HSM StrongLink (tape archive)

file version: 25 August 2022

Introduction

The DKRZ operates a hierarchal storage management system (HSM) used for the storage of all relevant data created and post processed on DKRZ systems. The hardware of the HSM consists of a disk cache and two tape libraries. The primary tape archive has a storage capacity of more than 300 PB and is located in the DKRZ building in Hamburg. Selected files are mirrored to the secondary tape archive located at the Max Planck Computing and Data Facility (MPCDF) in Garching. The software installed to operate the HSM is StrongLink. All command-line based user-interaction with the tape archive goes through StrongLink and its command line tool slk.

If you have questions, which are not answered on this page or on the linked pages, please have a look into our FAQ. If you do not find an answer there, please contact us via support@dkrz.de .

Storage options and quota

The amount of data that can be stored in the tape archive per project is limited by the available storage quota of that project. Individual users do not have a quota. Storage space on the HSM is applied for in conjunction with the (bi-)annual application for DKRZ compute and storage resources. You can check your storage quota via https://luv.dkrz.de . There is normal tape archive quota denoted as arch and quota for long term archival denoted as docu. Additionally, users might select very important files to be stored twice, i.e. one copy in Hamburg and one copy in Garching. The following table provides an overview:

Storage location, time and quota

File Storage	Storage Time	Used quota	How to achieve this	Past location (HPSS)	New location (StrongLink)
single copy on tape	1 year after expiration of DKRZ project	arch quota	default storage type	/hpss/arch/<prj>	/arch/<prj>
second copy on separate tape	1 year after expiration of project	arch quota; double file size used	store data in specific root namespace (see right column)	/hpss/double/<prj>	/double/<prj>
long-term storage for reference purpose	10 years after expiration of project	docu quota	Please contact data@dkrz.de	/hpss/doku/<prj>	/doku/<prj>

The tape archive delivers its best performance if the files to be archived are sufficiently large. The recommendations on packing developed on the basis of HPSS therefore remain effective for the time being. Like with the HPSS system, the accounting of used quota is done in increments of 1GB per archived file.

slk: command line tool for tape access

Note

slk stores a login token in the home directory of each user (~/.slk/config.json). The login token is valid for 30 days. By default, this file can only accessed by the respective user (permissions: -rw-------/600). However, users should be careful when doing things like chmod 755 * in their home directory. If you assume that your slk login token has been compromised please contact support@dkrz.de .

The StrongLink software comes with a command line tool suite slk. slk is the user interface to the StrongLink software and allows the user to interact with the HSM. The available commands are:

• help: displays the slk help page

• version: print the slk version

• login: log in to the system with LDAP credentials

• archive: copy files to the HSM

• chmod: modify permissions of archived files (same as chmod on the Linux shell)

• delete: delete a namespace (and all child objects for the namespace) or a specific file

• group: change group ownership of archived files; for file owners and admins only

• owner: change ownership of archived files; for admins only

• tag: modify metadata of archived files

• search: search archived files based on metadata; deactivated, please see slk_helpers search_limited instead

• list: list searched files and some of their metadata (similar to ls on the Linux shell)

• retrieve: retrieve files based on search result or based on absolute path; retrieves not more than 500 files; limited availability see Retrievals from tape

• recall: recall files based on search result or based on absolute path (recall: copy from tape to HSM-cache)

• move: move a file or a namespace from one namespace to another namespace (might be merged with slk rename in future)

• rename: rename a file or a namespace (might be merged with slk move in future)

Note

StrongLink uses the term “namespace” or “global namespace” (gns). A “namespace” is comparable to a “directory” or “path” on a common file system.

Note

slk does not provide its own shell. slk login simply creates a login token which allows other slk commands to be used. It is not possible to navigate through the emulated directory structure of the HSM using a cd-like command. Instead, each slk commands needs the full path of resources as input.

Note

Currently (May 2022), slk retrieve is only available on nodes of the compute, shared and interactive partitions on Levante. All other slk commands are available on all nodes. If you wish to use slk retrieve interactively, please start an interactive batch session via the interactive partition with salloc (Data Processing on Levante).

