Print | Rate this content

HP StorageWorks 1/8 G2 Tape Autoloader - Troubleshooting

Overview

This document contains information about various troubleshooting procedures for the HP StorageWorks 1/8 G2 Tape Autoloader.

CAUTION: Shipping Lock : The shipping lock must be removed for the robotics to work properly. The device displays a robot move error if the shipping lock is not removed.
NOTE: For an online troubleshooting tool, go to www.hp.com/support/MSLG3Tstree . The problems and solutions in the troubleshooting tree is updated more frequently than this manual.

top

SCSI installation problems

Problems encountered during installation are usually caused by improper SCSI bus configuration, application software configuration errors, or an incorrectly configured operating system. If the application software does not communicate with the device after installation, check the following:

  • SCSI ID : The device is pre-configured to SCSI ID 4. Depending on other devices attached to the same SCSI bus and their SCSI IDs, the SCSI ID will need to be changed before using the device. Review the manuals for the other devices on the SCSI bus or the operating system to determine which SCSI IDs are currently in use. Change the SCSI ID with the Operator control panel (OCP) or Remote management interface (RMI).
  • Multiple LUN support : The device uses two Logical Unit Numbers (LUNs) to control the tape drive (LUN 0) and robotic (LUN 1). The device requires an HBA with multiple LUN support and multiple LUN support must be enabled on the host computer. When multiple LUN support is not enabled, the host computer can see the tape drive, but not the Tape Autoloader.
    NOTE: Most RAID or array controllers do not provide multiple LUN support.
  • SCSI Cabling : Verify that all SCSI cables are securely connected on both ends. Check the length and integrity of the SCSI cabling. Check the SCSI connector for bent pins. The length of the internal SCSI cabling inside the device is 2 feet (60 cm). This length must be included in any calculations of cable length.
    • For LVD SCSI, the maximum length for a single device on the SCSI bus is 82 feet (25 meters). For Ultra 320 or multiple devices on a bus, the maximum combined internal/external length is 40 feet (12 meters).
    • If there is a combination of LVD and SE devices on the bus, the maximum cable length reverts to the SE specification, which for Ultra devices is 10 feet (3 meters) for four or fewer devices, and 5 feet (1.5 meters) for more than four devices.
  • Termination : If one of the ports on the device is not connected to another SCSI device, the port must be terminated. Only the devices physically located at the beginning and end of the SCSI bus should be terminated. Refer to the manuals supplied with the other devices on the SCSI bus for information on enabling or disabling termination on those devices.
    • To terminate the second SCSI port on the device, locate the terminator in the accessories package and press it firmly into either of the two SCSI connectors on the back panel. Secure the terminator by tightening the finger-screws until snug. The supplied terminator is “dual mode” and works on both Low-Voltage Differential (LVD) and Single Ended (SE) SCSI buses. Check all SCSI and power connections and confirm that the unit is attached to a valid SCSI SE or LVDS bus.
  • Compatibility : Ensure that the device is compatible with the SCSI host adapter and backup application being planned to use. For a list of compatible SCSI host bus adapters and application software, check with the SCSI host adapter manufacturer, backup application vendor, or the HP support website at http://www.hp.com/go/ebs .
    NOTE: The host bus adapter should be SCSI-3 LVDS. A single-ended SCSI host bus adapter severely degrades performance. Also, if any SE devices are on the same SCSI bus, all of the devices on the bus slow down to SE speed, which severely degrades performance.
  • SCSI Host Adapter Installation : Verify that the SCSI host adapter is installed correctly. Refer to the manual that came with the SCSI host adapter for installation and troubleshooting instructions. Pay particular attention to any steps describing configuration settings. Make sure that the host adapter is properly seated in the motherboard slot and the operating system correctly detects the host adapter. Make sure that the proper device driver is installed for the SCSI host adapter.
  • Backup Application Installation : Refer to the documentation included with the backup application for instructions on how to verify proper installation. Some backup software packages require an additional module to communicate with the robotics.
  • Device Driver Installation : Make sure that the proper device driver, if applicable, is installed for the device. Check the software provider’s website for the latest drivers and patches.
NOTE: Many backup applications use their own drivers. Before installing a driver, make sure it is not in conflict with the application software.

top

SAS installation problems

Problems encountered after installation are often caused by improper SAS cable connections, application software configuration errors, or an incorrectly configured operating system. If the application software or operating system does not communicate with the device after installation, determine the extent of the detection problem.

Based on the extent of the detection problem, check the following:

  • If neither the application software nor operating system detects the tape drive, or they do not detect both the tape drive and the Autoloader:

    • Verify that all SAS cables are securely connected on both ends. If the mini-SAS connectors that connect to the tape drive and some HBAs will not plug in, check the key. The mini-SAS connector on the tape drive is keyed at location four, which is the standard location for end devices. If the connector on the cable is keyed in a different location, not only will the connector not plug in, but the cable probably will not work.

    • Check the length and integrity of your SAS cabling. For reliable operation, do not use a SAS cable longer than six meters. Do not use a cable adapter or converters between the HBA and the Autoloader.

    • Check the SAS connectors for bent pins or debris.

  • If the application software or operating system detects the tape drive, but not the Autoloader:

    • Verify that multiple LUN support is enabled on the HBA. The device uses two Logical Unit Numbers (LUNs) to control the tape drive (LUN 0) and robotic (LUN 1). The device requires an HBA with multiple LUN support and multiple LUN support must be enabled on the host computer. When multiple LUN support is not enabled, the host computer can see the tape drive, but not the Autoloader.

      NOTE: Many RAID or array controllers do not provide multiple LUN support.
  • If the application software or operating system does not detect any devices on the HBA:

    • Verify that the SAS host adapter is installed correctly. Refer to the manual that came with your host adapter for installation and troubleshooting instructions. Pay particular attention to any steps describing configuration settings. Make sure that the host adapter is properly seated in the motherboard slot and the operating system correctly detects the host adapter.

    • Verify that the proper device driver is installed for the SAS host adapter.

  • If the Autoloader is detected by the operating system, but not by the application software:

    • Refer to the documentation included with your backup application for instructions on how to verify proper installation. Some backup software packages require an additional module to communicate with the robotics.

  • If the Autoloader is detected by the operating system, but is listed as an unknown or generic device:

    • Make sure that the proper device driver, if applicable, is installed for the device. Check your software provider's website for the latest drivers and patches.
      NOTE: Many backup applications use their own drivers. Before installing a driver, make sure it is not in conflict with the application software.

