you can use the command ioscan -C tape -fun to verify if the device was found and the driver was loaded. The following image shows an example:
The example shows two type devices and their associated special devices.
The lssf command can be used to verify that the special files point to the correct device paths, SCSI target ID and LUN.
|
Loading and Ejecting the Virtual Tape in HP-UX
Using the autoload Configuration Option
If autoload has been enabled in the configuration, the virtual tape is "loaded" and the container file created, if required, when the tape device is accessed.
Please note: Multi-volume backup software may not work correctly when autoload is enabled. If a new tape is requested by the software, it will not wait for a new tape to load but overwrite the existing file. If data is to be written to several tapes, this needs to be implemented manually (e.g., by a customized script).
For example, the following command will cause the container file for device /dev/rmt/0mnb (default tape device) to be created and opened:
# mt statusDrive: HP C1537AFormat:Status: [41114701] BOT online compression immediate-report-modeFile: 0Block: 0 |
The tape can be "ejected" by setting it to offline (device-name is the full path to the special file):
# mt -f <device-name> offl |
After the offline command, the container file can be removed (i.e., the tape can be archived).
To simulate the "swapping" of tapes, use the sequence:
- Release a loaded tape so its container file can be moved away:
mt -f <devicename> offl - Write to the tape as required (this will create a new container file, if needed).
- Release the loaded tape so its container file can be moved away:
mt -f <devicename> offl
Using Manual Tape Loading from the PA9 Console
In some cases it may be useful to load/unload tapes manually from the pa9-64 or pa9-32 console instead of using autoload. This is achieved by the following command (using PA9-64 as an example):
Load a tape:
pa9-64> MKXnnn.load |
Unload a tape:
pa9-64> |
Where MKXnnn stands for the virtual tape configuration name, for example, MKA500.
If no tape is "loaded" in the virtual tape device, the command
mt -f <device-name> status
will show a status of 0.
Reading and Writing Tapes in HP-UX
There are several commands that can be used to write data to / read data from the virtual tapes, for example,
- tar
- fbackup/frecover
- dump/restore
- vxdump/vxrestore
Please refer to your HP-UX documentation for details about the usage of these utilities.
© Stromasys, 1999-2024 - All the information is provided on the best effort basis, and might be changed anytime without notice. Information provided does not mean Stromasys commitment to any features described.