Chapter 2. IBM Linear Tape File System Library Edition

Search in book...
Toggle Font Controls
Create new playlist

Name your new playlist

Playlist description (optional)
Sign In

Email address

Password

Forgot Password?

or

Continue with Facebook

Continue with Google
Sign Up

Full Name

Email address

Confirm Email Address

Password

or

Continue with Facebook

Continue with Google

IBM Linear Tape File System Library Edition

IBM Linear Tape File System Library Edition (LTFS LE) extends the file manager capability of the IBM LTFS Single Drive Edition (LTFS SDE). LTFS LE is introduced with Version 2.0 of LTFS.

LTFS is the first file system that works with IBM System Storage tape technology to optimize ease of use and portability for open-systems tape storage. It manages the automation and provides operating system-level access to the contents of the library. LTFS LE is based on the LTFS format specification, enabling tape library cartridges to be interchangeable with cartridges that are written with the open source SDE version of LTFS. IBM LTFS LE supports most IBM tape libraries:

•TS2900 Tape Autoloader, TS3100 Tape Library, and TS3200 Tape Library

•TS3310 Tape Library

•TS3500 Tape Library

LTFS LE enables the reading, writing, searching, and indexing of user data on tape and access to user metadata. Metadata is the descriptive information about user data that is stored on a cartridge. Metadata enables searching and accessing of files through the graphical user interface (GUI) of the operating system. From LTFS V2R1.2, LTFS LE supports both Linux and Windows.

Did you know? IBM Linear Tape File System (LTFS) won a Pick Hit award at the National Association of Broadcasters (NAB) conference.

Ordering: LTFS LE is a Software License Product. LTFS LE has a Program PID number of 5639-LT1.

2.1 Introduction

LTFS LE is a file system that enables direct access to IBM Linear Tape-Option (LTO) generation 5, IBM LTO generation 6, and IBM TS1140 tape drives and tape libraries. LTFS LE eliminates the need for additional tape management software to access data. LTFS LE is based on the media partitioning feature that is introduced with the IBM Linear Tape-Option Ultrium 5 (LTO-5) and TS1140 tape drives.

LTFS is the first file system that works with LTO generation 5 tape technology to set a new standard for ease of use and portability for open systems tape storage. With LTFS, accessing data that is stored on an IBM tape cartridge becomes easy and intuitive. With LTFS and the operating system’s graphical file manager (Figure 2-1), reading data on a tape cartridge is as simple as dragging and dropping. In addition, it provides for caching tape indexes. It also provides for searching, querying, and displaying tape content within an IBM tape library without the requirement to mount tape cartridges.

Figure 2-1 IBM LTFS LE

Under LTFS, tape cartridge usage changes from being application-dependent to application independent. The tape cartridge is now a self-describing, independent storage object similar to a USB drive. Each LTFS cartridge contains the metadata to enable access to the user data on that cartridge. Tape cartridges are portable and robust, and far less prone to damage than a USB hard disk drive. They are also much smaller, lighter, and less expensive. Because nothing is ever overwritten on the tape cartridges, the tapes contain not only the current version of files, but also every previous version of every file that is written to the tape. And the metadata holds all of the necessary information to restore to previous “versions” of the tape.

LTFS LE provides the following product features:

•Direct access and management of data on tape libraries with LTO Ultrium 5 (LTO-5) tape drives, LTO Ultrium 6 (LTO-6) tape drives, and TS1140 tape drives

•Tagging of files with any text, allowing more intuitive searches of cartridge and library content

•Exploitation of the partitioning of the media in LTO-5 tape format standard

•One-to-one mapping of tape cartridges in tape libraries to file folders

•Capability to create a single file system mount point for a logical library that is managed by a single instance of LTFS and runs on a single computer system

•Capability to cache tape indexes and to search, query, and display tape content within an IBM tape library without having to mount tape cartridges

The following components are needed to operate the tape file system:

•Software in the form of LTFS LE

•Hardware, consisting of IBM LTO Ultrium 5 tape drive, IBM LTO Ultrium 6 tape drive, and media or IBM TS1140 Tape Drive and media

•On-tape data structures (created by LTFS)

Together, these components can handle a file system on LTO media as though it is a disk file system for accessing tape files, including the tape directory tree structures. After an LTFS tape is mounted and recognized by the LTFS software, its contents can be displayed quickly (that is, at the start of the tape) without a special backup or archive application. Data movement is application independent.

Because each LTFS formatted tape is self-describing, it holds all the needed information to determine the data contents on tape, including the instructions for data arrays, file elements and attributes, and the use of the metadata. An LTFS-formatted tape can be easily transported to other physical locations for reading and writing data.

The LTFS format is an ideal option for storing, on a long-term basis, large files that need to be shared with others, such as for the digital media and entertainment market. Most of these users use Linux or Macintosh systems and can see all the files on tape. With LTFS, users can recall any file from the tape without needing to read the entire tape.

The following sections provide a more detailed view of the design and implementation of the new tape file system:

•For index and data partition information, see 2.1.5, “Index and data partitions” on page 109.

•For Filesystem in User Space (FUSE), see 2.1.6, “Filesystem in Userspace (FUSE)” on page 111.

•For on-tape data structures that use an XML schema to record the contents of the tape, see 2.1.7, “XML schema” on page 112 and 2.1.8, “Recording the LTFS XML schemas” on page 120.

On-tape data structures: These structures are defined by the W3C Extensible Markup Language (XML) 1.0 standard.

2.1.1 LTO media partitioning

The new functionality of media partitioning is added to the LTO-5, LTO-6, and IBM TS1140 tape drives. Media partitioning is the logical and physical division of the tape into two segments to allow for faster data access and to update partitions independently.

In the LTO-5 implementation, two partition modes are supported:

•One partition on the full tape

•Two partitions with a variable size for each partition

Partitions on media all have the same beginning-of-tape (BOT) and end-of-tape (EOT) markers, as shown in Figure 2-2. The minimum partition size of the two partitions is equivalent to two wraps.

Figure 2-2 A logical view of media partitioning

The LTO-5 media default that is set at manufacturing is a single partition of 80 wraps. The host application or middleware can change the partition to one partition (partition 0, the default partition) or two partitions, where the second partition is called partition 1. Partitioning is done by using a reformat command that logically erases all data on a tape before it is partitioned.

Each LTO-5 partition (if there are two) has a bidirectional serpentine and must have an even number of wraps in size. This way, it begins and ends at Logical-Point 3 (LP3) (logical point to indicate the start of the user data region; see Figure 2-2). If there are two partitions, a guard band is between the two wraps, which is 2.5% of nominal LTO-5 cartridge capacity.

LTO-6 and TS1140 support the new partitioning scheme. By using the new scheme, LTFS can provide more space for data partitioning. Currently, LTFS V2R1.2 does not support the new scheme. The principle function of the media partitioning is to allow for faster data access by splitting the cartridge into up to four media partitions. Tape media can have a minimum of one and a maximum of four different partitions that are defined. Partitions on media all have the same BOT and EOT as shown in Figure 2-3. The minimum partition size of the multiple partitions is equivalent to two wraps.

Figure 2-3 Expanded Media Partitioning

Set capacity and partitioning are independent of each other. Partitioned media can be capacity scaled and capacity-scaled media can be partitioned. Each format media command causes a loss of data on media.

Important: The new partitioning scheme is applied when a cartridge is formatted. So, the user can read from and write to cartridges that are formatted by an older version of LTFS. LTFS V2R1.2 can read from and write to existing cartridges with the old partitioning scheme.

LTFSV2R1.2 does not provide the functionality to format with the “old” partitioning scheme. So, LTFS V2R1.2 can format with only the new partitioning scheme.

Encryption is supported by using one key for the entire cartridge in transparent encryption.

Write Once Read Many (WORM): Because WORM media is not partitionable, it is not supported in LTFS.

2.1.2 Interoperability

LTO Ultrium cartridges from earlier LTO generations (that is LTO-1 through LTO-4) are not partitionable. Those formats are not supported for partitioning. Figure 2-4 shows the interoperability matrix of media partitioning and LTO-5 cartridges.

Figure 2-4 Interoperability matrix for media partitioning

If LTO Ultrium 4 (LTO-4) cartridges are used in an LTO-5 drive to write data on, the LTO-4 cartridge is treated like an unpartitioned LTO-5 cartridge. Even if an application can work with partitions, it is not possible to partition the LTO-4 media that is mounted on an LTO Ultrium 5 drive.

2.1.3 Applications that use partitioning

To increase the usability of a tape drive system, a new tape file system interface acts similarly to any other storage disk and removable flash media, such as a USB disk drive. This file system, called the LTFS, is a self-describing tape format (SDTF) and open source file system software. Combined with LTO-5 partitioning technology, it brings functionality for easy handling of the tape cartridges.

To use the new tape file system, the file that is using the application can use middleware (such as LTFS) to handle the files. Alternatively, if the application is SDTF-capable, only the SCSI standard commands can be used (see Figure 2-5). Partitioning commands are already SCSI standard commands, but most tape applications that are written today do not use them.