If you continue to have problems with a SAS Autoloader, check the following:

  • Ensure that the device is compatible with the SAS host adapter and backup application you plan to use.

  • Verify that your HBA is supported by the host computer and qualified with the Autoloader.

top

Operation problems

Power Problems
Problem
Solution
Device does not power on.
  1. Check all power cord connections.
  2. Make sure the power button on the front panel has been pressed, and the green READY LED is lit.
  3. Make sure the outlet has power. Try another working outlet.
  4. Replace the power cord.
No display messages appear.
  1. Make sure the power cord is connected.
  2. Make sure the power button on the front panel has been pressed, and the green READY LED is lit.
  3. Power cycle the device.
Errors displayed on the front panel
Problem
Solution
“!” in operator panel inventory display.
Export the data cartridge.
The LCD displays an error code.
Look up the error code, try to resolve the failure, and power cycle the device.
Tape movement problems
Problem
Solution
Tape stuck in drive.
Try the following steps, in this order, to remove the stuck tape.
NOTE: The tape drive must rewind the tape before ejecting it. This can take as long as five minutes, depending on how much tape must be rewound. Once the tape is rewound, the eject cycle will take fewer than 16 seconds.
The READY light flashes while the tape rewinds. Wait for the tape to finish rewinding before attempting another operation.
  1. Attempt to unload the tape from the backup software.
  2. Shut down the back up software and stop the operating system’s removable storage services. From the OCP, attempt to unload or move the tape to a slot.
  3. Power down the unit, disconnect the SCSI cable, power up the unit, and wait until the tape drive is idle or ready. From the OCP, attempt to unload or move the tape to a slot.
  4. From the OCP, attempt a force eject or emergency unload operation.
NOTE: Inspect the tape cartridge that was stuck. Damage or misplaced labels on the cartridge could have caused the load/unload failure. Discard any tape cartridge found to have issues.
Tape stuck in storage slot.
To remove a stuck tape from a storage slot:
If the operator control panel or the remote management interface is still operational:
  1. Move the tapes from the drives to the magazines using the Move Tape command.

  2. Use the magazine removal process to release the magazine and remove it from the device.

  3. Manually remove the cartridge from the magazine by inserting a finger in the hole at the back of the magazine. Some tapes need to be inserted and removed several times to condition them for free movement in and out of the magazine.

Media problems
Problem
Solution
Cleaning or data cartridge incompatible with drive.
Make sure you are using data and cleaning cartridges that are compatible with the drive and model of the device (see the Tape cartridges section in Tape Cartridges and Magazines ). The device automatically unloads incompatible cartridges, the Attention LED flashes, and an exclamation point (!) displays in the inventory display for the indicated slot number. Export the media to clear the state.
Cannot write to or read from tape.
  • Make sure that the cartridge is not a WORM cartridge that has already been used.
  • Make sure that the cartridge is write enabled (move the write-protect switch to the enabled position).
  • Make sure that the appropriate data cartridge for the model is being used.
  • Make sure that an Ultrium cartridge that has not been degaussed is being used. Do not degauss Ultrium cartridges!
  • Make sure that the cartridge has not been exposed to harsh environmental or electrical conditions and is not physically damaged in any way.
  • Many backup applications do not read or write to cartridges that were created using a different backup application. In this case, you may have to perform an erase, format, or label operation on the cartridge.
  • Make sure that you understand any data protection or overwrite protection schemes that the backup application may be using, which could prevent you from writing to a given cartridge.
  • Retry the operation with a different, known good tape.
  • Clean the tape drive.
SCSI device not detected
Problem
Solution
Device not detected
  • Check that the HBA supports multiple LUNs and this feature is enabled. If not, only the tape drive will be detected.
  • Attach the device to an LVDS SCSI host adapter/bus.
  • The SCSI cable length might be too long. Use a shorter cable or remove other devices from the bus.
  • Check for conflicting SCSI IDs.
  • SCSI bus not properly terminated (see SCSI Installation problems ).

  • Power on device before powering on the host computer.
  • Check that the device is fully powered up and is not in an error state.
  • Check the SCSI connector and terminator for bent pins.
Changed drive SCSI ID, but the host server does not recognize the new ID.
  • Make sure that all SCSI devices on the same bus have unique SCSI ID numbers.
  • If the SCSI bus is narrow (50-pin), only SCSI IDs 0 through 7 are available.
  • Make sure that you cycle power on the device after changing the SCSI ID.
  • Reboot the host server.
Performance problems
Problem
Solution
The device is not efficiently backing up data.
  • In the backup software, verify that hardware data compression is turned on and software data compression is turned off.
  • Check the bandwidth from the host computer. If you are backing up data over a network, consider comparing the performance to a local-only backup.
  • Make sure the device and tape drive are on their own SCSI bus and not daisy-chained to another tape drive, or to the hard drive being backed up.
  • Make sure the device is connected to an LVDS SCSI bus and no SE devices are on the same bus, because this causes all the devices on the bus to slow down to SE speed.
  • Use an Ultra 320 SCSI bus and high-quality cabling.
  • Do not connect the device to an arrow SCSI bus.
  • Try a new cartridge. A marginal cartridge can cause performance problems when the tape drive has to retry writing to bad spots on the tape.
  • Backing up data that compresses poorly or is already compressed lowers performance.
  • Check the size of the files. Small file size can impact performance.
  • Confirm that the backup application is using block sizes of at least 32 KB, preferably 64KB. Refer to the backup application documentation for details.
  • Make sure the backup server has enough memory to handle the bandwidth of the backup or restore.
  • Clean the tape drive (see the Tape cartridges section in Tape Cartridges and Magazines ).

Attention LED Issues
Problem
Solution
Both the Attention and Cleaning LEDs are lit.
This is most likely caused by a dirty drive that cannot read a tape and marks the tape invalid.
  1. View the inventory with the RMI. Note the slots that have tapes marked with ! .

  2. Remove any magazines that contain tapes marked with ! .

  3. Remove the tapes that were marked with ! .

  4. Inspect each removed tape for damage, check that the tape is compatible with the drive, and ensure that it is not past its usage life. Discard any tapes that are damaged or past their usage life. Do not use cartridges that are incompatible with the tape drive.