Please read Known Issues before you start using the slk the first time. Please have a look into StrongLink Command Line Interface (slk) (on doc.dkrz.de) or into the StrongLink Command Line Interface Guide for a detailed description of the individual commands. Alternatively, the sections Switching: pftp to slk and slk Usage Examples contain several usage examples.

slk helpers: add-on to slk

The slk is lacking a few minor but very useful features. The slk_helpers program adds these features. Its usage is very similar to the usage of slk. The following commands are available:

• checksum: prints one or all checksums of a resource

• exists: check if a resource exists; the resource id is returned

• help: print help / usage information

• gen_file_query: generate a query to find files

• hostname: prints the hostname to which slk is currently connect to or to which slk will connect to

• iscached: checks if a file is in the HSM cache

• list_search: Get paths of all resources found by search id (similar slk list but prints full path of search results)

• mkdir: create a namespace in an already existing namespace (like mkdir on the Linux shell)

• metadata: get metadata of a resource

• resourcepath: get path for a resource id

• search_limited: submits and search and returns search ID if 1000 or less results are found

• session: prints until when the current slk session is valid

• size: returns file size in byte

• version: prints the version

Please have a look into slk helpers (on doc.dkrz.de) for a detailed description of the individual commands. Alternatively, the section slk Usage Examples contains several usage examples.

pyslk: python slk wrapper

We offer a python Wrapper package for the slk and the slk_helpers. Most commands of these two command line interfaces have corresponding wrappers in pyslk. Usage examples of pyslk are shown in the section pyslk in Getting Started with slk. Additionally, a pyslk API reference is provided. Feel free to download the package from pyslk availability.

External access via sftp

Currently, the HSM can only be accessed via Levante

Metadata: harvesting, manual manipulation and search

The StrongLink software reads and extracts extended file metadata from the headers of archived netCDF files. Users may edit some of these metadata and add further metadata via slk tag. The metadata, which are stored, are described in detail on our metadata manual page (Reference: metadata schemata). Additionally, extended metadata for some common file formats such as jpeg are extracted. Harvesting extended metadata from additional file formats used in the context of Earth System modeling is planned to be implemented.

Files can be searched and found based on their metadata. The StrongLink software provides the command line tools slk search, slk list and slk retrieve to search, list and retrieve files based on their metadata, respectively. Retrieval of files based on their absolute path in the HSM is also possible. Please see the sections Reference: metadata schemata, slk Usage Examples and StrongLink Command Line Interface (slk) for details on the usage of slk in this context.

Packing of data

The tape archive delivers its best performance if the files to be archived are sufficiently large. The recommendations on packing developed based on of HPSS therefore remain effective for the time being. Like with the HPSS system, the accounting of used quota is done in increments of 1GB per archived file.

The package packems, which was developed by MPI-M and DKRZ for the HPSS, has been partly adapted to the new HSM system. The process of packing & archiving of multiple data files to tape and their retrieval is simplified by this package. It consists of three command line programs:

• a pack-&-archive tools packems,

• a list-archived-content tool listems,

• a retrieve-&-unpack tool unpackems and

Please use module load packems to load the packems package (will be available mid-June 2022). For details on the usage of packems please have a look into the packems manual.

Backend data handling

Just like with the previous HSM system HPSS, the fast disk cache is installed upstream of the tape system. Files selected for archival are first copied to the disk cache and then successively written onto tape. Files selected for retrieval are first copied from tape to the cache and then copied to the specified target locations. The retrieval of files that are still/already stored in the disk cache is considerably faster than the retrieval of files that are located on tape only.

The distribution of the files in the disk cache, primary tape archive and secondary tape archive is automatically controlled by the software StrongLink.

A - or t is appended to the permissions string of each file in the output of slk list and slk_helpers search_list. The - indicates that the file is stored in the cache. The t indicates the file is stored on tape. Alternatively, the command slk_helpers iscached can be used to check whether a file is currently stored in the cache.

Comparison: HPSS and StrongLink HSM

The table below provides a comparison of a few key features of the HPSS and the new HSM (column “StrongLink”). Although “StrongLink” only is the name of the HSM software and is not related to the hardware, we denote the whole system as “StrongLink”.

performance information new HSM system

	HPSS	StrongLink
Disk cache	2.5 PB (configured; 5 PB installed)	1.2 PB
Throughput rate	15 GB/s (average)	30 GB/s (average; 15 GB/s upload; 15 GB/s download)
Annual throughput	75 PB	120 PB
Server setup and reliability	One HPSS server + one backup server on standby	11 equal StrongLink Nodes (node = server); HSM becomes read-only when two nodes fail
Expand HSM performance	replace HPSS server	add additional nodes as needed
Extended metadata	no	yes
Search files by metadata	no	yes

It is important to note that the StrongLink system consists of n equal nodes. n is 11 at the moment but can be extended in future when more bandwidth for data archival or retrieval is needed. The system as a whole still works without any restrictions if one node fails. If a second node fails, the system will go into read-only mode to prevent inconsistencies in the metadata database. Hence, the system is considerably more failsafe then the current system and easier to extend for future data workflows.