Figure 2-5 Application that uses partitioning, SDTF, and LTFS

2.1.4 LTFS tape format specification

LTFS tape format compatibility for versions 2.1.1 is described. The LTFS tape format version is updated from 1.0 to 2.1.1. See the following LTFS format specification website for a PDF file of the latest version:

http://www.ultrium.com/technology/LTFS_Format_To%20Print.pdf

The older version of the LTFS SDE program (1.0.1.xx or older for Linux and Mac OS X, and 1.1.0 for Microsoft Windows) supports only LTFS format version 1.0. The latest version of LTFS SDE (1.2.0.xx or later) and LTFS LE supports LTFS format version 2.1.1. Table 2-1 shows a compatibility matrix for media versions 1.0 and 2.1.1.

Table 2-1 Media version compatibility

LTFS LE version	LTFS 1.0 medium	LTFS 2.1.1 medium
2.0 Linux only	N/A	Read/write
2.1.1 Linux only	N/A	Read/write
2.1.2 Linux and Windows	Read/writea	Read/write¹

¹ The user can create, delete, and move the symbolic link file except on Windows.

Important: For LTFS LE, the mkltfs and ltfsck functions are provided by the ltfsadmintool command. The ltfsck utility in LTFS SDE checks and recovers media. The mkltfs utility in LTFS SDE formats media.

The ltfsck command in the older LTFS package cannot be used to check and recover a medium that is formatted to version 2.1.1. A medium that is formatted to version 2.1.1 cannot be converted to a 1.0 version format. If you use the mkltfs command from a 1.0.x LTFS version, you reformat the medium to a 1.0 format.

In summary, media that is written by LTFS SDE (LTFS 2.1.1 format) can be read by LTFS LE and vice versa.

2.1.5 Index and data partitions

LTFS uses the capability of LTO Generation 5 technology to partition the LTO media into two segments. The first partition is used to store the self-describing LTFS format based on an XML schema that contains the client data in the second partition (see Figure 2-6 and Figure 2-7).

Figure 2-6 Logical view of LTFS volume

One partition (partition 0), which is called the index partition, is small and allocates two LTO-5 wraps with about 37.5 GB on native storage capacity. The remaining cartridge capacity is used by partition 1, which is called the data partition. The data partition consists of 76 LTO-5 wraps with about 1425 GB of native storage capacity. Both partitions are separated by the two guard wraps as shown in Figure 2-2 on page 106.

Figure 2-7 Index and data partitions on tape

After LTFS formats a tape media, the index and data partitions contain a VOL1 label and an LTFS label. The LTFS label is based on XML schema (see 2.1.7, “XML schema” on page 112) and contains information about the LTFS partition on which the structure is stored.

VOL1 label: The VOL1 label is created according to ANSI Standard X 3.27. It contains ASCII-encoded characters that are 80 bytes in length.

Figure 2-8 on page 110 shows the LTFS-formatted tape without any client data on it. The label construct in the index and data partitions consists of the VOL1 and LTFS labels.

Figure 2-8 LTFS tape format after initialization (FM=file mark)

An LTFS volume is defined as a pair of LTFS partitions (one index partition and one data partition) that contain a logical set of data files and directories. The pair of partitions in an LTFS volume must have the same Universally Unique Identifier (UUID).

Figure 2-9 shows the content of a simple LTFS volume. This volume contains three files: File A, File B, and File C.

File A consists of three extents. Files B and C have one extent. The file blocks are followed by a file mark only when an XML index immediately follows the files. Figure 2-9 also indicates that the LTFS volume is still mounted. After you unmount an LTFS volume, double file marks are written at the end of the index and data partitions.

Figure 2-9 A simple LTFS volume

TS1140 new partitioning scheme support

LTFS V2R1.1 already supports the TS1140 drive. Using LTFS V2R1.1 and TS1140 drives consumes 600 GB for the index partition.

LTFS V2R1.2 supports an enhanced partitioning scheme, which allows the user to use less space for indexes and more space for data partitions. By using the new scheme on TS1140 drives, LTFS V2R1.2 consumes only 300 GB for the index partition.

Update firmware: To use the new partitioning scheme, TS1140 firmware needs to be updated to PGA3 or later. If older firmware is installed, a cartridge with the new partitioning scheme cannot be mounted.

2.1.6 Filesystem in Userspace (FUSE)

LTFS is implemented on FUSE technology. The FUSE framework provides a file system infrastructure for UNIX technology-type systems and implements a set of standard file system interfaces to LTFS. The loadable kernel module and the userspace library offer an application programming interface (API) so that users can develop their own file systems, independently of the knowledge of kernel programming.

The system was initially implemented on Linux and Macintosh OS X. It is portable to any other UNIX technology-type systems that support FUSE. Figure 2-10 shows a simplified diagram of the Linux FUSE-based LTFS implementation.

When an LTFS volume is mounted, the directory information is read from the index partition by using XML (see 2.1.7, “XML schema” on page 112) and placed in the in-memory directory. See Figure 2-10. During I/O tape operations, the in-memory directory is updated to reflect the new data on tape. At various points, the contents of the in-memory directory are written back to the tape partitions.

Figure 2-10 LTFS and FUSE implementation architecture

FUSE also provides ease of maintenance to the kernel API and supports the use of userspace libraries and tools without having access to kernel data. (System root access is not required for the users.)

Within the Linux system, FUSE can be compiled and loaded similarly to any other daemons (for example, /bin/my_fs_daemon/mnt/fuse). The FUSE kernel module (Figure 2-10) is responsible for the routing requests that come from applications to the corresponding file system daemon. In addition, FUSE communicates with the file system daemon by using a special file descriptor that is obtained by opening the /dev/fuse directory.

Because tape is a sequential device, access times might differ depending on the location of the data to be accessed. Using the speed of LTO-5 with 140-MBps native does not cause a delay time that is longer than 50 seconds on average.

LTFS for Linux: With LTFS V2R1.2 for Linux, the user can share LTFS by use of the Network File System (NFS). LTFS for Linux is implemented with FUSE. FUSE provides functions for a user space program to be able to implement a fully functional file system.

File sharing by the use of NFS is provided by a new version of FUSE (Version 2.8). LTFS V2R1.2 for Linux relies on FUSE. The current version of FUSE has a limitation. See the readme.NFS file in the FUSE package and use the -o noforget option to support NFS.

2.1.7 XML schema

LTFS uses an XML schema to record information about the data files on the tape media. XML provides the following benefits, among others, to the LTFS format:

•Easy to document and to understand due to its human-readability

•Simple to port to different systems to produce interchangeable data

•Usable for standardization

•Easy data recovery in the event of a failure

When tape media is formatted for use by LTFS, an empty root directory structure is written to the index partition. For robustness and to provide compatibility with tape utilities that are not multiple partition aware, the same structure is also written to the data partition. In both cases, a single file mark is written before the schema. A double file mark is written after it, delimiting the index schema and indicating the end of data in the partition. See Figure 2-8 on page 110.

When an LTFS volume is mounted to write data on, the last file mark in the index partition is removed and is put back when the volume is unmounted. During mount time, if one file mark is found at the end of the index partition, the volume is not correctly unmounted. Then, LTFS uses the recovery procedures to find the latest recorded version of the XML schema to get the volume to a consistent status.

The XML schema contains a back pointer to the index that describes the previous generation (Gen-1) of the file system, which contains a back pointer to Gen-2, and so on. This pattern continues until the earliest generation of the index is reached. Although initially designed as a recovery mechanism, this feature makes LTFS a convenient versioned file system. Data files are never overwritten or erased from the tape until the tape is reformatted.

The following examples specify the XML schema. Attributes with square brackets are optional, and attributes that end in an ellipsis (...) are optionally repeatable. Attribute values in braces indicate that one of a series of values must be chosen, with the default value underlined.

LTFS label

The LTFS label is an XML data structure that describes information about the LTFS volume
and the LTFS partition on which the LTFS label is recorded. See Example 2-1.

Example 2-1 LTFS label structure

<?xml version="1.0" encoding="UTF-8"?>

<creator>IBM LTFS 1.2.0 - Linux - mkltfs</creator>

</location>

</partitions>

</ltfslabel

Index

An index is an XML data structure that describes all data files, directory information, and associated metadata for files that are recorded on the LTFS volume. See Example 2-2 and Example 2-3.

Example 2-2 LTFS index structure

<?xml version="1.0" encoding="UTF-8"?>

<creator>IBM LTFS 1.2.0 - Linux - ltfs</creator>

<comment>A sample LTFS Index</comment>

</location>

</previousgenerationlocation>

</indexpartitioncriteria>

</dataplacementpolicy>

...

</directory>

</ltfsindex>

Example 2-3 shows an index.

Example 2-3 LTFS index

<?xml version="1.0" encoding="UTF-8"?>

...

<name>LTFS Volume Name</name>

<name>directory1</name>

<readonly>false</readonly>

<name>subdir1</name>

<readonly>false</readonly>

</directory>

</contents>

</directory>

<file>

<name>testfile.txt</name>

</extendedattributes>

</extent>

</extentinfo>

</file>

</contents>

</directory>

</ltfsindex>

Extended attributes elements

All directory and file elements in an index can specify zero or more extended attributes (extendedattributes). These extended attributes are recorded as xattr elements in the extendedattributes element for the directory or file. See Example 2-4.

Example 2-4 Extendedattributes elements structure

...