  5. Reload the magazines with tapes that have passed inspection and new tapes to replace cartridges that did not pass inspection.

  6. Replace the magazines.

  7. Clean the tape drive.

Contamination by loose debris
Avoid contamination by ensuring that the device is installed in a clean, contamination-free environment. Cartridges should be stored vertically in their plastic cases. Clean the tape drive and try the cartridge again.
A cartridge recently imported from a different environment is causing issues.
Media that is moved from one environment to another can cause issues until it has acclimated to the new conditions. A cartridge should be acclimated for at least 24 hours before being used, particularly if it has been stored at a substantially different temperature or level of humidity than the device.
The Attention LED is lit but the Cleaning LED is not lit after a cartridge load.
The Autoloader was unable to complete the requested operation with the selected tape cartridge.
  • Use only cartridges that are compatible with the drive type.
  • Use the correct type of cartridges for the operation. For example, use a cleaning cartridge for cleaning.

  • Make sure you are using an Ultrium universal cleaning cartridge (see the Tape cartridges section in Tape Cartridges and Magazines ).

Expired cleaning cartridge
A cleaning cartridge is good for 50 cleans.
A particular cartridge sets off the Attention LED and possibly the Cleaning LED.
If the Media Attention LED is cleared and the drive has been cleaned, and then immediately re-displays each time a particular cartridge is reloaded, that cartridge should be suspected as being defective.
  • If this occurs, export the cartridge and load a known good cartridge. In some cases, a cartridge can be worn out, have a defective Cartridge Memory, or have been formatted as a Firmware Upgrade Cartridge.
  • Any cartridge that is suspected of being defective or contaminated should NOT be reused in any drive.
  • If the bad cartridge is a cleaning cartridge, it might be expired.
Inventory problems
Problem
Solution
The inventory labels the cartridge Full instead of showing its bar code
  • Verify that the label is an HP label. The bar code reader might not be able to read other labels.

  • Verify that the label is properly applied.

  • Verify that the label is not soiled.

The inventory process takes a long time
Apply high-quality HP labels to all tape cartridges. During the inventory process, the bar code reader attempts to read the bar code on the cartridge or the bar code on the back of the storage slot until it identifies the cartridge or determines that the slot is empty. The reader can usually identify a properly-labeled cartridge the first time, while determining that an unlabeled cartridge is in a storage slot can take four times as long.
RMI network connection issues
Problem
Solution
RMI cannot connect to the device
  • Verify that the device is connected to the LAN with an Ethernet cable.
  • Verify that the link LED on the RJ45 (LAN) connector is lit when the device is powered up. If the LED is not lit, the device is not communicating with the LAN. See the network administrator for help.
  • Verify that the device has been configured with a valid static network address or DHCP has been enabled so the device can obtain a network address. If using DHCP, write down the device’s network address from the OCP Information menu. If the device did not obtain a valid address via DHCP, verify that the DHCP server is up and the device has network access to it. If necessary, set a static network address instead.
  • Enter the device’s IP address into the address bar of a web browser connected to the same LAN as the device. If the RMI web page does not display, ping the device’s IP address. If the ping fails, verify that the device has a valid network address and that there are no firewalls or other obstructions to network traffic between the computer with the web browser and the device. See the network administrator for help.
Cleaning problems
Problem
Solution
Cannot load the cleaning cartridge.
  • Make sure you are using an Ultrium universal cleaning cartridge (see the Tape cartridges section in Tape Cartridges and Magazines ).

  • Make sure the cleaning cartridge has not expired. A cleaning cartridge is good for 50 cleans.
  • Contact a service representative.

top

Performance problems

NOTE: For more performance troubleshooting information, see the Performance Troubleshooting Guide at www.hp.com/support/pat .

The process of backing up files goes through many devices, from the files in the file system on the disk, through the backup server, and out to the Autoloader, all managed by software running on an operating system. The backup process can only run as fast as the slowest link in this chain.

To find the performance bottlenecks in your system, check the specifications and performance of the:

  • Average file size

  • File system type

  • Connection from the host server to the disks

  • Operating system configuration

  • Backup server

  • Backup type

  • Connection from the host server to the Autoloader

  • Media

Average file size

The hard drive must seek to the position of a file before it can start reading. The more time the disks are seeking to files, the lower the performance.

To determine the average file size, divide the size of the backup by the number of files. See the performance impact of your system's average file size in Performance impact of various file sizes .

Performance impact of various file sizes
Average file size
Performance impact
Recommendations
<64 kb: small files
POOR . Lots of small files require the disk to perform many random accesses instead of a continuous read.
  • If possible, do NOT use a file-by-file backup method.

  • For backups with an average file size <64 kb, HP recommends using a sequential/image backup that backs up the hard drive or LUN image instead of the individual files.

The drawback with the sequential/image backup method is that you might only be able to restore the entire disk image and not individual files. If you can restore individual files, the restore operation will be very slow.
64 kb - 1 mb: medium files
NEUTRAL . Performance accessing medium-sized files should be okay. The disks will still need to do a fair number of random accesses.
No change is necessary but using a sequential backup method, such as an image backup, could offer some performance gains. See above for drawbacks.
>1 mb: large files
GOOD . Large files let the disk do less seeking and spend more time doing continuous reads.
None.
Image or sequential backup
None. File size is irrelevant for Image or sequential backups.
None.

top

File system type

The file system determines the organization of the files on the disks. When the files are spread over multiple disks with multiple controllers, some disks can be seeking while others are reading. Find the performance impact of your system's file system in Performance impact of various file systems

Performance impact of various file systems
File system
Performance impact
Recommendations
Disk array
GOOD . Disk arrays typically provide excellent access to data. They usually include many disks, which improves bandwidth.
None.
Server or workstation with RAID
VARIABLE . RAID uses a group of disks to improve performance, and in the case of RAID 5, provides some parity protection in case one of the drives fail.
To improve performance on a disk system, you can use faster disks or add disks. For good performance, each disk should have a sustained read speed of greater than 6 MB/s and there should be one disk per 4 MB/s of native transfer for each tape drive. The recommended number of disks per Ultrium tape drive is:
  • Ultrium 448 : 6

  • Ultrium 920 : 15

  • Ultrium 1760 : 20

Also, make sure that the server or workstation is used as little as possible during the backup, and close down programs such as virus scanners, which check each file that is opened.
Server or workstation with a single disk (spindle)
POOR . A single disk cannot retrieve or write data fast enough for any of the supported tape drives.
Upgrade to a disk array or RAID 5 with the minimum number of disks needed for your tape drives, or use concurrency or multithreading in the backup software to pull from more than one disk at a time. The drawback with concurrency or multithreading is that it slows the restore operation.
Also, make sure that the server or workstation is used as little as possible during the backup, and close down programs such as virus scanners, which check each file that is opened.

top

Connection from the host server to the disks

The connection between the host server and the disks determines how much data can be transferred from the disks to the host computer at a time. A connection with insufficient bandwidth cannot provide enough data for the tape drives to write at full speed. Find the performance impact of your system's disk connection in Performance impact of various disk connections .

