Swordfish Scalable Storage Management API User's Guide

Version 1.0.6

ABSTRACT: The Swordfish Scalable Storage Management API defines a RESTful interface and a standardized data model to provide a scalable, customer-centric interface for managing storage and related data services.

Publication of this Working Draft for review and comment has been approved by the Scalable Storage Management Technical Work Group. This draft represents a 'best effort' attempt by the Scalable Storage Management Technical Work Group to reach preliminary consensus, and it may be updated, replaced, or made obsolete at any time. This document should not be used as reference material or cited as other than a 'work in progress.' Suggestions for revision should be directed to http://www.snia.org/feedback.

Working Draft

Last Updated 13 February 2018

USAGE

Copyright © 2018 SNIA. All rights reserved. All other trademarks or registered trademarks are the property of their respective owners.

The SNIA hereby grants permission for individuals to use this document for personal use only, and for corporations and other business entities to use this document for internal use only (including internal copying, distribution, and display) provided that:

  1. Any text, diagram, chart, table or definition reproduced must be reproduced in its entirety with no alteration, and,

  2. Any document, printed or electronic, in which material from this document (or any portion hereof) is reproduced must acknowledge the SNIA copyright on that material, and must credit the SNIA for granting permission for its reuse.

Other than as explicitly provided above, you may not make any commercial use of this document, or any portion thereof, or distribute this document to third parties. All rights not explicitly granted are expressly reserved to SNIA.

Permission to use this document for purposes other than those enumerated above may be requested by emailing tcmd@snia.org. Please include the identity of the requesting individual and/or company and a brief description of the purpose, nature, and scope of the requested use.

All code fragments, scripts, data tables, and sample code in this SNIA document are made available under the following license:

BSD 3-Clause Software License

Copyright © 2018, The Storage Networking Industry Association

Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

DISCLAIMER

The information contained in this publication is subject to change without notice. The SNIA makes no warranty of any kind with regard to this specification, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. The SNIA shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use.

Suggestions for revisions should be directed to http://www.snia.org/feedback/.

Copyright © 2016-2018 Storage Networking Industry Association.

Revision History

Revision history
Date Revision Notes
19 September 2016 1.0.0 Initial Release
12 October 2016 1.0.1 General clean up and formatting consistency
A discussion of unused CoS and LoS entries in ServiceCatalog
Improve purpose for many use cases
1 November 2016 1.0.2 Corrected XREF link formatting
24 January 2017 1.0.3 Additional use cases and new document section addressing client considerations
25 April 2017 1.0.4 Update cross-references.
3 October 2017 1.0.5 Minor updates and corrections
13 February 2018 1.0.6 Added on-demand replication use cases

Suggestion for changes or modifications to this document should be sent to the SNIA Scalable Storage Management (SSM) Technical Working Group at http://www.snia.org/feedback/.

Contact SNIA

SNIA Web Site

Current SNIA practice is to make updates and other information available through their web site at http://www.snia.org.

FEEDBACK AND INTERPRETATIONS

Requests for interpretation, suggestions for improvement and addenda, or defect reports are welcome. They should be sent via the SNIA Feedback Portal at http://www.snia.org/feedback/ or by mail to the Storage Networking Industry Association, 4360 ArrowsWest Drive, Colorado Springs, Colorado 80907, U.S.A.

INTENDED AUDIENCE

This document is intended for use by individuals and companies engaged in storage management.

VERSIONING POLICY

This document is versioned material. Versioned material shall have a three-level revision identifier, comprised of a version number 'v', a release number 'r' and an errata number 'e'. Future publications of this document are subject to specific constraints on the scope of change that is permissible from one revision to the next and the degree of interoperability and backward compatibility that should be assumed between products designed to this standard. This versioning policy applies to all SNIA Swordfish versioned materials.

Version Number: Versioned material having version number 'v' shall be backwards compatible with all of revisions of that material that have the same version number 'v'. There is no assurance of interoperability or backward compatibility between revisions of a versioned material with different version numbers.

Release Number: Versioned material with a version number 'v' and release number 'r' shall be backwards compatible with previous revisions of the material with the same version number, and a lower release number. A minor revision represents a technical change to existing content or an adjustment to the scope of the versioned material. Each minor revision causes the release number to be increased by one.

Errata Number: Versioned material having version number 'v', a release number 'r', and an errata number 'e' should be backwards compatible with previous revisions of the material with the same version number and release number ("errata versions"). An errata revision of versioned material is limited to minor corrections or clarifications of existing versioned material. An errata revision may be backwards incompatible, if the incompatibility is necessary for correct operation of implementations of the versioned material.

Acknowledgements

The SNIA Scalable Storage Management Technical Work Group, which developed and reviewed the work in progress, would like to recognize the significant contributions made by the following members:

Member Representatives
Broadcom Limited Richelle Ahlvers
Dell Inc. Patrick Boyd
George Ericson
Michael Raineri
Rich Roscoe
Hitachi Data Systems Eric Hibbard
Hewlett Packard Enterprise Jeff Hilland
Chris Lionetti
John Mendonca
Doug Voigt
Inova Development Inc. Karl Schopmeyer
Intel Corporation Slawek Putyrski
Paul von Behren
Microsemi Anand Nagarjan
Microsoft Corporation Hector Linares
Jim Pinkerton
Michael Pizzo
Scott Seligman
NetApp, Inc. Don Deel
Nilesh Maheshwari
ScienceLogic Patrick Strick
VMware, Inc. Murali Rajagopal

About SNIA

The Storage Networking Industry Association (SNIA) is a non-profit organization made up of member companies spanning information technology. A globally recognized and trusted authority, SNIA’s mission is to lead the storage industry in developing and promoting vendor-neutral architectures, standards and educational services that facilitate the efficient management, movement and security of information.