<name>directory1</name>

<xattr>

<key>binary_xattr</key>

<value type="base64">/42n2QaEWDSX+g==</value>

</xattr>

<xattr>

<key>empty_xattr</key>

</xattr>

<xattr>

<key>document_name</key>

<value type="text">LTFS Format Specification</value>

</xattr>

</extendedattributes>

</contents>

</directory>

...

Data placement policy

An index can specify a data placement policy. This policy defines when the data extents for a file can be placed on the index partition. A data placement policy specifies the conditions under which it is allowed to place data extents on the index partition. See Example 2-5.

Example 2-5 Data placement policy structure

<?xml version="1.0" encoding="UTF-8"?>

...

</indexpartitioncriteria>

</dataplacementpolicy>

...

</directory>

</ltfsindex>

LTFS label XML schema

See Example 2-6 to get details about the definition. See also Example 2-7 on page 117.

Example 2-6 XML schema structure

<?xml version="1.0" encoding="UTF-8"?>

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">

<xs:element name="ltfslabel">

<xs:complexType>

<xs:all>

<xs:element name="creator" type="xs:string"/>

<xs:element name="formattime" type="datetime"/>

<xs:element name="volumeuuid" type="uuid"/>

<xs:element name="location">

<xs:complexType>

<xs:sequence>

<xs:element name="partition" type="partitionid"/>

</xs:sequence>

</xs:complexType>

</xs:element>

<xs:element name="partitions">

<xs:complexType>

<xs:all>

<xs:element name="index" type="partitionid"/>

<xs:element name="data" type="partitionid"/>

</xs:all>

</xs:complexType>

</xs:element>

<xs:element name="blocksize" type="blocksize"/>

<xs:element name="compression" type="xs:boolean"/>

</xs:all>

<xs:attribute name="version" use="required" type="version"/>

</xs:complexType>

</xs:element>

<xs:simpleType name="blocksize">

<xs:minInclusive value="4096"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="version">

<xs:restriction base="xs:string">

<xs:pattern value="[0-9]+.[0-9]+.[0-9]+"/>

<xs:enumeration value="2.0.0"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="datetime">

<xs:restriction base="xs:string">

<xs:pattern

value="[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:[0-9]

{2}.[0-9]{9}Z"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="partitionid">

<xs:restriction base="xs:string">

<xs:pattern value="[a-z]"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="uuid">

<xs:restriction base="xs:string">

<xs:pattern

value="[a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-

[a-fA-F0-9]{4}-[a-fA-F0-9]{12}"/>

</xs:restriction>

</xs:simpleType>

</xs:schema>

LTFS index XML schema

Example 2-7shows the LTFS index XML schema.

Example 2-7 LTFS index XML schema

<?xml version="1.0" encoding="UTF-8"?>

<xs:schema xmlns:xs="http://www.w3.org/2001/XMLSchema">

<xs:element name="ltfsindex">

<xs:complexType>

<xs:all>

<xs:element name="creator" type="xs:string"/>

<xs:element name="comment" type="xs:string" minOccurs="0"/>

<xs:element name="volumeuuid" type="uuid"/>

<xs:element name="generationnumber" type="xs:nonNegativeInteger"/>

<xs:element name="updatetime" type="datetime"/>

<xs:element name="location" type="tapeposition"/>

<xs:element name="previousgenerationlocation" type="tapeposition" minOccurs="0"/>

<xs:element name="allowpolicyupdate" type="xs:boolean"/>

<xs:element name="dataplacementpolicy" type="policy" minOccurs="0"/>

<xs:element name="highestfileuid" type="xs:nonNegativeInteger"/>

<xs:element ref="directory"/>

</xs:all>

<xs:attribute name="version" use="required" type="version"/>

</xs:complexType>

</xs:element>

<xs:element name="directory">

<xs:complexType>

<xs:all>

<xs:element name="fileuid" type="xs:nonNegativeInteger"/>

<xs:element name="name" type="xs:string"/>

<xs:element name="creationtime" type="datetime"/>

<xs:element name="changetime" type="datetime"/>

<xs:element name="modifytime" type="datetime"/>

<xs:element name="accesstime" type="datetime"/>

<xs:element name="backuptime" type="datetime"/>

<xs:element name="readonly" type="xs:boolean"/>

<xs:element ref="extendedattributes" minOccurs="0"/>

<xs:element name="contents">

<xs:complexType>

<xs:choice minOccurs="0" maxOccurs="unbounded">

<xs:element ref="directory"/>

<xs:element ref="file"/>

</xs:choice>

</xs:complexType>

</xs:element>

</xs:all>

</xs:complexType>

</xs:element>

<xs:element name="file">

<xs:complexType>

<xs:all>

<xs:element name="fileuid" type="xs:nonNegativeInteger"/>

<xs:element name="name" type="xs:string"/>
<xs:element name="length" type="xs:nonNegativeInteger"/>

<xs:element name="creationtime" type="datetime"/>

<xs:element name="changetime" type="datetime"/>

<xs:element name="modifytime" type="datetime"/>

<xs:element name="accesstime" type="datetime"/>

<xs:element name="backuptime" type="datetime"/>

<xs:element name="readonly" type="xs:boolean"/>

<xs:element ref="extendedattributes" minOccurs="0"/>

<xs:element name="extentinfo" minOccurs="0">

<xs:complexType>

<xs:sequence minOccurs="0" maxOccurs="unbounded">

<xs:element name="extent">

<xs:complexType>

<xs:all>

<xs:element name="partition" type="partitionid"/>

<xs:element name="startblock" type="xs:nonNegativeInteger"/>

<xs:element name="byteoffset" type="xs:nonNegativeInteger"/>

<xs:element name="bytecount" type="xs:positiveInteger"/>

<xs:element name="fileoffset" type="xs:nonNegativeInteger"/>

</xs:all>

</xs:complexType>

</xs:element>

</xs:sequence>

</xs:complexType>

</xs:element>

</xs:all>

</xs:complexType>

</xs:element>

<xs:element name="extendedattributes">

<xs:complexType>

<xs:sequence minOccurs="0" maxOccurs="unbounded">

<xs:element name="xattr">

<xs:complexType>

<xs:all>

<xs:element name="key" type="xs:string"/>

<xs:element name="value">

<xs:complexType mixed="true">

<xs:attribute name="type">

<xs:simpleType>

<xs:restriction base="xs:token">

<xs:enumeration value="base64"/>

<xs:enumeration value="text"/>

</xs:restriction>

</xs:simpleType >

</xs:attribute>

</xs:complexType>

</xs:element>

</xs:all>

</xs:complexType>

</xs:element>

</xs:sequence>

</xs:complexType>
</xs:element>

<xs:complexType name="policy">

<xs:sequence>

<xs:element name="indexpartitioncriteria">

<xs:complexType>

<xs:sequence>

<xs:element name="name" type="xs:string" minOccurs="0" maxOccurs="unbounded"/>

<xs:element name="size" type="xs:nonNegativeInteger"/>

<xs:element name="name" type="xs:string" minOccurs="0" maxOccurs="unbounded"/>

</xs:sequence>

</xs:complexType>

</xs:element>

</xs:sequence>

</xs:complexType>

<xs:complexType name="tapeposition">

<xs:all>

<xs:element name="partition" type="partitionid"/>

<xs:element name="startblock" type="xs:nonNegativeInteger"/>

</xs:all>

</xs:complexType>

<xs:simpleType name="version">

<xs:restriction base="xs:string">

<xs:pattern value="[0-9]+.[0-9]+.[0-9]+"/>

<xs:enumeration value="2.0.0"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="datetime">

<xs:restriction base="xs:string">

<xs:pattern

value="[0-9]{4}-[0-9]{2}-[0-9]{2}T[0-9]{2}:[0-9]{2}:

[0-9]{2}.[0-9]{9}Z"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="partitionid">

<xs:restriction base="xs:string">

<xs:pattern value="[a-z]"/>

</xs:restriction>

</xs:simpleType>

<xs:simpleType name="uuid">

<xs:restriction base="xs:string">

<xs:pattern

value="[a-fA-F0-9]{8}-[a-fA-F0-9]{4}-[a-fA-F0-9]{4}-

[a-fA-F0-9]{4}-[a-fA-F0-9]{12}"/>

</xs:restriction>

</xs:simpleType>

</xs:schema>

2.1.8 Recording the LTFS XML schemas

The following steps are taken by LTFS when files are written to the partitions. The following illustrations show how the XML schemas are recorded during the partition changes when a policy exists to have small files written to the index partition. They also show how multiple generations of indexes are stored on tape. In this scenario, the tape is already formatted as shown in Figure 2-8 on page 110.

The recording of the LTFS XML schemas follows these steps:

1. An LTFS formatted tape is mounted in read/write mode. Then, the following changes to the partitions are performed (see Figure 2-11):

– The last file mark in the index partition is erased.

This action indicates that a new session is started. Any file changes on the data partition are not recorded.

– New files F1 and F2 are added to the data partition.

The second file mark from the data partition is overwritten by file F1 to erase the end-of-data (EOD) indicator.

Figure 2-11 Files F1 and F2 added to the data partition

2. A small file that is called f3 is written to the index partition. Assume that LTFS is configured to cache a maximum of one file. The file is written as soon as it is closed by the application that writes it.