Performance impact of various disk connections
Connection type
Performance impact
Recommendations
Fibre: 1 GB
  • LTO-2 : GOOD . When fully used, a 1 GB fibre connection can provide enough bandwidth for two LTO-2 tape drives.

  • LTO-3, LTO-4 : POOR . A 1 GB fibre connection is inadequate for an LTO-3 or LTO-4 tape drive.

For good performance, use a faster connection with LTO-3 and LTO-4 tape drives.
Fibre: 2 GB
GOOD . When fully used, a 2 GB fibre connection can provide enough bandwidth for the following number of tape drives:
  • LTO-2 : GOOD . When fully used, a 2 GB fibre connection can provide enough bandwidth for 4 LTO-2 tape drives.

  • LTO-3 : GOOD . When fully used, a 2 GB fibre connection can provide enough bandwidth for an LTO-3 tape drive.

  • Ultrium 1760 : GOOD . When fully used, a 2 GB fibre connection can provide enough bandwidth for an Ultrium 1760 tape drive.

For good performance, use a faster connection if you have multiple LTO-3 or LTO-4 tape drives.
Fibre: 4 GB
GOOD . When fully used, a 4 GB fibre connection can provide enough bandwidth for the following number of tape drives:
  • Ultrium 448 : 8

  • LTO-3, LTO-4 : 2

Parallel SCSI: Ultra 320
  • Ultrium 448 : GOOD . When fully used and assuming that the connection to the disks does not force the speed to a lower parallel SCSI type, an Ultra 320 connection can provide enough bandwidth for five Ultrium 448 tape drives.

  • LTO-3, LTO-4 : GOOD . When fully used and assuming that the connection to the disks does not force the speed to a lower parallel SCSI type, an Ultra 320 connection can provide enough bandwidth for a single LTO-3 or LTO-4 tape drive.

If cables not designed for Ultra 320 speeds are used or there are more than a couple devices on the bus, the bus may slow down to Ultra 160 speeds.
Parallel SCSI: Ultra 160
  • Ultrium 448 : GOOD . When fully used and assuming that the connection to the disks does not force the speed to a lower parallel SCSI type, an Ultra 160 connection can provide enough bandwidth for two Ultrium 448 tape drives.

  • LTO-3 : MARGINAL . An Ultra 160 connection is barely adequate for an Ultrium 920 tape drive. The bus would be saturated when providing 2:1 compressible data to an LTO-3 tape drive.

  • Ultrium 1760 : MARGINAL . An Ultra 160 connection is barely adequate for an Ultrium 1760 tape drive. The bus would be saturated when providing 2:1 compressible data to an Ultrium 1760 tape drive.

Upgrade to Ultra 320 for parallel SCSI disk drives when using an LTO-3 or LTO-4 tape drive, especially if your data is greater than 2:1 compressible.
Parallel SCSI: Ultra 2 (80)
  • Ultrium 448 : MARGINAL . When fully used and assuming that the connection to the disks does not force the speed to a lower parallel SCSI type, an Ultra 2 (80) connection barely provides enough bandwidth for an Ultrium 448 tape drive.

  • LTO-3, LTO-4 : POOR . An Ultra 2(80) connection is inadequate for an LTO-3 or LTO-4 tape drive.

Upgrade to Ultra 320 for parallel SCSI disk drives when using an LTO-3 or LTO-4 tape drive.
Parallel SCSI: Ultra Wide, Fast Wide
POOR . Ultra Wide and Fast Wide are inadequate transport mechanism for Ultrium tape drives.
Upgrade to a faster transport medium, such as Ultra 320 SCSI.
Ethernet: 1 Gigabit
  • Ultrium 448 : MARGINAL . A 1 Gigabit Ethernet connection barely provides enough bandwidth for an Ultrium 448 tape drive when two concurrent streams are used. The drawback of using concurrency is that restore operations take longer.

  • LTO-3, LTO-4 : POOR . A 1 Gigabit Ethernet connection is inadequate for an LTO-3 or LTO-4 tape drive.

Upgrade to Fibre Channel if using LTO-3 or LTO-4 tape drives.
Ethernet: 100 Base T, 10 Base T
POOR . 100 Base T and 10 Base T Ethernet do not have enough bandwidth to support any of the drives and would require too many streams to make concurrency practical.
Upgrade to Gigabit Ethernet or Fibre Channel.
Internal disks on backup server
POOR . Add-in RAID controllers, RAID on a chip (ROC), RAID on the mother board (ROMB), internal parallel SCSI, and internal IDE (non-RAID) are inadequate transport mechanisms for tape drives.
Use a dedicated disk array or upgrade to RAID 5.

top

Operating system configuration

The operating system configuration and other programs running on the host computer can impact the ability of the host computer to transfer files from the disks to the tape drive.

Windows

To improve backup performance and improve SAN stability, shutdown and disable the Windows Removable Storage Manager unless the backup software requires it:

  1. From the Windows start menu, select Control Panel .

  2. In the Control Panel, select Administrative Tools .

  3. In the Administrative Tools, select Services .

  4. In Services, right-click Removable Storage and select Stop . The Status should be blank as shown in Removable Storage service (Figure 1) .

    Figure 1: Removable Storage service

  5. Right-click Removable Storage again and select Properties .

  6. Change the Startup type to Disabled and click OK .

Stop as many programs as possible while a backup or restore is in progress. Be especially aware of monitoring applications, such as anti-virus software, which check every file that is read. If you are concerned about viruses and backup performance, perform a virus scan before starting the backup, and shut down the anti-virus software while the backup is running.

top

Novell

You might need to tweak some Novell operating system parameters to achieve good backup performance.