Acknowledgements

The SNIA Scalable Storage Management Technical Work Group, which developed and reviewed work in progress, would like to recognize the significant contributions made by the following members:

Member Contributors
Broadcom Limited Richelle Ahlvers
Dell Inc. Patrick Boyd
George Ericson
Michael Raineri
Rich Roscoe
Hitachi Data Systems Eric Hibbard
Hewlett Packard Enterprise Jeff Hilland
John Mendonca
Inova Development Inc. Karl Schopmeyer
Intel Corporation Slawek Putyrski
Paul von Behren
Microsoft Corporation Hector Linares
Jim Pinkerton
Michael Pizzo
Scott Seligman
NetApp, Inc. Don Deel
Nilesh Maheshwari
Nimble Storage Chris Lionetti
VMware, Inc. Murali Rajagopal

Table of Contents


1. Introduction
1.1. Audience
1.2. Documentation structure
1.3. Configuration assumptions
1.4. Knowledge assumptions
1.5. Related documents
2. General query syntax
2.1. OData conformance
2.2. Query method
2.3. Query Headers
2.4. Service root
2.5. Resource path
2.6. Query options
2.7. Filter expressions
2.8. HTTP status codes
3. Actors
3.1. Actors
3.2. CloudAdmin
3.3. DevOps
3.4. Storage Admin
4. Management domains
4.1. Management Domain Overview
4.2. Application storage management domain
4.3. Block storage management domain
4.4. Service catalog management domain
4.5. File system storage management domain
5. User Guidance
5.1. Default Class of Service
6. Use Cases
6.1. Add Volume
6.2. Allocate VM space
6.3. Allocate Volume
6.4. Allocate Volume using Default Class of Service
6.5. Health updates
6.6. Create class of service
6.7. Create file share
6.8. Create file system
6.9. Create line of service
6.10. Create StorageGroup
6.11. Create storage pool
6.12. Discover Class of Service
6.13. Expand capacity of a storage volume
6.14. Find storage service
6.15. Get capacity by class of service
6.16. List supported line of service options
6.17. Subscribe to Threshold Events
6.18. Set Capacity Usage Threshold
6.19. Protect Volumes of an application

1. Introduction

1.1. Audience

This guide is intended to provide a common repository of best practices, common tasks and education for the users of the Swordfish API. Each use case includes an indication of the classes of API users who are most likely to find the case useful.

1.2. Documentation structure

This document begins with a set of information intended to provide a solid foundation for readers new to restful APIs in general and Swordfish in particular. While this material is no replacement for a thorough understanding of the Swordfish specification and the material that it references, it is intended as a stand alone document that can provide a solid introduction to Swordfish.

Based on that foundational material, this document then presents a set of Use Cases intended to capture common tasks and best practices that can be used to exercise the breadth and strength of the Swordfish API. In general, the guide is structured to provide more basic use cases first, and examine common refinements and options at the same time. More advanced tasks are handled later in the guide, and assume the prior skills and assumptions have been mastered.

For each use case, this guide will use a common template. Table 1 lists each field of the template and its description.

Table 1: Guidelines for the Use Case Template

Name Description
Title A description of the high-level scope of the Use Case
Summary A high-level summary of the use case
Purpose The intended goals or motivations for the use case
Who The Actor(s) who are likely to use this use case. The actor description also includes a list of other use cases of interest for a named actor
Management Domain The Management Domain(s) applicable to this use case. The domain description also includes a list of other use cases of interest for the named management domains
Triggers A description of likely business conditions or goals that would make this use case useful
Detailed Context A detailed description of the operations environment and configuration assumptions for this use case
Preconditions Pre-existing knowledge, configurations or capabilities
Inputs A set of parameters and values that are used to adapt a generic use case to a specific business needs. Where appropriate, the parameter description will include a data type (e.g., {CAPACITY}: desired storage capacity (int64))
Basic Course of Events A sequence of API requests, including required headers, the body of the request, and the expected reply
Configuration Impacts Changes to the storage configurations caused by the use case
Failure Scenario Common failure conditions encountered in this use case
See Also Other Use Cases that may be of interest

1.3. Configuration assumptions

This document assumes that some fundamental configuration issues have been properly addressed, and will not need to be addressed in any detail. In particular, this document assumes:

1.4. Knowledge assumptions

The Swordfish API conforms to the standards defined in the Redfish API. More generally, it is provides a RESTful interface. The reader is assumed to be familiar with common conventions for RESTful APIs. Those readers who are interested in additional background information are encouraged to refer to the following sources:

This User's Guide is part of the documentation suite for the Swordfish API. Readers are encouraged to refer to the following for additional information: - Swordfish API Specification - Swordfish Tutorials

2. General query syntax

2.1. OData conformance

Swordfish, like RedFish, conforms to the OData Version 4. It conforms to the query syntax and URL structure that that standard defines. The high-level assumptions are summarized here. Readers who need more information should refer to the OData standard.

All Swordfish URLs follow a general form of:

2.2. Query method

Swordfish queries support four query methods. Each query URL must include exactly one of of the query methods listed in Table 1: Query Methods.