Before LTFS switches partitions, a new XML schema (which is aware of F1 and F2) is written to the end of the data partition (Figure 2-12). The new Index #1 contains a back pointer to the previous valid Index #0. It is enclosed by two file marks to indicate the boundaries of the XML schema and to ensure that its data is written to tape.

Figure 2-12 New index with a back pointer to the previous index

3. LTFS switches the partitions and writes the cached file f3 to the index partition (Figure 2-13). The file marks that are used to delimit the previous XML schema (Index #0 at the index partition) are overwritten by the f3 file.

Figure 2-13 Small file f3 that is written to the index partition

4. The LTFS volume becomes unmounted. Figure 2-14 shows the partition changes:

– The new XML schema that is called Index #2 is aware of the f3 file. Index #2 is written to the end of the index partition.

– Index #2 points back to the previous generation Index #1, which is stored in the data partition. A double file mark is appended to the index partition to indicate the EOD.

Figure 2-14 LTFS volume gets unmounted

5. The data partition XML schema is updated (Figure 2-15).

By adding the f3 file to the index partition, Index #1 at the end of the data partition is now out-of-date. Index #2 is now copied back to the data partition to ensure that both partitions provide a consistent view of the file system. Adding the f3 file also provides the ability to roll back to previous generations of the files on the tape.

Figure 2-15 Final partition view after tape unmount

6. The LTFS tape is unmounted and can be physically removed.

2.1.9 Reserved extended attribute definitions

In an LTFS index, all extended attribute names that start with the prefix “ltfs” with any capitalization are reserved for use by the LTFS format. (That is, any names that start with a not case-sensitive match for the letters “ltfs” are reserved.)

Any writer of an LTFS volume can use reserved extended attribute names to store extended attribute values only in conformance with the following tables.

The defined, reserved extended attributes are described in Table 2-2, Table 2-3, Table 2-4 on page 124, and Table 2-5 on page 127. Support for each of these defined, reserved extended attributes is optional for implementations that are in compliance with the LTFS format specification.

Software metadata

Table 2-2 lists the names of the defined, reserved extended attributes for the software metadata.

Table 2-2 Software metadata

Name	Accessibility	Availability	Description
ltfs.softwareVendor	Read only	Files and directories	This name is the software vendor name of this LTFS program. The LTFS program returns IBM.
ltfs.softwareProduct	Read only	Files and directories	This name is the product name of the LTFS program. The LTFS program returns LTFS SDE.
ltfs.softwareVersion	Read only	Files and directories	This name is the version number of this LTFS program.
ltfs.softwareFormatSpec	Read only	Files and directories	This name is the format specification version that is supported by this LTFS program.

Object metadata

Table 2-3 lists the names for the defined, reserved extended attributes of the object metadata.

Table 2-3 Object metadata

Name	Accessibility	Availability	Description	LTFS format version 1.0
ltfs.accessTime	read/write	Files and directories	This value is the date and time of the last access to the object.	N/A
ltfs.backupTime	read/write	Files and directories	This value is the date and time of the last archive or backup of the object.	N/A
ltfs.changeTime	read/write	Files and directories	This value is the date and time of the last status change to the object.	N/A
ltfs.createTime	read/write	Files and directories	This value is the date and time that the object was created.	ltfs.creationtime
ltfs.fileUID¹	Read-only	Files and directories	This value is the integer identifier for objects in the file system. This value is guaranteed to be unique within the LTFS volume.	N/A
ltfs.modifyTime	read/write	Files and directories	This value is the date and time of the last object modification.	N/A
ltfs.partition	Read-only	Non-empty file objects only	This value is the partition on which the first extent of the file is stored.	ltfs.partition
ltfs.startblock	Read-only	Non-empty file objects only	This value is the block address where the first extent of the file is stored.	ltfs.startblock

¹ Unique identifiers (UIDs) for each file/directory are permanently stored on a version 2.0.0 formatted medium. When a version 1.0 formatted medium is mounted, this attribute returns UIDs for each file/directory, but the values are not permanently stored. The UIDs are stored only while the medium is mounted. When the medium is remounted, different UIDs can be assigned.

Volume metadata

Table 2-4 lists the defined, reserved extended attributes for the volume metadata.

Table 2-4 Volume metadata

Name	Accessibility	Availability	Description	LTFS format version 1.0
ltfs.commitMessage	read/write	Root directory only	This message is meant to be saved in the tape index on the subsequent synchronization of LTFS to the medium. If the commit message is not changed afterward, all indexes that are written to the medium have that same message in the XML <comment> field.	ltfs.commit_message
ltfs.indexCreator	Read only	Root directory only	This string is the creator string for the index. This string provides a human-readable identifier for the product that generated the index.	N/A
ltfs.indexGeneration	Read only	Root directory only	This number is the last index generation number that is written to the tape.	ltfs.index_generation
ltfs.indexLocation	Read only	Root directory only	This value is the location of the last index on the medium in the form p:1, where p is an alphabetic character value that indicates the internal LTFS partition identifier, and 1 is the logical block number within the partition. For example, the value a:1000 indicates that the last index starts on partition “a” at logical block 1000.	N/A
ltfs.indexPrevious	Read only	Root directory only	Location of the previous index on the medium in the form p:1, where p is an alphabetic character value that indicates the internal LTFS partition identifier, and 1 is the logical block number within the partition. For example, the value b:55 indicates that the previous index starts on partition “b” at logical block 55.	N/A
ltfs.indexTime	Read only	Root directory only	This value is the date and time when the last LTFS index is written to the medium.	ltfs.index_time
ltfs.labelCreator	Read only	Root directory only	This value is the creator string for the LTFS label. This string provides a human-readable identifier for the product that generated the LTFS label.	N/A
ltfs.partitionMap	Read only	Root directory only	This value is the only medium partition layout for the LTFS volume. The form of the value is W:x,Y:z where W and Y have the value I, indicating an index partition, or D, indicating a data partition. The alphabetic character values x and y indicate the internal LTFS partition identifier. For example, the value I:a,D:b indicates that LTFS partition “a” is used as the index partition, and LTFS partition “b” is used as the data partition.	N/A
ltfs.policyAllowUpdate	Read only	Root directory only	This value indicates whether the data placement policy for the volume can be updated.	ltfs.policy_allowupdate
ltfs.policyExists	Read only	Root directory only	This value indicates whether the policy that allows certain files to be written to the index partition is enabled. The policy is set at the time of formatting.	ltfs.policy_is_set
ltfs.policyMaxFileSize	Read only	Root directory only	This value is the maximum file size for files that match the data placement policy for the volume.	ltfs.policy_maxfilesize
ltfs.sync	Write only	Root directory only	Writing any value to this extended attribute initiates a file system sync on any implementation that supports this extended attribute.	N/A
ltfs.volumeBlocksize	Read only	Root directory only	This block size is used on read/write operations as specified at the time of formatting.	ltfs.volume_blocksize
ltfs.volumeCompression	Read only	Root directory only	This value is the compression setting for the volume.	N/A
ltfs.volumeFormatTime	Read only	Root directory only	This value is the date and time when the volume was formatted.	ltfs.volume_format_time
ltfs.volumeName	Read	Files and directories	This name is the tape volume name as set at the time of formatting.	ltfs.volume_name
ltfs.volumeName	Write	Root directory only		ltfs.volume_name
ltfs.volumeSerial	Read only	Root directory only	This value is the tape serial number as specified at the time of formatting.	ltfs.tape_serial
ltfs.volumeUUID	Read only	Files and directories	This value is the Universally Unique Identifier that is assigned to the tape medium.	ltfs.volume_uuid

Media metadata

Table 2-5 lists the defined, reserved extended attributes for the media metadata.