Further reading

• Getting Started with slk
  • Introduction
  • Load slk module
  • slk memory footprint
  • slk data transfer rate
  • slk in SLURM jobs
  • Login
  • Archival
  • List content
  • Validate archivals
  • Retrieve a file or namespace
  • Move, rename and delete files or namespaces
  • Search files
  • Debugging
  • packems
  • pyslk
• Known Issues (read this!)
  • slk issues on Levante
  • slk archive/retrieve may use much memory and CPU time – careful with parallel slk calls on one node
  • slk is hanging / unresponsive
  • Non-recursive is semi-recursive
  • slk writes no output in non-interactive mode
  • slk never writes to stderr
  • slk move and slk rename
  • slk archive compares file size and timestamp prior to overwriting files
  • Availability of archived data and modified metadata might be delayed by a few seconds
  • A file listed by slk list is not necessarily available for retrieval yet
  • failed/canceled slk archive/retrieve calls leave file fragments
  • Pagination mode of slk list
  • slk tag cannot be applied on individual files
  • slk does not have a –version flag
  • Update interval of progress bars (slk archive, group, owner, retrieve, tag)
  • Using slk list to print search results
  • slk performance on different node types
  • group memberships of user updated on login
  • LDAP user not known to StrongLink prior to first login
  • slk retrieve does not overwrite files but creates duplicates
  • “slk retrieve /source/ /target” and “slk retrieve /source /target” are not the same
  • Difficult to recognize if “slk group” fails to change all target files
  • slk archive might create namespaces with “.”
  • “slk bad_input” returns exit code 0
  • slk cannot handle a path with “//” (double slash)
  • “Login Unsuccessful - Incorrect Credentials” or “Session key has expired” when StrongLink not available
  • Filtering slk list results with “*”
  • slk chmod -R modifies permissions of each file several times
  • How to search non-recursively in a namespace
  • Terminal cursor disappears if slk command with progress bar is canceled
  • error “conflict with jdk/…” when the slk module is loaded
  • slk needs a specific Java version
  • slk search yields RQL parse error
  • slk login asks me to provide a hostname and/or a domain
  • Archival fails and Java NullPointerException in the log
  • slk ERROR: Unhandled error occurred, please check logs
  • slk login ERROR: Unhandled error occurred, please check logs
  • permissions of retrieved files are “rw——-” although umask is set differently
  • slk archive: Exception …: lateinit property websocket has not been initialized
  • slk archive runs infinitely when files of 0 byte size are archived
  • slk delete failed, but nevertheless file was deleted
• Archivals to tape
  • Introduction and Summary
  • Useful information on slk archive
  • What to do when slk archive was interrupted/killed?
  • Validate archivals
  • Archival script templates
• Retrievals from tape
  • Introduction and Summary
  • Recommendations for usage of slk retrieve
  • Speed up your retrievals
  • Retrieval script templates
• File Search and Metadata
  • Metadata in StrongLink
  • Set metadata
  • Print metadata
  • Search files by metadata
• slk usage examples
  • Obtain Access Token
  • Check if access token is still valid
  • Archival
  • Search files
  • Generate search queries
  • List files
  • Retrieve files
  • tag files (set metadata)
  • Change permissions and group of files and directories
  • Get user/group IDs and names
  • slk in batch jobs on compute nodes
• Switching: pftp to slk
  • Nomenclature
  • Login
  • Archival of a single file
  • Archival of a whole directory
  • Retrieval of a single file
  • Retrieval of a whole directory
  • Listing the content of a folder in the archive
• manual pages slk
  • slk: official StrongLink Command Line Interface
  • slk helpers: slk extension provided by DKRZ
  • Reference: metadata schemata
  • Reference: StrongLink query language
  • Official StrongLink Command Line Interface Guide
• FAQ
  • General information about the HSM system
  • Data Migration
  • Training, Questions and Adaption of Workflows
  • Archiving and Retrieval
  • Additional features
  • Advanced Technical Aspects
  • Common issues
  • Changelog