Table 1: Query Methods

Method Action
GET Retrieve the current state or settings of the named Resource Path as seen through the Service Root
POST Create a new object under the named Resource Path
PUT Replace the object referenced by the named Resource Path
DELETE Delete the object referenced by the named Resource Path
PATCH Update the object referenced by the named Resource Path
HEAD Validates a GET request against the named Resource Path without returning the HTML headers for the response without the result of the query

2.3. Query Headers

All HTTP requests and responses in a compliant Swordfish implementation support the HTTP headers required by the Redfish Protocol Specification. The supported headers are reproduced here for convenience.

Request headers

Header Supported Values Notes
Accept RFC 7231 Indicates to the server what media type(s) this client is prepared to accept. Services shall support requests for resources with an Accept header including application/json or application/json;charset=utf-8. Services shall support requests for metadata with an Accept header including application/xml or application/xml;charset=utf-8.
Content-Type RFC 7231 Describes the type of representation used in the message body. Content-Type shall be required in requests that include a request body. Services shall accept Content-Type values of application/json or application/json;charset=utf-8.
OData-Version 4.0 Services shall reject requests which specify an unsupported OData version. If a service encounters a version that it does not support, the service should reject the request with status code 412. If client does not specify an Odata-Version header, the client is outside the boundaries of this specification.
Authorization RFC 7235, Section 4.2 Required for Basic Authentication
User-Agent RFC 7231 Required for tracing product tokens and their version. Multiple product tokens may be listed.
Host RFC 7230 Required to allow support of multiple origin hosts at a single IP address.
Origin W3C CORS, Section 5.7 Used to allow web applications to consume Redfish Service while preventing CSRF attacks.
If-Match RFC 7232 If-Match shall be supported on PUT and PATCH requests for resources for which the service returns ETags, to ensure clients are updating the resource from a known state.
X-Auth-Token Opaque encoded octet strings Used for authentication of user sessions. The token value shall be indistinguishable from random.

Response headers

Header Supported Values Notes
OData-Version 4.0 Describes the OData version of the payload that the response conforms to.
Content-Type RFC 7231 Describes the type of representation used in the message body. Services shall specify a Content-Type of application/json when returning resources as JSON, and application/xml when returning metadata as XML. ;charset=utf-8 shall be appended to the Content-Type if specified in the chosen media-type in the Accept header for the request.
ETag RFC 7232 An identifier for a specific version of a resource, often a message digest. Etags shall be included on responses to GETs of ManagerAccount objects.
Server RFC 7231 Required to describe a product token and its version. Multiple product tokens may be listed.
Link   Link headers shall be returned as described in the clause on Link Headers
Location RFC 7231 Indicates a URI that can be used to request a representation of the resource. Shall be returned if a new resource was created. Location and X-Auth-Token shall be included on responses which create user sessions.
Cache-Control RFC 7234 This header shall be supported and is meant to indicate whether a response can be cached or not.
Access-Control-Allow-Origin W3C CORS, Section 5.7 Prevents or allows requests based on originating domain. Used to prevent CSRF attacks.
Allow POST, PUT, PATCH, DELETE, GET, HEAD Shall be returned with a 405 (Method Not Allowed) response to indicate the valid methods for the specified Request URI. Should be returned with any GET or HEAD operation to indicate the other allowable operations for this resource.
WWW-Authenticate RFC 7234, Section 4.1 Required for Basic and other optional authentication mechanisms. See the Security clause for details.
X-Auth-Token Opaque encoded octet strings Used for authentication of user sessions. The token value shall be indistinguishable from random.

2.4. Service root

This is the base of all Swordfish URL's. A GET request to the Service Root will return an overview of the services provided by a given Swordfish service. In addition, the Service Root will include versioning information.

All Service Root URLs that are compliant with the Swordfish specification will be of the form https://hostName/redfish/v1, where hostName specifies the system (and optionally port number), of the Swordfish service provider.

2.5. Resource path

The Resource Path identifies the specific object (or collection of objects) that is the target of the Swordfish query. As with OData, Swordfish Resource Paths can identify: - A singleton object (e.g., a specific storage LUN or Volume) - A collection of objects (e.g., the list of all LUNs provided by a specific storage array)

At the highest level, Swordfish identifies two major collections: Storage Systems and Storage Services.

2.6. Query options

Swordfish queries can include arbitrary sets of query options to further refine the target of given query or the actions being requested of that target. These general query options are summarized in Table 3: Query Options.

Note: Additional query options may be supported (or constrained) for a specific query target or resource path. These target-specific query options will be addressed in specific use case descriptions, as required.

Table 3: Query Options

Parameter Name Arguments Notes
$skip=n Integer Omit the first n entries in the collection from the returned set of objects (required by redfish)
$top=n Integer Return, at most, the first n entries in the returned set of objects (required by redfish)
$filter=condition Filter Expression Returns only the members of the named collection that match the provided logical expression (required by swordfish)
$expand=target Expand Expression Expand additional detail on the target property(s) in the returned result set (required by swordfish)
$select=property list Comma-separated list of object properties Return the named properties for each object in the result set, rather than the entire object (required by swordfish)
$orderby=filter condition Filter Expression sort the result set by the output values from the filter expression (required by swordfish)

2.7. Filter expressions

 

  1. Simple example $filter=(age gt 30) A group of people never to trust.

For more information see Filter Expression in the OData specification.

2.8. HTTP status codes

Swordfish clients may receive any of the standard HTTP status codes. For status codes greater than 400, the server can also return extended status information as a simple JSON object (see Redfish Specification for more information).

3. Actors

3.1. Actors

This document covers a broad range of common use cases and storage management operations.