Table 2-5 Media metadata

Name	Accessibility	Availability	Description
ltfs.mediaBeginningMediumPasses	Read only	Root directory only	This value is the total number of times the beginning of a medium position is passed. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaDatasetsRead	Read only	Root directory only	This value is the total number of data sets read from the medium over its lifetime. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaDataPartitionAvailableSpace	Read only	Root directory only	This value is the total available space in the data partition on the medium. The value is an integer count that is measured in units of 1048576 bytes.
ltfs.mediaDataPartitionTotalCapacity	Read only	Root directory only	This value is the total capacity of the data partition on the medium. The value is an integer count that is measured in units of 1048576 bytes.
ltfs.mediaDatasetsWritten	Read only	Root directory only	This value is the total number of data sets written to the medium over its lifetime. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaEfficiency	Read only	Root directory only	This value is the overall measure of the condition of the loaded medium. The value 0x00 indicates that the condition is unknown. The range of known values is from 0x01 (best condition) to 0xFF (worst condition). If the storage hardware cannot report this data, the value is -1.
ltfs.mediaIndexPartitionAvailableSpace	Read only	Root directory only	This value is the total available space in the index partition on the medium. The value is an integer count that is measured in units of 1048576 bytes.
ltfs.mediaIndexPartitionTotalCapacity	Read only	Root directory only	This value is the total capacity of the index partition on the medium. The value is an integer count that is measured in units of 1048576 bytes.
ltfs.mediaLoads	Read only	Root directory only	This value is the number of times that the medium is loaded in a drive. For example, with tape media, this value is the tread count. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaMBRead	Read only	Root directory only	This value is the total number of megabytes of logical object data that is read from the medium after compression over its lifetime. The value is rounded up to the next whole megabyte. The value that is reported includes bytes read as part of reading file marks from the medium. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaMBWritten	Read only	Root directory only	This value is the total number of megabytes of logical object data that is written to the medium after compression over its lifetime. The value is rounded up to the next whole megabyte. The value that is reported includes bytes written as part of writing file marks to the medium. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaMiddleMediumPasses	Read only	Root directory only	This value is the total number of times that the physical middle position of the user data region of a medium is passed. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaPermanentReadErrors	Read only	Root directory only	This value is the total number of unrecovered data read errors over the lifetime of the medium. This value is the total number of times that a read-type command terminated with a sense key of MEDIUM ERROR, HARDWARE ERROR, or equivalent over its lifetime. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaPermanentWriteErrors	Read only	Root directory only	This value is the total number of unrecovered data write errors over the lifetime of the medium. This value is the total number of times that a write-type command terminated with a sense key of MEDIUM ERROR, HARDWARE ERROR, or equivalent over the lifetime of the medium. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaPreviousPermanentReadErrors	Read only	Root directory only	This value is the total number of unrecovered read errors that occurred during the previous load of the medium. This value is the total number of times that a read-type command terminated with a sense key of MEDIUM ERROR, HARDWARE ERROR, or equivalent during the previous load session. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaPreviousPermanentWriteErrors	Read only	Root directory only	This value is the total number of unrecovered write errors that occurred during the previous load of the medium. This value is the total number of times that a write-type command terminated with a sense key of MEDIUM ERROR, HARDWARE ERROR, or equivalent during the previous load session. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaRecoveredReadErrors	Read only	Root directory only	This value is the total number of recovered read errors over the lifetime of the medium. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaRecoveredWriteErrors	Read only	Root directory only	This value is the total number of recovered data write correction errors over the lifetime of the medium. If the storage hardware cannot report this data, the value is -1.
ltfs.mediaStorageAlert	read/write	Root directory only	This value is a 64-bit value that contains alert flags for the storage system. For data tape media, this value is equal to the standard tape alert flags. If the storage hardware cannot report this data, the value is the string “UNKNOWN.” By writing a value of 1, the corresponding bit is cleared. By writing a value of 0, the current value is retained.

2.2 Installation

The system hardware and storage requirements are listed. The required steps to install the LTFS LE on a Linux system and also on a Windows system are described.

2.2.1 Hardware and software requirements

The system hardware and software requirements that relate to LTFS LE are listed. The current information for LTFS LE supported hardware and software configurations and notices or limitations is in the readme file of the software package.

Supported tape drives and media

The following tape drives and tape media are currently supported by the LTFS (Table 2-6).

Table 2-6 Supported tape drives and media

Tape drive	Tape media¹	Native/compressed capacity²	Firmware level³
IBM System Storage TS1140 Tape Drive⁴	Advanced data (JC)	4000/12000 GB (E07 Format)	69E (D3I3_69E) or later
	Extended data (JB)	1600/4800 GB (E07 Format)
	Economy data (JK)	500/1500 GB (E07 Format)
IBM System Storage LTO Ultrium 5 Tape Drive	LTO-5	1500/3000 GB	C7R2 or later (full-height tape drive) C7R3 or later (half-height tape drive)
IBM System Storage LTO Ultrium 6 Tape Drive	LTO-6	2500/6250 GB	C7S4 or later (full-height tape drive) C7S5 or later (half-height tape drive)

¹ WORM media is not supported.

² 1 GB = 1,000,000,000 bytes

³ This tape drive firmware level is required to fully use LTFS functions.

⁴ The TS1140 tape drive supports enhanced partitioning for cartridges. Enhanced partitioning allows more space for the data partition by using only 300 GB for the index partition on the IBM System Storage TS1140 Tape Drive. This partitioning scheme is applied when a cartridge is formatted for LTFS. To use enhanced partitioning, any TS1140 tape drives must be at firmware level D3I3_694 or later. If older firmware is installed on a tape drive, a cartridge with the enhanced partitioning scheme cannot be mounted. Beginning with version 2.1.2, the LTFS LE checks the firmware versions of all TS1140 tape drives in the library. If there is a TS1140 tape drive with older firmware, LTFS is unable to start. In this case, LTFS shows an error message that indicates that the drive firmware for all TS1140 tape drives in the library needs to be updated.

⁴

To learn more about fixes and updates for your system software, hardware, and operating system, see the Fix Central website:

http://www.ibm.com/support/fixcentral

System requirements

The system requirements are described for the LTFS. LTFS LE supports both Linux and Windows operating systems and hardware platforms. See Table 2-7 for Linux requirements. See Table 2-8 on page 132 for Windows requirements.

Linux system requirements

Table 2-7 shows the Linux system requirements.

Table 2-7 Linux system requirements

Linux computers
Supported operating systems (x86 and x86_64 only)	Red Hat Enterprise Linux Server 5.5, 5.6, 5.7, 5.8, 6.1, and 6.2
Supported operating systems (x86 and x86_64 only)	Novell SUSE Linux Enterprise Server 11 Service Pack (SP) 1 and SP2
Supported hardware	TS3500 Tape Library with TS1140 tape drives: • Library firmware level: latest • Drive firmware level: 69E (D3I3_69E) or later WORM media is not supported.
Supported hardware	TS2900 (1U), TS3100 (2U), TS3200 (4U), TS3310, and TS3500 tape libraries with LTO-5 and later generation LTO drives: •Library firmware level: latest •Drive firmware level: - LTO-5: C7R2 or later for full height; C7R3 or later for half height - LTO-6: C7S4 or later for full height; C7S5 or later for half height WORM media is not supported.
Server PC	x86 or x86_64 processor
	1 GB (0.93 GiB) of available RAM for each 1,000,000 files that are stored on a tape (The on-disk metadata storage (dcache) utilizes the loop device on the file system. When using dcache, the maximum number of files and directories depends on the file system that is used and the dcache image size, which, in turn, affects the amount of additional RAM that is required. The following file and directory limits exist for each file system: •ext3 file system: 130,000 files (and directories) per gibibyte (GiB) •ext4 file system: 65,000 files (and directories) per gibibyte (GiB) By default, LTFS reserves 5.37 GB (5 GiB) of disk space for the dcache image for each mount point. If the file system is ext4, LTFS can manage up to 325,000 files (65,000 x 5) under the mount point. Note: The ext3 file system is used on RHEL 5.5, 5.6, 5.7, and 5.8, and SUSE 11.1, by default. The ext4 file system is used on RHEL 6.1 and 6.2, by default.)
	5.91 GB (5.5 GiB) of available disk space (if on-disk metadata storage is enabled)
	Serial-attached Small Computer System Interface (SCSI) or Fibre Channel (FC) host bus adapter (HBA) is supported by IBM LTO-5 and later generation tape drives
	FC HBA is supported by IBM TS1140 tape drives
	TCP/IP based protocol network

Windows systems requirements

Table 2-8 shows the Windows system requirements.

Table 2-8 Windows system requirements

Windows computers
Supported operating systems	Windows Server 2008 R2 (SP1)
Supported hardware	TS3500 Tape Library with TS1140 tape drives: • Library firmware level: latest • Drive firmware level: 69E (D3I3_69E) or later WORM media is not supported.
Supported hardware	TS2900 (1U), TS3100 (2U), TS3200 (4U), TS3310, and TS3500 tape libraries with LTO-5 and later generation LTO drives: •Library firmware level: latest •Drive firmware level: - LTO-5: C7R2 or later for full height; C7R3 or later for half height - LTO-6: C7S4 or later for full height; C7S5 or later for half height WORM media is not supported.
Server PC	i386 or x86_64 processor
	1 GB (0.93 GiB) of available RAM for each 1,000,000 files that are stored on a tape
	Serial-attached Small Computer System Interface (SCSI) (SAS) or FC HBA supported by IBM LTO-5 and later generation tape drives
	FC HBA supported by IBM TS1140 tape drives

Important:

•LTFS LE generates a volume for every cartridge in the library.

•LTO media can be loaded only to LTO-5 and LTO-6 drives that are installed in the library.

•The TS2900 Tape Autoloader supports the Half-High tape drive.

•LTFS LE supports only the random mode for TS2900 Tape Autoloader, TS3100 Tape Library, and TS3200 Tape Library.

•In a TS3500 Tape Library, you can have a mixture of LTO-5 and LTO-6 drives in one logical library. The limiting factor in this case is that all tapes must be LTO-5 media. The TS1140 and the LTO-5 and LTO-6 cannot be in the same logical library.

•LTFS LE supports Control Path Failover (CPF).

Required software

The software requirements are described for the LTFS. The LTFS LE requires that the following software is installed:

•Linux:

– openssl 0.9.8e or later

– FUSE 2.7.4 or later, but not higher than 2.8.3 (kernel and shared library modules are required). To identify the FUSE version, use one of the following commands:

• ls -l /lib64 |grep -i fuse
lrwxrwxrwx 1 root root 16 Sep 24 17:12 libfuse.so.2 → libfuse.so.2.7.4 -rwxr-xr-x 1 root root 136352 Mar 2 2011 libfuse.so.2.7.4

• ls -l /lib | grep -i fuse

• ls -l /usr/local/lib | grep -i fuse

• ls - /usr/local/lib64 | grep -i fuse

– The following packages are included in the service pack:

• libxml2-2.6.26 or later

• libuuid 1.39 or later, or equivalent Universally Unique Identifier (UUID) functions that conform to OSF DCE 1.1 (including the e2fsprogs package)

• libicu 3.6 or later

•IBM tape device driver for Linux (lin_tape) 1.73 or later

•Python 2.4 or later, but earlier than 3.0

•Windows:

– Microsoft Visual C++ 2010 Redistributable Package

– Microsoft .NET Framework 4

– IBM tape device driver for Windows version 6225 or later

– Python 2.7.2 or later, but earlier than 3.0

Python: The ltfsadmintool script does not run on Python version 3.0 or later.