  • The MAXIMUM PACKET RECEIVE BUFFERS and MINIMUM PACKET RECEIVE BUFFERS parameters may need to be increased (if currently maxed out, try setting the MINIMUM to the current MAXIMUM and then double the value for the MAXIMUM on the backup server and any remote servers that are backed up).

  • The MAXIMUM DIRECTORY CACHE BUFFERS and MINIMUM DIRECTORY CACHE BUFFERS parameters may need to be increased (if currently maxed out, try setting the MINIMUM to the current MAXIMUM and then double the value for the MAXIMUM on the backup server and any remote servers that are backed up.)

  • Set TCP DELAYED ACKNOWLEDGEMENTS to OFF.

  • Since Novell volumes may be compressed, hardware compression may lower the performance and capacity. If the volumes are compressed, make sure that the backup software does not uncompress the data on read and has hardware and software compression disabled.
    NOTE: Disabling the hardware compression should not be necessary on Ultrium drives as they sense the compression ratios and can automatically adjust if they receive non-compressable data.
  • Stop as many programs as possible while a backup or restore is in progress. Be especially aware of monitoring applications, such as anti-virus software, which check every file that is read. If you are concerned about viruses and backup performance, perform a virus scan before starting the backup, and shut down the anti-virus software while the backup is running.

top

top

Backup server

The backup server must have enough RAM and processor power to transfer the files from the disk to the tape drive. See Backup server requirements for the minimum RAM and CPU speed needed for each tape drive. In some cases, a multi-processor server may be required.

NOTE: Processor speed numbers as based on Intel x86 type processors. Use the equivalent on a RISC or other type of processor.
Backup server requirements
Tape drive
Minimum RAM/drive
Processor power/drive
LTO-2
512 MB
1 GHz
LTO-3, LTO-4
1 GB
2 GHz

In addition to having enough RAM and processing power, ensure that the PCI bus is at least 64 bit, has a speed of 66 MHz or better, and is not overloaded by too many high-bandwidth cards. PCI-X is preferred.

top

Backup type

Each type of backup has its own impact on performance, depending on how well it can keep data streaming to the tape drive.

File-by-file with a full-featured backup application

Performance impact: VARIABLE .

File-by-file backup with a full-featured backup application can be fast enough if the average file size is at least 64 k and there are not too many fragmented files. Full-featured backup applications also offer the best speeds for single file restores and allows for backing up only specific data. Check the compatibility matrix on the EBS website for a list of full-featured backup applications supported on your operating system for your Tape Autoloader: www.hp.com/go/ebs .

If the average file size is less than 64 kb or if the file system is very fragmented, file-by-file backup will have poor performance. If the file system has a lot of fragmentation, use a de-fragmentation utility to make the files contiguous again. If the average file size is less than 64 kb, HP recommends that you use a sequential/image backup that backups the hard drive or LUN image instead of the individual files. The drawback with sequential/image backups is that they may only be able to restore the entire disk image and not individual files. If they can restore individual files, the restore operation will be very slow.

top

File-by-file with a native application

Performance impact: POOR .

Native backup applications based on tar, cpio, NT Backup, etc. do not have the extra features needed to manage the bandwidth requirements of the faster tape drives and should only be used to test basic functionality.

To get the best backup and restore performance, use a full-featured backup application. If the average file size is less than 64 kb, use a sequential/image backup for best performance. However, a sequential/disk image backup might not allow you to restore individual files or the restore process will be very slow.

top

Disk image, flash, or sequential

Performance impact: GOOD .

A disk image or sequential backup backs up an entire disk, partition, or LUN by looking at the disk sector by sector instead of file by file. The entire disk contents is backed up contiguously, without the disk seeking, which prevents performance degradation caused by small or fragmented files.

If you are more concerned about backup performance than single-file restore, disk image or sequential backups can offer a real performance benefit. The disadvantage is that backup and restore operations work on an entire disk, partition, or LUN. You might not be able to backup a subset of files or restore a single file. If you can restore a single file, the restore process will be slow.

top

Database backup

Performance impact: VARIABLE .

To improve performance when backing up data from a database:

  • Use specific backup agents for the database.

  • Use the latest versions of the databases.

  • Do not backup individual mailboxes.

  • Do not backup specific records or do a record-by-record backup.

  • Do not backup when the database is in heavy use.

top

top

Connection from the host server to the Autoloader

For the best performance, the connection from the host server to the Autoloader must have enough bandwidth to provide enough data to keep the tape drive streaming.

NOTE: Unlike most tape technologies, an LTO tape drive can write data as fast as the server can send it, even if the tape drive is not streaming.

Find the performance impact of your system's Autoloader connection in Performance impact of various Autoloader connections

Performance impact of various Autoloader connections
Connection type
Performance impact
Recommendations
Parallel SCSI: Ultra 320
GOOD . When fully used, an Ultra 320 SCSI connection can provide enough bandwidth for the following number of tape drives, assuming 2:1 compression:
  • LTO-2 : 2

  • LTO-3, LTO-4 : 1

Ultra 320 is an excellent transport medium, but only has enough bandwidth for one LTO-3 or LTO-4 drive per bus.
If the parallel SCSI cables are not designed for Ultra 320 speeds or more than a couple devices are on the parallel SCSI bus, the bus may slow down to Ultra 160 speed.
Parallel SCSI: Ultra 160
  • LTO-2 : GOOD . When fully used and assuming that the connection to the disks does not force the speed to a lower parallel SCSI type, an Ultra 160 connection can provide enough bandwidth for two LTO-2 tape drives.

  • LTO-3 : MARGINAL . An Ultra 160 connection is barely adequate for an LTO-3 tape drive. The bus would be saturated with 2:1 compressible data.

  • LTO-4 : POOR . An LTO-4 tape drive is unable to sustain maximum transfer speeds with 2:1 compressible data.

Ultra 160 is a good transport medium, but does not have enough bandwidth for LTO-3 or LTO-4 tape drives. If using LTO-3 or LTO-4, upgrade to an Ultra 320 HBA, especially if you have greater than 2:1 compressible data.
Parallel SCSI: Ultra 2 (80)
  • LTO-2 : MARGINAL . Ultra 2 (80) barely has enough bandwidth for one LTO-2 drive.

  • LTO-3, LTO-4 : POOR . An Ultra 2 (80) connection is inadequate for an LTO-3 or LTO-4 tape drive.