In an attempt to serve as both a introductory text and as a reference tool, the use cases have been grouped in a number of different ways:

3.2. CloudAdmin

A Cloud Administrator (CloudAdmin) is a converged infrastructure administrator, working with systems that are:

The CloudAdmin role in an enterprise or service provider is the individual or group primarily responsible for managing the operational lifecycle of a cloud, virtualization, converged environment that consists of the workloads, resource abstractions, storage, networking, and compute.

Also referred to as a "cloud architect", this role:

Use Cases

3.3. DevOps

A member of the DevOps group:

This role is typically aligned with the business unit. Their focus is the delivery, deployment, and maintenance of apps on IaaS and PaaS resources. This role is not typically a deep subject matter expert in compute, storage, or networking. From a development perspective, their desire is to treat the infrastructure as a programmable subsystem that presents resources on-demand.

This role:

Use Cases

3.4. Storage Admin

A storage administrator designs storage solutions for modern environments, including:

The storage admin role in an enterprise or service provider can have responsibility for managing the operational lifcycle of storage in the datacenter. In particular:

Use Cases

4. Management domains

4.1. Management Domain Overview

This document covers a broad range of common use cases and storage management operations.

In an attempt to serve as both a introductory text and as a reference tool, the use cases have been grouped in a number of different ways:

4.2. Application storage management domain

This domain manages the interface between applications and the storage that they rely upon.

StorageGroups provide a means to collectively manage the Volumes and FileShares utilized by an Application. The StorageGroup specifies whether the collected resources are managed so that storage is updated or replicated consistently across all members. Additionally, the StorageGroup provides the means to atomically expose (or hide) the collected resources to (or from) host endpoints.

Use cases

4.3. Block storage management domain

Many devices and services provide their storage capacity to external devices and applications through block-based protocols to storage devices. This domain includes the management of resources that provide block-based access to storage.

Block-based storage is represented by a Volume. This domain provides for the discovery and provisioning of Volumes and for maintaining relationships to Device, Endpoint, StorageService, StorageGroup, StoragePool, and ComputerSystem resources.

Use Cases

4.4. Service catalog management domain

Swordfish supports access to, and management of, a catalog of service options, (see ITIL glossary and abbreviations), supported by storage services.

The ClassOfService resource represents a service option that may be used to specify requirements for utility or warranty when provisioning a resource. Currently ClassOfService is defined for use in the Block Storage, File System, and ApplicationStorage domains.

The service catalog for each StorageService is represented by a collection of references to supported ClassOfService resources. Each ClassOfService is known minimally by a Name and a unique Identifier. When a ClassOfService is specified for a resource, the StorageService shall attempt to maintain that resource in compliance to the requirements of that ClassOfService. The requirements may be specified informally by text in the Description property or may be specified formally by the property values of embedded options related to specific lines of service.

The embedded service options are described by values of complex types representing lines of service.

Over time, as the service catalog is continually updated to match evolving user needs and service provider offerings, it is expected that the catalog will contain entries (one or more ClassOfSerice or LineOfService instances) that are not currently active.

Data protection

The primary storage is described by a ClassOfService resource. That ClassOfService may aggregate any number of data protection service options. Each instance of a data protection service option describes the characteristics a replication session that shall be maintained for the containing primary storage resource.

For additional information, see the definitions for DataProtectionLineOfService and DataProtectionLoSCapabilities.

Data security

An instance of a data security service option describes an optional set of data security requirements. A data security Service option is typically aggregated into a ClassOfService resource that is associated with storage. At most one data security service option may be aggregated into a ClassOfService resource. When storage is provisioned with that ClassOfService, it will provide the currently specified data security characteristics.

A data security service option may specify data security characteristics that enable the storage system to be used in an environment where compliance with an externally-specified security standard or standards is required. Examples of such standards include FIPS-140, HIPAA and PCI. In this case, the names of the standard or standards can usefully be included in the user/admin-visible name of the instance. With the notable exception of FIPS-140, compliance requires measures well beyond the means of a storage system to provide (e.g., both HIPAA and PCI impose significant requirements on administration and operation of the data center), so this approach promises that the storage system will do its part in supporting compliance, but does not (and cannot) promise that the storage system will deliver full compliance by itself.

The description attribute value may include human readable information including:

NOTE For comparable cryptographic strengths, (see NIST SP 800-57 part 1)

NOTE For symmetric encryption algorithm key sizes, 112 bits is the 3DES key size and 128, 192, and 256 bits are options for AES key sizes.`

NOTE MediaEncryptionStrength includes the case where metadata about the data must be encrypted. (e.g. data presence vs. absence in a thin volume, array filesystem metadata) The implementation may be self-encrypting drives or encryption in the storage system’s drive controller. Keys may be drive or array resident or externally managed (e.g., via KMIP).

For additional information, see the definitions for DataSecurityLineOfService and DataSecurityLoSCapabilities.

Data storage

Each data storage service option describes characteristics of the storage at a particular location. A class of service will have at most one data storage service option, which describes the storage specified by that class of Service.

For additional information, see the DataStorageLineOfService and DataProtectionLoSCapabilities.

IO connectivity

An IO connectivity service option specifies the characteristics of storage connectivity. For each value of AccessProtocol, at most one IO connectivity service option may be aggregated into a class of service.

NOTE: If used within a ClassOfService for Storage Provisioning, this value constrains the set of connections used to expose that storage.

For additional information, see the IOConnectivityLineOfService and IOConnectivityLoSCapabilities.