To learn more about fixes and updates for your system software, hardware, and operating system, see the Fix Central website:

http://www.ibm.com/support/fixcentral

2.2.2 Unsupported Windows features, drivers, and commands

The Windows features, drivers, and commands that are not supported by the LTFS are listed.

New Technology File System (NTFS)

LTFS does not support the following NTFS features:

• Compression of files within a folder (LTFS uses the compression of the tape drive.)

•Encrypted files and directories

•Quota management

•Reparse points

•Defragmentation

•Change journals to monitor file activity

•Scanning of all files or directories that a security identifier owns (FSCTL_FIND_FILES_BY_SID)

•Access control lists. The Security tab is not displayed when the user views the file property by right-clicking, then selecting Properties.

•Alternate data streams. The Details tab is not available when the user views the file property by right-clicking, then selecting Properties.

•Shadow copies. The Previous Version tab is not displayed when the user views the file property by right-clicking, then selecting Properties.

•Opportunistic locks

•Recycle bins

•Short names

•Audit and alarm events (as specified in access control lists)

•Windows sparse files application interfaces

•Transactional NTFS (TxF)

•Symbolic and hard links

Symbolic links: It is possible to access symbolic links that are created in another platform (Linux or OS X); however, it is not possible create a symbolic link with Windows.

File system filter drivers

Although Windows has a mechanism to support file system filter drivers, LTFS Windows does not support their use. File system filter drivers cannot attach to LTFS by using formally documented mechanisms or application interfaces.

Native commands for a disk drive

LTFS does not support the following commands:

•chkdsk

•format

•defrag

•recover

•label

Symbolic links

The following limitations exist when you use symbolic links on a Windows system:

•The user cannot create a new symbolic link between another file system and LTFS.

•The user cannot create a new symbolic link in LTFS.

•The user cannot delete a symbolic link in LTFS.

•The user cannot rename a symbolic link in LTFS.

•Accessing the symbolic link is always forwarded to the target of the link.

•The target of the link is shown as a file, even if it is a directory.

•The target of the link is not shown.

2.2.3 Installation on Windows

The LTFS program installation on a Windows system is described.

Installing Microsoft Visual C++ 2010 Redistributable package (required software)

Follow this procedure:

1. Go to the Microsoft Visual C++ 2010 Redistributable package download site for the corresponding system type (x86 or x64). There are a couple of Visual C++ packages that are available for download. Be careful to download the correct x64 package. The LTFS LE supports only x64.

2. Download the package to your local folder.

3. Double-click the package in Windows Explorer to display the User Account Control window. See Figure 2-16.

Figure 2-16 Windows Explorer

4. Click Yes when the message “Do you want to allow the following program to make changes to this computer?” displays.

5. Follow the program instructions until the installation is completed. See Figure 2-17.

Figure 2-17 Microsoft Visual C++ 2010 Redistributable package

Installing Microsoft .NET Framework 4 (required software)

Follow this procedure:

1. Go to the Microsoft .NET Framework 4 (Stand-alone Installer) download site.

2. Download the file to your local folder.

3. Double-click the file in Windows Explorer. See Figure 2-18.

Figure 2-18 Microsoft .NET Framework 4

4. Click Yes when the message “Do you want to allow the following program to make changes to this computer?” displays.

5. Follow the installer instructions until the installation is completed. See Figure 2-19 and Example 2-20 on page 161.

Figure 2-19 Microsoft .NET Framework 4 installer

Figure 2-20 Microsoft .NET Framework 4 installation complete

Installing the IBM tape driver for Windows (required software)

Follow these instructions before you begin the installation. Before you install the tape device driver, ensure that the following steps are completed:

1. The HBA and its device driver are installed automatically.

Emulex adapter: If the HBA that is attached to the tape library is an Emulex adapter, set the HBA configuration parameter "ExtTransferSize"=2 for Windows 2008 R2 by using the Emulex OneCommand Manager.

2. The tape drive and HBA are connected.

3. The tape drive is powered on.

4. The tape drives in the library are powered on.

5. The server is powered on.

Installation procedure

Complete the following steps to install the tape device driver:

1. Download the current version of the device driver from the IBM Fix Central portal:

a. Open the following URL in your web browser:

http://www.ibm.com/support/fixcentral

b. For Product Group, select System Storage.

c. For System Storage, select Tape systems.

d. For Tape systems, select Tape drivers and software.

e. For Tape drivers and software, select Tape device drivers.

f. For Platform, to download the latest version of atape, select the correct platform.

g. Click Continue.

h. Select the version that you want to download.

Installers: The IBM tape device driver has two installers: exclusive and non-exclusive. Use the exclusive installer with the LTFS LE.

i. Follow the instructions on the Fix Central download page to download the new version.

2. Download the PDF document, IBM Tape Device Drivers Installation and User’s Guide, GC27-2130-10, from the IBM Support portal:

a. Open the following URL in your web browser:

http://www.ibm.com/support/us/en/

b. Enter the guide name in the Search support text box and then click the blue button.

c. Click IBM Tape Device Drivers Installation and User’s Guide.

d. Click Device_Driver_IUG. PDF to download the user’s guide.

e. Follow the procedures in the guide to install the tape device driver.

Installing Python for Windows (required software)

Complete the following steps in order to install Python:

1. Go to the Python Windows download website.

2. Download the file to your local folder.

3. Double-click the file in Windows Explorer. See Figure 2-21.

Figure 2-21 Python

4. Follow the installer instructions until the installation is completed. See Figure 2-22 and Figure 2-23 on page 139.

Figure 2-22 Python installer

Figure 2-23 Python installation complete

Installing or upgrading LTFS LE Windows from an .exe file

Follow this procedure:

1. Complete the following steps to install or upgrade LTFS from an .exe file:

a. Open Windows Explorer and double-click the .exe file that you downloaded in advance from the IBM website. A dialog box displays with a drop-down menu from which you can select a language for the installation. See Figure 2-24, Figure 2-25, and Figure 2-26 on page 140.

Figure 2-24 Windows Explorer

Figure 2-25 LTFS LE installation wizard

Figure 2-26 LTFS LE installation wizard

b. Click OK to display a Software License Agreement window.

c. Select one of the following options to display the Destination Folder window:

• I accept the terms in the license agreement and click Next (Figure 2-27).

• Read Non-IBM Terms (to open a non-IBM license agreement) and click Close.

Figure 2-27 LTFS LE installation wizard

Destination folder: With the Destination Folder wizard, you can change the folder to install the program. The default destination folder is set to C:Program FilesIBMLTFS. If you want to change the default folder, click Change. Otherwise, click Next to skip to step g.

d. If you click Change, the Change Current Destination Folder window displays the following options. Either choose a folder from the Look in menu (Figure 2-28) or set the folder name from the Folder name text box.

Figure 2-28 LTFS LE installation wizard

e. After you select an option, click OK to proceed to the next step.

f. Click Next in the Destination Folder window to display the Ready to Install the Program window (Figure 2-29).

Figure 2-29 LTFS LE installation wizard

Changes: If you want to change an installation parameter, click Back, and then follow the same steps that got you to step g.

g. Click Install to display the “Installing IBM Linear Tape File System Library Edition” window.

Progress: You can track the progress of the installation by viewing the status bar.

h. Click Yes when the User Account Control window displays the message “Do you want to allow the following program to make changes to this computer?” When the installation completes, the InstallShield Wizard Completed window displays (Figure 2-30).

Figure 2-30 User account control window

Administrator password: If you log on to the system as a user account, the User Account Control window displays. To proceed to the next step, type an administrator password and then click Yes.

i. Click Finish and then Yes when the “Installing IBM Linear Tape File System Library Edition Installer Information” window displays to restart the system to launch LTFS. See Figure 2-31 and Figure 2-32.

Figure 2-31 LTFS LE installation wizard

Figure 2-32 LTFS LE installation wizard

Figure 2-33 shows how the physical tapes are mapped to the Windows file system. Each physical tape cartridge volume serial numbers (VOLSER) is displayed as a directory on Windows Explorer.

Figure 2-33 Tape library tapes mapped to Windows

Uninstalling LTFS from a Windows system

Complete the following steps to uninstall LTFS:

1. From the desktop, click Start → All Programs → IBM → LTFS → Uninstall LTFS to open the Windows Installer program.

2. Click Yes when the message “Are you sure you want to uninstall this product?” displays.

3. Click Yes when the message “Do you want to allow the following program to make changes to this computer?” displays.

Administrator: If you log on to the system as a user account, the User Account Control window displays. You must type an administrator password to proceed to the next step and then click Yes.

4. If the message “All LTFS tape libraries must be unmounted before uninstalling LTFS program” displays, unmount all tape libraries from LTFS by using the ltfsunmount command, and then try again. If this message does not display, proceed to step 5.