Upgrade to Ultra 320.
Parallel SCSI: Ultra Wide, Fast Wide
Ultra Wide and Fast Wide are inadequate for all of the currently supported tape drives.
Upgrade to Ultra 320.

top

Media

Once the data gets to the tape drive, it must be written onto the tape. If there are no bottlenecks and the Autoloader starts having performance problems:

  • Try a new cartridge. A marginal cartridge can cause performance problems when the tape drive has to retry writing to bad spots on the tape.

  • Clean the tape drive.

top

top

Service and repair

Removing stuck tapes from slots

NOTE: Contact an HP support representative if a cartridge is stuck in the drive.

In the event of a severe mechanical problem or if circumstances requires removing tape cartridges, perform the following:

If the operator control panel or the remote management interface is still operational:

  1. Move the tapes from the drives to the magazines using the Move Tape command or the Force Drive to Eject command.
  2. Use the magazine removal process to release the magazine and remove it from the device. To use the operator control panel.

top

Releasing the magazines manually

If the tape cartridge cannot be removed from the OCP or RMI, do the following:

  1. Unplug the power cord from the device.
  2. From the back of the device, find the access holes for the right and left magazines (see Figure 2).

    Figure 2: Access holes for the left and right magazine

    1 - Right magazine release
    2 - Left magazine release
  3. To manually release a magazine, push the end of a straightened paper clip into the access hole for a magazine at the back of the device. While holding the paper clip, have a second person attempt to pull the magazine out of the front of the unit (see Figure 3).

    Figure 3: Releasing the magazine

    1 - Pushing a paper clip into the access hole
    2 - Pulling the magazine out of the front of the unit
  4. Repeat Step 4 for the other magazine if necessary.
  5. If other tapes are still in the device, or if it is unable to manually remove the magazines and drive, contact HP service for further instructions.

top

top

Error codes

If an error occurs during operation of the device, the device stops the current operation and displays an error code on the LCD screen. Unless otherwise listed in Operation problems , try to resolve the error by cycling power to the device and retrying the operation.

To check the overall operation of the device, run the wellness test from the RMI or OCP. The wellness test exercises all robotic movements and checks the status of the electrical components and communication.

If the error persists, contact support personnel.

There are three ways to obtain error codes from the device:

  • On the OCP

  • On the RMI

  • On an L&TT support ticket or report

Finding error code information on the OCP

When an error first occurs, the error message and error code are displayed on the OCP, as shown in Initial OCP error message (Figure 4) .

Figure 4: Initial OCP error message

The code 9B is the main error code, and 37 is the error sub-code.

If you review the Error Log in the Support menu, the OCP error log displays the error code, as shown in Error code in the OCP Error Log (Figure 5) .

Figure 5: Error code in the OCP Error Log

The code 9B is the main error code, 37 is the error sub-code, and 31 is sub-code specific information for factory use only. When you press Enter , the OCP displays the error message, as shown in Error message in the OCP Error Log (Figure 6) .

Figure 6: Error message in the OCP Error Log

If you press Next , the OCP will display additional information, if available, as shown in No additional information in the OCP error log (Figure 7) .

Figure 7: No additional information in the OCP error log

If you press Next again, the OCP will display the date and time in the format: YYYY-MM-DD followed by the time in 24–hour clock format, where 1:00 pm is 13:00, as shown in Date and time in the OCP error log (Figure 8) .

Figure 8: Date and time in the OCP error log

top

Finding error code information on the RMI

You can find error codes on the RMI Support: Autoloader logs page (Figure 9) . The available logs are: Error Trace, Informational Trace, Warning Trace, Configuration Change Trace, and Standard Trace.

The log entries are displayed in order of most recent to oldest. The format for the log entries is: YY.MM.DD HH.MM.SS.ss LIB/ERR<80 89 62 40

  • YY.MM.DD - the date displayed as Year.Month.Day

  • HH.MM.SS.ss - the time displayed as Hour.Minute.Second.Hundredths fo a second

  • First code - hard or soft error. The code after LIB/ERR (80 in the example) will be 80 or 40. 80 indicates a hard error, 40 indicates a soft error.

  • Second code - the main error code (89 in this example). See Error codes for a list of error codes and recovery procedures.

  • Third code - the sub-code (62 in this example). See Device error sub-codes for a list of sub-codes.

  • Fourth code - sub-code-specific information for factory use only

Figure 9: Support: Autoloader logs page

top

Finding error code information on an L&TT support ticket or report

An L&TT support ticket or report contains detailed information about the device configuration, along with errors and warnings. The support ticket and report contain the same information. The report is easier to read, but must be generated and read on the host computer. The support ticket can be downloaded from the device and then viewed on any computer with L&TT installed.

To generate and view a report or support ticket from L&TT:

  1. In the L&TT By Product or By Connection tab, select the device from the device list.

  2. Click the Health button on the main toolbar to generate and display a standard report. Or click the Support button on the main toolbar to display the Support screen for additional report or support ticket options.

To download a support ticket from the device, do one of the following:

  • From the RMI Support > Support ticket screen, click Download .

  • Insert a USB flash drive into the USB port on the rear panel and then from the OCP, select Download support ticket to USB .

NOTE: The support ticket downloaded from the RMI will only contain information about the Autoloader. To download a support ticket containing drive information, download the support ticket with L&TT.

To view a downloaded support ticket:

  1. From the L&TT File menu, select Load Support Ticket .

  2. Select the support ticket file in the browser.

The top of the support ticket contains basic configuration information about the device, as shown in Support ticket in viewer (Figure 10) .

Figure 10: Support ticket in viewer

Expand HP Event Logs to see events divided into three categories:

  • Events in the last 24 hours

  • Events in the last 31 days

  • Events older than 31 days

Set the Current Detail Level to see additional types of events:

  • Normal will only show critical events or hard errors.

  • More details will also show warning and configuration events.

  • Everything shows all events.

Critical events are designated with a STOP sign icon. Expand the event for more information, as shown in Critical event details (Figure 11) .

Figure 11: Critical event details

  • The time stamp is in the format hours : minutes : seconds. The hours are in 24-hour clock format. For example, in this case 14
    is 2 p.m.

  • The date is in the format year/month/day.

  • The event ID is the number on the header line, 0x006E
    in this example. It uniquely maps to an error code.