IO performance

An IO performance service option specifies a choice of performance characteristics as viewed through the data path to the storage. This is affected by choices of storage and connection technologies.
At most one IO performance service option may be aggregated into a ClassOfService for a storage resource. When storage is provisioned with that ClassOfService, it should provide at least the specified performance.

For additional information, see the IOConnectivityLineOfService and IOConnectivityLoSCapabilities.

Use cases

4.5. File system storage management domain

FileSystems provide access to byte-accessible storage through file-based protocols. This domain includes the management of resources that provide file-based access to storage.

File-based storage is represented by a FileSystem resources. Remote access to portions of a FileSystem is provided by FileShare resources.

Use cases

5. User Guidance

5.1. Default Class of Service

If a pool, storage volume or other contruct is created with no specified class of service when a class of service exists, the implementation will attempt to apply the DefaultClassOfService.

6. Use Cases

Title Description
Add volume Add a volume to a StorageGroup
Allocate VM Space Allocate space for a new VM
Allocate volume Allocate a new volume
Allocate volume using default class of service Allocate a new volume without default class of service
Create class of service Create a class of service
Create file share Create a file share
Create file system Create a files system
Create line of service Create a line of service
Create StorageGroup Create a StorageGroup
Create Storage Pool Create a StoragePool
Discover class of service Discover a class of service
Expand Volume Expand capacity of a Volume
Find storage service Find a services support a class of service
Get capacity by class of service Summarized capacity by a class of service
Health Updates Example of health status monitoring
List supported line of service options List Supported Line Of Service Options
Set space thresholds Set Capacity Usage Thresholds
Subscribe to thresholds events Subscribe to Threshold Events
Protect Volumes of an application Protect Volumes of an Application

6.1. Add Volume


Summary: Add Volumes to a StorageGroup

Purpose: Add Volumes to a StorageGroup

Who: CloudAdmin

Management Domain: Block storage management

Triggers: A need for expanded capacity in a StorageGroup

Detailed Context: A Volume has been created, and needs to be added to an existing StorageGroup.

Preconditions: None.

Inputs: - {SGURL}: URL for storage group - One or more URLs for Volumes to add [{VURI1}, {VURI2}, ...]

Basic Course of Events:

  1. Add the Volume(s) the Members collection of the StorageGroup

Postconditions: The selected Volume(s) are added to the Members collection for the Storage Group.

Failure Scenario: None defined.

See also: None defined.


6.2. Allocate VM space


Summary: Allocate space for a VM (e.g., Docker) deployment

Purpose: Allocate space for a VM (e.g., Docker) deployment

Who: DevOps

Management domain: Block storage management

Triggers: DevOp needs to set up a new VM to host a Ceph OSD service.

Detailed context: Some cloud web services deploy a large number of VMs (or containers) that do not keep data in traditional files or block storage; instead, they store data in remote object stores. This use case relates to setting up systems hosting Ceph as an object store. Ceph is deployed as a set of collaborating systems providing fault tolerance; this use case addresses one VM dedicated as a Ceph Object store (OSD) system. The VM is allocated with a virtual disk for the OS and Ceph software; an external disk is allocated for Ceph data. For this use case, the DevOP prefers using an HDD rather than an SSD and needs at least 400 Gigabytes (400,000,000,000 bytes).

Preconditions: The catalog (see JBOD discovery use case) must exist and include at least one available HD

Inputs:

Basic course of events:

Postconditions: The Devop can now install and configure Ceph to use the selected drive

Failure Scenario: None defined.

See also: None defined.


6.3. Allocate Volume


Summary: Allocate a Volume

Purpose: Allocate a Volume with a known capacity and class of service.

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: Block storage management

Triggers: Need to allocate storage for a new application.

Detailed Context: The admin needs to satisfy a service request to provide a given amount of storage to an application, and to assure a given class of service.

Preconditions: None.

Inputs:

Basic Course of Events:

  1. Post the definition of the new volume to the Volumes resource collection.

This instructs the service to allocate a new volume of the requested size that meets the requirements of the specified class of service. Since additional details are not provided, the service is free to allocate the storage from any of its storage pools that can satisfy the request.

Postconditions: The selected volumes are added to the collection for the Storage Group.

Failure Scenario: None defined

See also: None defined.


6.4. Allocate Volume using Default Class of Service


Summary: Allocate a Volume using the Default Class of Service

Purpose: Allocate a Volume with a known capacity and without specifying the class of service, invoking the DefaultClassOfService.

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: Block storage management

Triggers: Need to allocate storage for a new application.

Detailed Context: The admin needs to satisfy a service request to provide a given amount of storage to an application, but does not specify a given class of service. The system will attempt to provision the system with the specified ClassOfService in DefaultClassOfService in either the given StoragePool or the StorageService as available.

Preconditions: The DefaultClassOfService property is set in either the StoragePool or StorageService.

Inputs:

Basic Course of Events:

  1. Post the definition of the new volume to the Volumes resource collection.

This instructs the service to allocate a new volume of the requested size that meets the requirements of the specified class of service. Since additional details are not provided, the service is free to allocate the storage from any of its storage pools that can satisfy the request.

Postconditions: The selected volumes are added to the collection for the Storage Group.

Failure Scenario: None defined

See also: None defined.


6.5. Health updates


Summary: Retrieve storage metrics information for a storage volume.