5. Click OK when either of the following messages displays:

– “The following applications should be closed before continuing the uninstall.”

– “The setup was unable to automatically close all required applications.”

Progress: Use the status bar to track the progress of the uninstall.

6. Restart the system to completely remove the LTFS program from your operating system.

Alternative: You also can uninstall the LTFS program from the desktop by clicking Start → Control Panel → Uninstall a program under Programs. Then, restart your operating system.

2.2.4 Installation on Linux

Complete these required steps before you install LTFS LE.

Installing the prerequisite Linux software

LTFS LE requires the following prerequisite packages to be installed:

•openssl 0.9.8e or later

•FUSE 2.7.4 or later (including kernel and shared library modules)

•libxml2-2.6.26 or later

•libuuid 1.39 or later, or equivalent UUID functions that conform to OSF DCE 1.1 (included in the e2fsprogs package)

•libicu 3.6 or later

•Python 2.4 or later, but earlier than 3.0

Tip: The ltfsadmintool script does not run on Python version 3.0 or later.

Installing the HBA and HBA device driver

To install the HBA and its device driver for use with LTFS LE, see the documents that are provided by the HBA manufacturer. For a list of supported adapters, go to the following LTO page:

http://www.ibm.com/systems/storage/tape/lto/

For information about connectivity and supported configurations, see the IBM System Storage Interoperation Center (SSIC) website:

http://www.ibm.com/systems/support/storage/config/ssic/

Emulex adapter: If the HBA that is attached to the tape library is an Emulex adapter, add one of the following lines to the /etc/modprobe.conf file (for RHEL 5.5, 5.6, 5.7, or 5.8) or to the /etc/modprobe.d/lpfc.conf file (for RHEL 6.1, 6.2, and SLES 11). Then, restart the system for the changes to take effect.

For RHEL 5.5, 5.6, 5.7, or 5.8, add this line:

modprobe lpfc lpfc_sg_seg_cnt=256

For RHEL 6.1, 6.2, and SLES 11, add this line:

options lpfc lpfc_sg_seg_cnt=256

Installing the IBM tape device driver

Before you install the tape device driver for use with LTFS LE, complete the following tasks:

1. Install the HBA and its device driver:

a. Download the current version of the tape device driver from the IBM Fix Central portal (see 2.1.4, “LTFS tape format specification” on page 108):

http://www.ibm.com/support/fixcentral

b. Download the IBM Tape Device Drivers Installation and User’s Guide (a PDF document) from the IBM Support and Download site:

http://www.ibm.com/support/us/en/

c. Follow the procedures in the user’s guide to install the tape device driver or see 1.2.4, “Installation procedure” on page 16.

2. Connect the tape drive and HBA.

3. Power on the tape drive in the tape library.

4. Power on the server.

Enabling the system log

Error information for LTFS LE operations is displayed on the terminal console. The level of error reporting is based on the log trace level.

When the LTFS commands are used, error and trace information is recorded by using the system log mechanism. This mechanism is disabled by default. To enable the LTFS system log and locate additional information about LTFS logs, follow these steps:

1. Log on to the operating system as root.

2. Edit the system log configuration file on any available editor to enable the system log. Apply one of the following edits, depending on your base system. See Example 2-8.

Example 2-8 System log configuration file

For RHEL 5.5, 5.6, 5.7, or 5.8 add the following line to /etc/syslog.conf

user.* /var/log/userlog

For RHEL 6.1, 6.2, or SLES 11 add the following line to /etc/rsyslog.conf

:msg, contains, "LTFS" /var/log/ltfs.log

3. Enable the setting by restarting syslog (for RHEL 5.5, 5.6, 5.7, or 5.8) or ryslog (for RHEL 6.1, 6.2, or SLES 11). Issue the following command:

/etc/init.d/syslog restart

4. Open the /etc/logrotate.d/syslog file on any available editor (see Example 2-9). Add or edit the following lines.

Example 2-9 Add or edit syslog file

For RHEL 5.5, 5.6, 5.7, or 5.8:

/var/log/userlog {

size 1M

rotate 4

missingok

compress

sharedscripts

postrotate

/bin/kill -HUP `cat /var/run/syslogd.pid 2> /dev/null` 2> /dev/null || true

/bin/kill -HUP `cat /var/run/rsyslogd.pid 2> /dev/null` 2> /dev/null || true

endscript

}

For RHEL 6.1, 6.2, or SLES:

/var/log/ltfs.log {

size 1M

rotate 4

missingok

compress

sharedscripts

postrotate

/bin/kill -HUP `cat /var/run/syslogd.pid 2> /dev/null` 2> /dev/null || true

/bin/kill -HUP `cat /var/run/rsyslogd.pid 2> /dev/null` 2> /dev/null || true

endscript

}

LTFS records and displays the following types of logs:

Error A message that indicates an unrecoverable error occurred or an operation unexpectedly failed

Warning A message that indicates an unexpected condition occurred, but the operation can continue

Information A message that provides additional information about the current operation

Trace A message that provides information about debugging and troubleshooting

Table 2-9 describes the logging command options, types of results, and reasons to use the command options.

Table 2-9 Logging command options and results

Command option	Command	Results	Use
No option designated	Any	Error, warning, and information messages are logged and reported. This option is the default setting.	LTFS is designed to use this default setting to isolate any problem that occurs and to identify the appropriate action.
-o quiet	ltfs	Error and warning messages are logged and reported.	The quiet option can be used to improve performance by reducing the level and number of messages logged and reported.
-q or --quiet	ltfsadmintool -f -- ltfsadmintool -c --	Error and warning messages are logged and reported.
-o trace	ltfs	Error, warning, information, and trace messages are logged and reported.	The trace option can be used if intermittent errors occur, or the default log information is insufficient to reproduce the problem in the IBM lab. The trace option is set under these special conditions to gather additional information to help re-create and solve the problem.
-t or --trace	ltfsadmintool -f --
-a or --trace	ltfsadmintool -c --

Logging command options: The ltfsadmintool does not support logging command options, but when the -f option or -c option is specified, the logging command options for mkltfs and ltfsck functions can be specified after the -- tag. See “Optional mkltfs parameters” on page 191 and “Optional ltfsck parameters” on page 194.

For more information about how to set the log trace level, see the LTFS LE information center and search for “setting log trace level”:

http://publib.boulder.ibm.com/infocenter/ltfsle/cust/index.jsp

Latest information: Information in the readme and install files that are provided with the LTFS LE distribution package supersedes information that is presented and the information in the information center.

Extracting binary rpm files from an installation package

Extracting binary rpm files from an installation package for LTFS on a Linux system is described.

Installation package: The installation package name is ltfs-2.1.1-[revision].bin. It includes rpm files for the revision and supported platforms and architectures.

Ensure that the following steps are completed before you begin to create a binary rpm package:

1. Log on to the operating system as root.

2. Confirm the execution permission of ltfs-2.1.1-[revision].bin (Example 2-10) by issuing the following command:

# ls -l ltfs-2.1.1-[revision].bin

Example 2-10 LTFS confirm execution permission

[root@double ~]# ls -l ltfs-2.1.1-201109201.i386.bin

-rwxr-x--- 1 root root 68924904 Sep 29 16:23 ltfs-2.1.1-201109201.i386.bin

3. Set the execution permission, if it is not already set, by issuing the following command:

# chmod +x ltfs-2.1.1-[revision].bin

Create the rpm packages: If the software shipped as an rpm source file named “src.rpm”, issue the following command to create the rpm packages:

# rpmbuild --rebuild ltfs-2.1.1-[revision].src.rpm

After you verify the execution permission, complete the following steps to extract the binary rpm files:

1. Run the binary file on the system by following the steps that are shown in step 2 of the preceding list. You can launch it on a terminal by issuing the following command in the directory in which the file resides:

# ./ltfs-2.1.1-[revision].bin

The extracted rpm files are in the directory: /root/LTFSLE/rpm[revision]/[platform]/

2. Select the rpm files that correspond to your platform.

The following files in the rpm folder are extracted:

•ltfs-2.1.1-[revision]-[platform].rpm

•ltfs-license-2.1.0-[revision]-[platform].rpm

•ltfs-library-2.1.1-[revision]-[platform].rpm

•ltfs-admin-utils-2.1.1-[revision].noarch.rpm

The ltfs-admin-utils-2.1.1-[revision].noarch.rpm file is common to all platforms. Both ltfs and ltfs-library rpm file names include platform tags.

For example, if you want to install LTFS rpm files for revision 201109201, by using a Red Hat Linux distribution on a 64-bit architecture, the following files in the rpm folder are required:

•ltfs-2.1.1-201109021-RHEL61.i386.rpm

•ltfs-license-2.1.0-201109021-RHEL61.i386.rpm

•ltfs-library-2.1.1-2011009201-RHEL561.i386.rpm

•ltfs-admin-utils-2.1.1-201109201.noarch.rpm

The procedure to use the files is described.

Installing or upgrading from a binary rpm file

Follow these steps to install or upgrade the IBM LTFS LE program on a Linux system from a binary rpm file.

ltfsadmintool utility: LTFS LE does not support remote access by using the ltfsadmintool utility. Port 2112 must be blocked by a firewall program. If this port is set as a trusted port by using a firewall program, remove port 2112 as the trusted port before you install LTFS.