  • HE
    designates a hard error. The STOP sign icon and the word Crit
    before the event ID also indicate a hard error.

  • The text description in the header (“robotic controller error” in this example) is the simple text description of the main error code.

  • The main error code (0x83)
    is displayed in parenthesis as the Global error code . The error codes are described in Main error codes . The text after the main error code (Robotic controller generic problem
    in this example) is the text description for the error code.

  • The error sub-code (0x02)
    is displayed in parenthesis as the Module error code . The error sub-codes are described in Error sub-code descriptions . The text after the error sub-code (Robotic: connection to slave robotic failed
    in this example) is name of the component followed by the text description of the error sub-code.

  • The Current command provides information for factory use only.

top

Main error code descriptions

The following table lists the main error codes with details and a solution.

Here is a brief description of a few parts listed in the table below:

Gripper : The part of the robotics assembly that pinches media in order to grip it.
Slider : The part of the robotics assembly that plunges in and out for get and put operations.
Elevator : The part of the robotics assembly that moves in the vertical direction.
Rotation : The part of the robotics assembly that turns the robot to face each magazine and the drive.
Sled : The part of the robotics assembly that moves the robot towards the OCP or back towards the drive.

Main error codes
Error code
Description
Details and solution
80
Cannot initialize bar code reader
Power-cycle the unit and retry the operation
81
No response from bar code reader
82
No response from EEPROM on robotic controller
83
Robotic controller generic problem
84
Setting of gripper motor parameters failed
85
Setting of slider motor parameters failed
86
Setting of elevator motor parameters failed
87
Setting of rotation motor parameters failed
88
Setting of sled motor parameters failed
89
Gripper obstructed
Run the wellness test
8A
Slider obstructed
8B
Elevator obstructed
8C
Rotation obstructed
8D
Sled obstructed
  • If this error occurs on the first power-on after unpacking or moving the device, or after replacing the chassis, ensure that the shipping lock was removed from the top and stored on the back panel.
  • If the device was moved or shipped with tape cartridges in the magazines, verify that the cartridges did not come out of the magazines to obstruct the drive.
  • Ensure that nothing is stacked on top of the device because any weight on top of the device can bow the top cover down and interfere with the robotics. If the Autoloader has a tabletop conversion cover, verify that no more than 15 kg (33 lbs.) is on top of the cover.
  • Ensure that the device is either mounted in a rack with its rack kit or in a tabletop conversion cover. If the Autoloader is not mounted in a rack or in the tabletop conversion cover, ensure that the six plastic feet are installed and that all of them are standing on a flat surface. The device must be supported under both of its side edges or the bottom can bow and impede robotic movement.
  • Run the wellness test.
8E
Ends of gripper movement not in expected range
Run the wellness test
8F
Ends of slider movement not in expected range
  • If the error consistently happens on the same slot, try different tape cartridges in that slot.
    • If the failure remains with the same slot, the magazine is probably at fault. Replace the magazine.
    • If the failure follows the tape cartridge, the tape cartridge is probably at fault. Replace the tape cartridge.
  • If you are doing an operation that automatically returns a tape cartridge to a certain slot, make sure another tape is not loaded in that slot.
  • Run the wellness test
90
Ends of elevator movement not in expected range
Run the wellness test
91
Ends of rotation movement not in expected range
92
Ends of sled movement not in expected range
93
Gripper reached a position beyond expected range
94
Slider reached a position beyond expected range
95
Elevator reached a position beyond expected range
96
Rotation reached a position beyond expected range
97
Sled reached a position beyond expected range
98
Cartridge present sensor not found
99
Slider home sensor not found
9A
Rotation home sensor not found
9B
Sled position sensor not found
9C
Gripper range of motion out of specification
9D
Slider range of motion out of specification
9E
Elevator range of motion out of specification
9F
Rotation range of motion out of specification
Run the wellness test
A0
Sled range of motion out of specification
A1
Open Mail Slot failed
  • Retry the operation
  • If not successful, contact technical support
B0
Command from Autoloader controller to robotics controller did not complete in time allotted
  • Reset the unit
  • Retry the operation
  • If not successful, contact technical support
B1
Robot controller reported format error on command from Autoloader controller
B2
Communication to robot from Autoloader controller failed
B3
Robot stopped due to a released magazine
  • Check whether the magazine is completely inserted
  • Retry the operation
  • If not successful, contact technical support
B4
“Tape in gripper” sensor did not report the expected value
  • Remove and inspect the magazine or drive for a stuck tape
  • Run the wellness test
B5
Robotic controller not responding to command from Autoloader controller
  • Reset the unit
  • Retry the operation
  • If not successful, contact technical support
C0
Network initialization failed
  • Check network cable
  • Check the network configuration
  • Power cycle the unit
C1
Telnet interface initialization failed
C2
Web server initialization failed
C6
Ping command did not reach target
C7
Cannot upgrade firmware from USB
  • Retry firmware upgrade, if not successful,
  • Attempt a different upgrade method
D0
ROM checksum incorrect
Retry firmware upgrade
D1
RAM error during Power-On-Self-Test
Power-cycle the unit
D2
Read or Write to NVRAM on Autoloader controller failed
D3
Time controller failed during Power-On-Self-Test
D4
Internal UART serial communication error
D5
Communication to display failed
D6
Autoloader controller memory error
D7
Fatal system error
D8
Autoloader controller data base error
D9
No SCSI IC detected
DA
When running the wellness test, the bar code did not match the previous value for that tape
  • Check the bar code label
  • Run the wellness test again
DB
External cooling fan error.
Check if the fan is operational and not obstructed
F0
Drive exceeded temperature specification
Check ambient temperature conditions and fans
F1
Autoloader controller lost communication with the drive
Re-seat the tape drive
F2
Drive sled not present
F3
Drive hardware error
  • Cycle power, after several occurrences,
  • Contact technical support
F4
Time allotted for drive loading exceeded
  • Retry the operation, if not successful,
  • Contact technical support
F5
Time allotted for drive unloading exceeded
  • Attempt Force Tape Eject test
  • Inspect tape for defects after unloading

top

Drive error codes

The drive error codes are displayed on the RMI drive status page.

Drive error codes
Error code
Description
01
Drive broken
02
Temperature exceeds limit
03
Tape error
04
Cleaning cartridge is expired
05
Drive needs cleaning
06
Autoloader lost communication with the drive
07
Warning that the tape is nearing its end of life

top

top

Error sub-code descriptions

Robotic error sub-codes

The following table lists the robotic error sub-codes.