Purpose:

Who: DevOps at an enterprise

Management domain: Block storage management

Trigger: Lower than expected application requests completed per second

Detailed context: Devops creates diagnostics scripts that retrieve performance information from multiple layers in the application stack, including infrastructure, to help identify potential bottlenecks during production hours. Devops already understand that the issue is not in the application layers so now they have to dig deeper. Members of devops are typically not experts in infrastructure compute, storage, and networking so they need simple scripts that can provide the information they require without a deep understanding of underlying hardware. With the storage administrator's help, devops use a few simple GETs on storage objects related to their application. To help simplify the query, the storage administrator tags the volumes associated with the application. Devops can use stack wide performance metrics to quickly isolate potential bottlenecks that may be contributing to the slow down in the application.

Preconditions:

Inputs: DevOps knows the storage system name or IP address with the Swordfish service and the storage pool volume names.

Basic course of events:

  1. uses GET operations to look at metrics of the storage volume

Postconditions: None defined.

See also: None defined.


6.6. Create class of service


Summary: Create a class of service

Purpose: Create a new class of service in the service catalog to match a newly available type of storage

Who: StorageAdmin

Management Domain: Block storage management, Service catalog management

Triggers: The administrator has determined that a new class of service needs to be created to reflect a new class of SSD storage in the infrastructure.

Detailed Context: This is a simple scenario where the primary characteristic is the enhanced performance available of SSD drives.

Preconditions: None defined.

Inputs:

Basic Course of Events:

  1. Create ClassOfService
  1. Response:

Postconditions: The requested class of service is added to the ClassesOfService collection.

Failure Scenario: None defined.

See also: None defined.


6.7. Create file share


Summary: Create a file share

Purpose: Share an existing file system as /Shares/MyShare

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: File system storage management

Triggers: None defined.

Detailed Context: Create a share starting at /Shares/MyShare.

Preconditions: None defined.

Inputs:

Basic Course of Events:

1. Create a file share

{
        "Name": "MyShare",
        "Description": "Share of files under MyShare.",
    "SharedFilePath": "/Shares/MyShare"
    }

Postconditions: The requested file share is added to the ExportedShares collection for the file system.

Failure Scenario: None defined.

See also: None defined.


6.8. Create file system


Summary: Create a file system

Purpose: Create a file system with a given capacity and performance level.

Who: StorageAdmin

Management Domain: File system storage management

Triggers: None defined.

Detailed Context: Create a 100 TB file system based on SSD class storage.

Preconditions: None defined.

Inputs:

json { "Data": { "ProvisionedBytes": 100000000000000; "IsThinProvisioned": true; } }

Basic Course of Events:

  1. Create FileSystem

Postconditions: The requested file system is added to the FileSystems collection for the Storage Service.

Failure Scenario: None defined.

See also: None defined.


6.9. Create line of service


Summary: Create a line of service to reflect the performance characteristics of SSD storage

Purpose: The definition is created here in preparation of creating ClassOfService instances that include a requirement for SSD storage performance.

Who: StorageAdmin

Management Domain: Block storage management, Service catalog management

Triggers: None defined.

Detailed Context: SSD storage is introduced and need a new performance line of service to reflect their capability.

Preconditions: None defined.

Inputs:

json { "Name": "NewSSDLoS", "IoOperationsPerSecondIsLimitedBoolean": false, "MaxIoOperationsPerSecondPerTerabyte": 100000, "AverageIoOperationLatencyMicroseconds": 10 }

Basic Course of Events:

1. Get existing supported lines of service

GET /redfish/v1/StorageServices(1)/Links/IOPerformanceLoSCapabilities/SupportedIOPerformanceLinesOfService

2. Create new line of service

PATCH /redfish/v1/StorageServices(1)/Links/IOPerformanceLoSCapabilities

Postconditions: The requested line of service is added to the SupportedIOPerformanceLinesOfService of the Storage Service.

Failure Scenario: None defined.

See also: None defined.


6.10. Create StorageGroup


Summary: Create a StorageGroup

Purpose: Create a StorageGroup

Who: CloudAdmin

Management Domain: Block storage management

Triggers: None defined.

Detailed Context: Create a collection of application storage that is exposed to an application and managed as a unit.

Preconditions: None defined.

Inputs:

Basic Course of Events:

  1. Create StorageGroup

Postconditions: The requested StorageGroup is added to the Members collection for the Storage Service.

Failure Scenario: None defined.

See also: None defined.


6.11. Create storage pool


Summary: Create a StoragePool

Purpose: Create a StoragePool

Who: StorageAdmin

Management Domain: Block storage management

Triggers: Users need to allocate storage with characteristics satisfied by a class of service.

Detailed Context: Create a storage pool containing an amount of storage that can be used to create a requested class of service. The storage pool implementation will attempt to find and allocate enough storage that will satisfy the request. No metadata or snapshot storage is reserved.

Preconditions: None defined.

Inputs:

{
    "Data": {
        "ProvisionedBytes": 100000000000000,
        "IsThinProvisioned": false
    }
}

Basic Course of Events:

  1. Create StoragePool
{
    "Name": "SSD",
    "Description": "100 TB pool of SSD class storage.",
    "Capacity": {
        "Data": {
            "ProvisionedBytes": 100000000000000
        },
        "IsThinProvisioned": false
    },
    "ClassesOfService": {
        "Members": [{
            "@odata.id": "/redfish/v1/StorageServices/Members(1)/ClassesOfService(SSD)"
        }]
    }
}

Post-Conditions: The requested StoragePool is added to the collection for the Storage Service.

Failure Scenario: None defined.

See also: None defined.