If port 2112 cannot be blocked because it is used by another program, specify another blocked port number with the -o admin_port option and the ltfs command.

Installing the binary rpm files: Install the binary rpm files by following the procedure in “Select the rpm files that correspond to your platform.” on page 149.

Complete the following steps from a Linux command prompt:

1. Log on to the operating system as a root user.

2. If LTFS is already installed and you want to upgrade it, unmount the LTFS file system by entering the following command:

umount /mnt/ltfs

To continue the upgrade, skip to step 5.

3. Run these commands in the following sequence to install the rpm packages:

# rpm -ihv ./ltfs-license-2.1.0-[platform].[architecture].rpm

# rpm -ihv ./ltfs-2.1.2-[revision]-[platform].[architecture].rpm

# rpm -ihv ./ltfs-library-2.1.2-[revision]-[platform].[architecture].rpm

# rpm -ihv./ltfs-admin-utils-2.1.2-[revision].noarch.rpm

Required: It is required to install the ltfs-license rpm file before you install the other packages. It is required to install the ltfs rpm file before you install the ltfs-library file. The commands are the same, whether you install for the first time or upgrade the rpm file. For example, if you install the ltfs-library rpm, revision 201111110, by using a Red Hat Linux distribution 6.1 on a 64-bit architecture, you issue the following command:

# rpm -ihv ./ltfs-library-2.1.2-201111110-RHEL61.x86_64.rpm

4. Optional: Run the following commands to verify the current version of LTFS LE:

# rpm -q ltfs-license

# rpm -q ltfs

# rpm -q ltfs-library

# rpm -q ltfs-admin-utils

5. Run these commands in the following sequence to upgrade the binary rpm package:

# rpm -Uhv ./ltfs-2.1.2-[revision]-[platform].[architecture].rpm

# rpm -Uhv ./ltfs-library-2.1.2-[revision]-[platform].[architecture].rpm

# rpm -Uhv ./ltfs-admin-utils-2.1.2-[revision].noarch.rpm

6. Optional: Run the following commands to verify the current version of LTFS LE:

# rpm -q ltfs-license

# rpm -q ltfs

# rpm -q ltfs-library

# rpm -q ltfs-admin-utils

Uninstalling from a Linux system

Follow these steps to uninstall LTFS from a Linux system:

1. Log on to the operating system as a root user.

2. If LTFS is mounted, unmount it by entering the following command:

# umount /mnt/ltfs

3. Run these commands in the following sequence to remove the installed packages:

# rpm -e ltfs-admin-utils-2.1.2-[revision]

# rpm -e ltfs-library-2.1.2-[revision]

# rpm -e ltfs-2.1.2-[revision]

4. Issue the following commands to check that the packages are uninstalled:

# rpm -q ltfs

# rpm -q ltfs-library

# rpm -q ltfs-admin-utils

Trials: If you uninstall to upgrade to a higher version, keep the license module installed. You cannot install an upgrade package without the license module. If you have a license module for a trial version and the license expires, LTFS works in read-only mode. To use all LTFS functions, uninstall the trial license module and install the license module for the product version.

Optional: The extracted rpm files for those operating systems other than your system can be deleted. Retain the rpm files that belong to LTFS LE for your operating system in case you need to reinstall.

2.3 Managing LTFS LE

Information is provided about the interfaces that are supported for use with LTFS. The operations to mount and unmount media in a tape library for LTFS LE are described.

2.3.1 Typical user scenario on a Windows system

Performing a series of typical administrative tasks in LTFS LE is described.

The following commands (Table 2-10) are used for common and basic tasks, such as mounting and unmounting a library, displaying available libraries, and managing LTFS.

Table 2-10 Basic LTFS LE administrative tasks

Command	Purpose
ltfs.exe	This command starts LTFS to mount a tape library. A drive letter must be specified.
ltfsunmount.exe	This command unmounts LTFS.
ltfsadmintool	This command is used to manage LTFS. It is used for operations, such as formatting a medium, checking a medium, and importing or exporting a medium. Python must be installed to use this tool.

Starting LTFS and mounting a library on a Windows system

Use these procedures to start LTFS and mount a library on a Windows system.

Follow this procedure:

1. Issue the ltfslsdev.exe command to view the available tape libraries as shown in the following example:

ltfslsdev.exe

2. Review the results that display the mapping between the address and serial number of the tape library as shown in the following example. The address is used to mount the library.

C:Program FilesIBMLTFS>ltfslsdev.exe

Assigned Address Serial
---------- ------------ ------------------------

2.0.14.1 00L4U78C6339_LL1

Issue the ltfs.exe command with the address and a drive letter as shown in the following example to start LTFS:

ltfs.exe [options] DriveLetter

Drive letter: The ltfs.exe command on a Windows system requires a drive letter. There is no rule for the order of the options that are used with the ltfs.exe command. However, the drive letter must be after any options. The last argument for the ltfs.exe command must be the drive letter.

The %TEMP%ltfs directory is set as the default work_directory on a Windows system.

The following LTFS options are available:

-o changer_devname=<dev>
Changer device (For example, 1.2.3.4)

-o work_directory=<dir>
LTFS work directory

-o sync_type=<type> Specify sync type (default: time@5). <type> must be specified with one of the following values:

- time@min: LTFS attempts to write an index each 'min' minutes. min must be a decimal number 1 - 35791394. It is equivalent to "-o sync_type=unmount" when 0 is specified (default: min=5).

- close: LTFS attempts to write an index when a file is closed.

- unmount: LTFS attempts to write an index when the medium is unmounted.

-o device_list Show available tape devices

-o release_device Clear device reservation (must be specified with -o changer_devname)

-o run_inventory Run a full inventory to create the LTFS namespace

-o library_cache Set to use and create caches from the medium auxiliary memory (MAM), index, and labels (default)

-o nolibrary_cache Set to disable the use and creation of MAM, index, and label caches

-o volser_only_folder
Set to exclude volume name from top-level folder name

-o novolser_only_folder
Set to include volume name in top-level folder name (default)

-o six_char_folder Display first six characters of directory name

-o strict_drive Mount only the corresponding generation of the medium with the tape drive. For example, an LTO-6 drive mounts an LTO-6 medium only.

-o nostrict_drive Mount any supported generation of the medium for the drive (default). For example, an LTO-6 drive mounts either an LTO-5 medium or an LTO-6 medium.

-o scsi_lbprotect=<on|off>
Enable logical block protection (default=off)

-o symlink_type=<type>
Specify symbolic link type (default: posix). <type> must be specified with one of the following values:

- posix: LTFS behavior is the same as standard symbolic link

- live: LTFS replaces mount point path by current mount point

-o quiet Disable informational messages

-o trace Enable diagnostic output

-o fulltrace Enable full call tracing

-a Advanced help

-h Display this help and exit

The following LTFS Admin Interface option is available:

-o admin_port=<num> TCP/IP port on which to listen for connections (default: 2112)

Example 2-11 shows the command and the results.

Example 2-11 ltfs -o changer_devname command

C:Program FilesIBMLTFS>ltfs -o changer_devname=2.0.14.1 L

LTFS61259I Sync type is "time", Sync time is 300 sec

LTFS61509I Set trace level in library (LTFS_INFO)

LTFS61206I LTFS starting. Drive:L, (Sep 20 2012 08:47:36) log level 2

LTFS17085I Plugin: Loading "scsilib" driver

LTFS17085I Plugin: Loading "unified" iosched

LTFS17085I Plugin: Loading "scsilib" changer

LTFS11593I LTFS starts with a product license version (20120712_1200)

LTFS62611I Opening a device through scsilib driver (af5fb0h)

LTFS62613I Changer serial is 00L4U78C6339_LL1

LTFS62160I Drive serial is 1068000264

LTFS17160I Maximum device block size is 1048576

LTFS62160I Drive serial is 1068000256

LTFS17160I Maximum device block size is 1048576

LTFS61509I Set trace level in library (LTFS_INFO)

LTFS11545I Rebuilding the cartridge inventory

LTFS11571I State of tape '001106L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape 'DG I69L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '3TX070L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '3DA491L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '3DA493L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '059AHDL4' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '3SS081L4' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape 'DG I73L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS11571I State of tape '3TX071L3' in slot 0 is changed from 'Not Initialized' to 'Non-supported'

LTFS14506I LTFS admin server is listening on port 2112

LTFS61280I Symbolic link type is (posix)

LTFS61237I Read-only directory mode is Windows

Important: To enable any of these options, specify them with the mount command.

Example 2-12 shows how to set the option sync_type to “close”.

Example 2-12 Setting sync_type option to close

C:Program FilesIBMLTFS>ltfs -o changer_devname=2.0.14.1 -o sync_type=close L

LTFS61260I Sync type is "close"

LTFS61509I Set trace level in library (LTFS_INFO)

LTFS61206I LTFS starting. Drive:L, (Sep 20 2012 08:47:36) log level 2

LTFS17085I Plugin: Loading "scsilib" driver

LTFS17085I Plugin: Loading "unified" iosched

LTFS17085I Plugin: Loading "scsilib" changer

LTFS11593I LTFS starts with a product license version (20120712_1200)

LTFS62611I Opening a device through scsilib driver (a091f0h)