Robotic error sub-codes
Sub-code
Description
01
Mechanical initialization failure
02
Connection to slave robotic failed
03
Error motor initialization
04
Error during gripper close
05
Error slider home positioning
06
Error elevator home movement
07
Error during sled movement to rotation4 position
08
Error during rotation initialization, get range failed
09
Error elevator init
0A
Error during rotation to far position
0B
Error first sled init, move to sensor failed
0C
Error during sled movement to rotation4 position
0D
Error during rotation to slide position
0E
Error slider init, get range failed
0F
Error during slider forward movement
10
Error gripper init, get range failed
11
Error during slider home movement
12
Error during rotation to FAR position
13
Error sled init, move to sensor failed
20
Error inventory scan
21
Error during gripper close
22
Error slider home movement
23
Error during move gripper to scan pos
24
Error reading bar code label
28
Error Extra inventory scan
29
Error during closing gripper
2A
Error slider preposition movement
2B
Error during opening gripper
2C
Error during sled movement up to sensor
2D
Error slider preposition backwards movement
30
Error slot preposition
31
Error during sled movement in FLMoveRotation
32
Command sending to robotic failed
33
Error during elevator movement in FLMoveRotation function
34
Error during rotation in FLMoveRotation function
35
Error during elevator movement in FLMoveRotation function
36
Error during sled movement in FLMoveSled function
37
Error during sled positioning to mail slot in FLMoveSled function
38
Error during sled positioning to mail slot in FLMoveSled function
39
Error during sled positioning without sensor
40
Movement to/from slot failed
41
Error during first slider movement
42
Error during first gripper movement
43
Error during second slider movement
44
Error during second gripper movement, get range failed
45
Error during third slider movement, move home failed
50
Preposition to drive failed
51
Elevator movement to home sensor failed
52
Sled movement to home sensor failed
53
Error during sled movement to drive position
54
Error during rotation to drive position
55
Error during elevator movement in drive position
60
Move from/to drive failed
61
Error during first slider movement
62
Error during first gripper movement
63
Error during second slider movement
64
Error during second gripper movement, get range failed
65
Error during third slider movement, move home failed
70
Release magazine failed
71
Error during sled movement to rotation4 position
72
Error during rotation to unlock position
73
Error during move sled to block
80
Opening mail slot failed
82
Error during moving back, sensor was found
90
Movement to home position failed
91
Elevator movement to home position failed
92
Error during sled movement to rotation position
93
Error during rotation to home or far position
94
Sled movement to home sensor position failed
95
Sled movement to transport position failed
A0
Movement to mail slot failed
A1
Sled movement to sensor failed
A2
Sled movement to rotation position failed
A3
Elevator movement to home position failed
A4
Error during rotation to far position
A5
Sled movement to mail slot position failed

top

Device error sub-codes

The following table lists the device error sub-codes.

Device error sub-codes
Error code
Description
81
Drive wake up failed
88
Error accessing slot status
90
Robotic load not reached Cartridge Present sensor
91
No activity after Load command
92
Time-out while loading tape
93
No activity after load command
94
Time-out drive Unload
95
Drive terminated unsuccessfully
96
Tape not ejected at robot unload
97
Slot not free at robot unload

top

top

Warning events

Warning event codes
Event code
Description
Details and Solution
60
Cleaning tape installed.
Complete the cleaning process and retry the operation.
61
Cleaning failure. Cleaning process could not be performed.
A data cartridge may have been loaded instead of a cleaning cartridge.
  • Check cleaning tape and exchange if necessary.

  • Retry the operation.

62
Cleaning tape expired.
Exchange the cleaning tape.
63
Invalid cartridge. Drive has rejected the cartridge as invalid.
  • Check the cartridge for damage. It could be a defective cartridge.

  • Make sure that the cartridge is compatible with the drive.

  • Verify that the cartridge is not a firmware upgrade tape. Initiate the firmware upgrade process before loading a firmware upgrade tape into the drive.

64
Invalid cleaning cartridge. Drive has rejected the cartridge as invalid.
  • Verify that the cartridge is a cleaning cartridge.

  • Check the cartridge for damage. It could be a defective cartridge.

  • Make sure that the cartridge is compatible with the drive.

  • Verify that the cartridge is not a firmware upgrade tape. Initiate the firmware upgrade process before loading a firmware upgrade tape into the drive.

65
Invalid upgrade cartridge. Drive has rejected the cartridge as invalid.
  • Verify that the cartridge is a firmware upgrade cartridge, not a data or cleaning cartridge.

  • Check the cartridge for damage. It could be a defective cartridge.

  • Make sure that the cartridge is compatible with the drive.

DB
External cooling fan error (fan motion has stopped). The subcode indicates which drive sled fan is affected.
Subcode 00: drive sled #1
Subcode 01: drive sled #1
  • Verify that the fan for the indicated fan is operational and not obstructed.

  • If the problem continues, contact HP customer support.

DD
Power supply fan has failed, redundancy may be at risk.
  • Verify that the fan for the indicated fan is operational and not obstructed.

  • Check ambient temperature conditions.

  • Upgrade the Autoloader firmware to the latest release.

  • If the problem continues, contact HP customer support.

DE
Power supply has failed. Redundancy is not available.
  • Verify that the power supply is inserted correctly and that the thumbscrews are adjusted.

  • Power cycle the unit.

  • Upgrade the Autoloader firmware to the latest release.

  • If the problem continues, contact HP customer support.

top

Using HP Library & Tape Tools to diagnose problems

With HP Library & Tape Tools installed on the host server, the following tasks can be accomplished:

  • Identify all SCSI devices connected to the system.
  • View detailed configuration, identification, inventory, and drive information for the device.
  • Easily update device and drive firmware.
  • Run advanced diagnostic tests, including connectivity, read/write, media validation, and testing the functionality of the device.
  • View device and drive error logs.
  • Generate a detailed support file that can be e-mailed or faxed to a support representative for analysis.

The HP Library & Tape Tools diagnostic provides an intuitive graphical user interface with integrated context-sensitive help. It can be downloaded free of charge from http://www.hp.com/support/TapeTools .

top

Provide feedback

Please rate the information on this page to help us improve our content. Thank you!