6.12. Discover Class of Service


Summary: Discover a suitable class of service

Purpose: Identify an existing Class of Service that meets given requirements

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: Block storage management, File system storage management, Application storage management, Service catalog management

Trigger: Need to provide a ClassOfService reference in another request (e.g., Create Volume )

Detailed Context: Identify an existing class of service that would satisfy a set of characteristics (e.g., capacity, service level, throughput)

Preconditions: None.

Inputs: A set of storage attributes that must be satisfied, phrased as a Filter Expression. For this example: we want at least 10000 IOPS per Terabyte as a desired minimum supported performance level

Basic Course of Events:

  1. Use GET operation to retrieve possible ClassOfService entries
  1. Select a Class of Service based on returned values.

Postcondition: The catalog now has entries for each new disk drive including class of service and capacity.

Failure Scenario:

Error Code Cause
204 No existing Class of Service entries match the required criteria

See also: Allocate Volume


6.13. Expand capacity of a storage volume


Summary: Retrieve storage capacity information for a storage pool to check available capacity before expanding a storage volume.

Purpose:

Who: StorageAdmin at an enterprise

Management domain: Block storage management

Trigger: Low available capacity warning

Detailed context: The enterprise administrator offers managed storage services to business units in her organization. The administrator has established an SLA with application owners for storage volume expansion:

The application monitoring tool sends a warning alert to the administrator's datacenter monitoring tool indicating that a storage volume's capacity is at 80% and must be increased to ensure the application does not experience unplanned downtime. The administrator first checks the available capacity on the storage pool for a given storage class of service. Since the storage system is new, there is sufficient capacity available. Next the administrator figures out that the storage volume has never been expanded based on the original size tag. At the start of the maintenance window, non-business hours, the administrator initiates the expand action. For this particular storage system, the expand is a long running action so the administrator tracks the progress using the associated task.

Preconditions:

Inputs: Enterprise administrator knows the storage system name or IP address with the Swordfish service. Administrator knows the storage pool name and storage volume name.

Basic Course of Events:

  1. uses GET operations to look at information about storage volume and confirm low capacity and the current size is less than 2TB
{
  "@Redfish.Copyright": "Copyright 2015-2016 SNIA. All rights reserved.",
  "@odata.context": "/redfish/v1/$metadata#Volume.Volume",
  "@odata.id": "/redfish/v1/StorageServices/1/Volumes/61001234876545676100123487654567",
  "@odata.type": "#Volume_1_0_0.Volume",
  "Id": "61001234876545676100123487654567",
  "Name": "Volume 1",
  "Description": "application storage",
  "Identifiers": [
    {
      "DurableNameFormat": "NAA6",
      "DurableName": "61001234876545676100123487654567"
    }
  ],
  "Manufacturer": "SuperDuperStorageProvider",
  "Status": {
    "State": "Enabled",
    "Health": "OK"
  },
  "BlockSizeBytes": 512,
  "LowSpaceWarningThresholdPercent": [
    80,
    null,
    null,
    null,
    null
  ],
  "Capacity": {
    "Data": {
      "ConsumedBytes": 879609302221,
      "AllocatedBytes": 879609302221,
      "GuaranteedBytes": 549755813888,
      "ProvisionedBytes": 1099511627776
    },
  }
}
  1. Uses GET operations to look at information about storage pool to confirm it has more than 25% available capacity
  1. Use expand action to increase size of storage volume by 25%

Postconditions: The volume's capacity expands by 25%. Administrator needs to track the associated task to know when the volume expansion completes.

See also: None defined.


6.14. Find storage service


Summary: Find storage services that support specified class of service characteristics.

Purpose: Identify storage services that support needed service options.

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: Block storage management, File system storage management, Application storage management, Service catalog management

Trigger: Need to provide a Storage Service reference to support a more specific request (e.g., Create Volume)

Detailed Context: Identify existing storage services that would satisfy a set of service characteristics (e.g., capacity, service level, throughput)

Preconditions: None.

Inputs: A set of storage attributes that must be satisfied, phrased as a Filter Expression. For this example:

Basic Course of Events:

  1. Use GET operation to retrieve possible StorageService entries
  1. Select a Class of Service URL from the returned values.

Postconditions: None defined.

Failure Scenario:

Error Code Cause
204 No existing Class of Service entries match the required criteria

See also: None defined.


6.15. Get capacity by class of service


Summary: Get capacity by class of service

Purpose:

Who: CloudAdmin, StorageAdmin, DevOps

Management domain: Block Storage Management, File system storage management, Application storage management, Service catalog management

Trigger: An alert indicating low available storage or simple monitoring of available storage.

Detailed context: Need to identify available storage capacity for a specified class of service. Input will be the StorageService URI and a class of service name. The use case will look across all storage pools.

Preconditions: None.

Inputs: The URI for the Storage Service that will provision storage is /redfish/v1/StorageServices(1) and the requested class of service name. In this example, GoldBoston.

Basic course of events:

  1. Use GET operation to retrieve data protection capabilities and supported line of service options.
[ {  
         "@odata.context":"/redfish/v1/$metadata#StoragePool.StoragePool",
         "@odata.Id":"/redfish/v1/$metadata#StoragePool.StoragePool/BasePool",
         "Name":"BasePool",
         "Capacity":{  
            "Data":{  
                "ConsumedBytes":0,
                "AllocatedBytes":0,
                "GuaranteedBytes":0,
                "ProvisionedBytes":0
             },
             "Metadata":null,
             "Snapshot":null
          }
      },
      {  
          "@odata.context":"/redfish/v1/$metadata#StoragePool.StoragePool",
          "@odata.Id":"/redfish/v1/$metadata#StoragePool.StoragePool/SpecialPool",
          "Name":"SpecialPool",
          "Capacity":{  
             "Data":{  
                 "ConsumedBytes":549755813888,
                 "AllocatedBytes":1099511627776,
                 "GuaranteedBytes":70368744177664,
                 "ProvisionedBytes":140737488355328
             },
             "Metadata":null,
             "Snapshot":null
      }
   }]
  1. For each returned StoragePool, the requester subtracts ConsumedBytes from ProvisionedBytes for each category , (i.e. Capacity/Data, Capacity/Metadata, and Capacity/Snapshot), and sums the results for each category to get the available capacity for the class of service.

Postconditions: None defined.

Failure Scenario:

Error Code Cause
204 The addressed StorageService is not found.

See also: None defined.


6.16. List supported line of service options


Summary: List supported line of service options

Purpose:

Who: CloudAdmin, StorageAdmin

Management domain: Block storage management, File system storage management, Application storage management, Service catalog management

Trigger: Need to define a new a ClassOfService for use in another request (e.g., Create Volume )

Detailed context: The requester has determined that a new class of service is needed. This use case identifies existing service options for specified lines of service and the capabilities to create new line of service options. For this example, the assumption is that the administrator wants to see all Data Protection capabilities and line of service options.

Note: End users will not typically need to get this level of detail.

Preconditions: None.

Inputs: The URI for the Storage Service that will provision storage is /redfish/v1/StorageServices(1).

Basic Course of Events:

  1. Use GET operation to retrieve data protection capabilities and supported line of service options.
  1. Repeat as needed for other lines of service using the service properties:

Postconditions: None defined.

Failure Scenario:

Error Code Cause
204 The addressed StorageService is not found.

See also: None defined.


6.17. Subscribe to Threshold Events


Summary: Subscribe to Trigger/Clear events for LowSpaceWarningThresholds for a named Volume.

Purpose: Provide an event stream to support utilization management. This is used in conjunction with LowSpaceWarningThresholds to provide a means for on-going monitoring of resource consumption.

Who: CloudAdmin, StorageAdmin, DevOps

Management domain: Block storage management, Application storage management

Triggers: None defined.

Detailed Context: This provides the basis for monitoring capacity consumption.

Preconditions: None defined.

Inputs:

Basic course of events:

  1. Retrieve the event destination

Postconditions: Newly-created event subscription is added to the EventService.

Failure Scenario: None defined.

See also: None defined.


6.18. Set Capacity Usage Threshold


Summary: Set a low space warning threshold on a Volume.

Purpose: Define capacity threshold(s) that will trigger the creation of a LowSpaceWarningThresholdTriggered event and/or LowSpaceWarningThresholdCleared event against a given Volume.

Who: CloudAdmin, StorageAdmin, DevOps

Management domain: Block storage management, Application storage management

Triggers: None defined.

Detailed Context: This provides the basis for monitoring capacity consumption.

Preconditions: Volume exists.

Inputs:

Basic course of events:

  1. Get the LowSpaceWarningThresholdPercents array for the selected Volume.
  1. Response:

json { "@odata.metadata": "/redfish/v1/$metadata#Volume.Volume/v1?$select=LowSpaceWarningThresholdPercents", "value": [ { "LowSpaceWarningThresholdPercents": [0,5,10,35] } ] }

  1. Update the values in the array of thresholds

Postconditions: The new threshold values are set for the identified volume.

Failure Scenario: None defined.

See also: None defined.


6.19. Protect Volumes of an application


Summary: Protect Application Volumes

Purpose: Protect the Volumes of an application represented by a storage group with daily clones and hourly snapshots.

Who: CloudAdmin, StorageAdmin, DevOps

Management Domain: Block storage management

Triggers: Need to protect a set of volumes.

Detailed Context: A storage group collects a set of related storage entities (volumes or file systems). The primary purpose of the collection is to govern access to the members by clients or to add service requirements for the members of the collection. It is expected that the collected storage is used by a set of related client applications.

In this use case, the goal is to add data protection services to all of the storage members, in the form of hourly snapshots and daily clones. This is done by adding a ClassOfService to the storage group or by updating an existing ClassOfService for the storage group. In this use case, the assumption is that the storage group does not already have a class of service, (but this is checked). The added class of service is additive to the class of service requirements of each member storage entity. If the added requirement are in conflict with the requirements of any member, then the request to add the new class of service shall be rejected.

Preconditions: None.

Inputs:

Basic Course of Events:

  1. The first step is to check to see that the desired ClassOfService for the storage group has not been assigned.

  2. Request: GET /redfish/v1/StorageServices(1)/StorageGroups(1)/Links/ClassOfService

  1. Response: 204 No Content
  1. Add a ClassOfService that meets the requirements to create hourly snapshots and daily clones. This is added to the classes of service supported by the storage service. If the storage service is not able to provide the requested service, this request shall be rejected.

  2. Request: POST /redfish/v1/StorageServices(1)/Links/ClassesOfService

  1. Response:
  1. Add the new Class of Service to the storage group. In the process of adding this new class of service the storage service shall check that the new requirements can be met for each member of the storage group. If not, this request shall be rejected.

  2. Request:

PATCH /redfish/v1/StorageServices(1)/StorageGroups(1)/Links

  1. Response:

Postconditions: The updated storage group is now creating hourly snapshots and daily clones for all volumes or filesystems that are members of that group.

Failure Scenario: None defined.

See also: None defined.