ThePub @ NetApp: OpenStack Liberty Summit in Vancouver

OpenStack Liberty Summit in Vancouver – Vote for Session Proposals!

Voting is open now (until February 23rd) for presentation proposals for the OpenStack Liberty Summit in Vancouver. While there are lots of great proposals out there covering a wide variety of topics, we’d like to call your attention to several proposals that involve OpenStack and NetApp, including technology deep-dives on Cinder and Manila, and the use of OpenStack Manila with a variety of use cases. We’d appreciate your vote for these sessions - click on the session title to be taken to the voting page for that session!

Manila: An Update on OpenStack’s Shared File Services Program

Presenters: Bob Callaway, NetApp; Sean Cohen, Red Hat; Mark Sturdevant, HP

Abstract: This presentation will provide an overview of Manila, the OpenStack File Shares Service. Manila is a community-driven project that presents the management of file shares (e.g. NFS, CIFS) as a core service to OpenStack. Manila currently works with a variety of storage platforms, as well as a reference implementation based on a Linux NFS server.

In this session, members of the Manila community will:

Introduce the Manila File Shares Service
Provide an overview of Manila, discussing its architecture through use case examples
Provide details of what is new with the project in the Kilo release
Discuss upcoming areas of focus for Manila towards graduation

Attendees will leave this session with an understanding of Manila, the use cases it was designed to solve, what was added to the project in the Kilo release, and the future directions for the project.

Manila: Taking Open Stack Shared File Storage to the Telco Cloud

Presenters: Marc Koderer, Deutsche Telekom; Thomas Lichtenstein, NetApp; Christian Fey, SVA

Abstract: Manila is the shared file system service and is an incubated OpenStack project since the Juno release. Manila allows tenants to manage shares, e.g., NFS exports, take snapshots, and adjust access rules - among other things.

In this talk, we will present our lessons learned during an evaluation of Manila for use in a telco environment. Deutsche Telekom, NetApp, and SVA have established a joint qualification lab. The solution is using a fully automated open source deployment and testing framework that leverages OpenStack Tempest. Part of this presentation is validation of Manila’s applicability for common use cases, its functionality, as well as the maturity of different backend implementations.

Attendees will leave this talk with a clear understanding of Manila, a better understanding of its suitability for real use, and an outlook on Manila’s roadmap.

Private Cloud POC with RHEL OpenStack Platform on FlexPod

Presenters: Dimitar Ivanov, Telus

Abstract: The talk will share our approach and experience in the process of starting an OpenStack based project for building a private cloud and demonstrate that the barrier to entry into OpenStack based solutions is not high, the key is for the interested organizations to make the right technology selections and form the strong partnerships with their key vendors. The attendees will also be able to learn about the technical aspects of an OpenStack deployment using the RHEL OpenStack Platform on FlexPod converged infrastructure from Cisco and NetApp. This will include the logical and physical architecture and design of the solution and the lessons learned during the implementation process

Use Cases for Manila: Shared File Services in OpenStack Clouds

Presenters: Greg Loughmiller, NetApp

Abstract: Manila is a recently-incubated OpenStack program that adds shared file services as a first-class service along size block (Cinder) and object (Swift). Given that there is no equivalent service in major public clouds like Amazon or Microsoft Azure, we thought it might be useful to give some examples of use cases where Manila can fill a critical gap in bringing certain workloads to OpenStack clouds.

In this session, we’ll give:

A brief introduction to the Manila File Shares Service
Examples of use cases, including:
- Database as a service
- VDI
- Big data (Hadoop) workloads

Attendees will leave this session with an understanding of Manila and tangible use cases where it can be leveraged as a key component of an application deployment.

Ask the Experts: Designing Storage for the Enterprise

Presenters: Greg Knieriemen, Hitachi; Aaron Delp, SolidFire; Neil Levine, Red Hat; Manju Ramanthpura, Hitachi; Alex McDonald, NetApp; Mark Twomey, EMC

Abstract: This all-star panel discussion will focus on the challenges for OpenStack storage and to get direct feedback from users and developers in the audience of their adoption of OpenStack storage. This panel may be recorded as a “Speaking in Tech” podcast which is distributed by Europe’s largest tech publication, The Register.

Topics that will be discussed include:

Enterprise storage adoption observations from the panel
Both Cinder and externally attached storage
Feedback directly from audience members on storage adoption
Discussion of Manila
Ceilometer and Storage metrics

Cinder 102: Pouring a solid foundation for block storage services

Presenters: Jeff Applewhite, NetApp

Abstract: Cinder, at its core, provides basic provisioning and management of block storage volumes to users of OpenStack clouds. However, beyond basic volumes and snapshots, there is a lot of advanced capability for cloud administrators that can be leveraged to provide value added capability to end users, as well as for managing the lifecycle of volumes and storage subsystems.

This talk serves as a “102” level discussion of Cinder and will cover:

Consistency groups
Replicated volumes
Backup/Restore of volumes
Administrative operations
- Volume migration
- Foreign volume import/export
- Resource pool management
- Scheduler optimizations

Attendees will leave with a greater appreciation for the art-of-the-possible with Cinder in the Kilo release, knowledge of the various tools available to Cinder administrators, and practical knowledge around how a Cinder deployment can be optimized for both end-user and cloud administrator outcomes.

Crossing the Chasm: Why Enterprises Are Not Embracing OpenStack

Presenters: Gordon Stitt, Nebula and Jeff O’Neal, NetApp

Abstract: Enterprise IT is constantly challenged to react faster to business demands, which is driving cloud adoption. The default is the public cloud, but there are a number of issues enterprises need to resolve. OpenStack provides hope for enterprise IT to help regain control and deliver infrastructure nimbleness, yet why is there is still reluctance to adopting OpenStack in the Enterprise?

Join Nebula’s CEO Gordon Stitt and NetApp’s Sr. Director of OpenStack Cloud Solutions Jeff O’Neal, as they highlight enterprise IT’s expectation of any technology to seamlessly integrate and operate with their existing enterprise workflows and architecture – these include identity, storage, networking, and monitoring workflows. The session will also review key expectations for enterprise availability and reliability, security, and the challenges of cross-organizational alignment for private clouds.

Making Swift a platform for object lifecycle management

Presenters: Shiva Chaitanya, NetApp

Abstract: Swift has proven to be an extensible platform for object stores and with the introduction of Storage Policies provides a range of performance, reliability and availability. As data in a Swift cluster grows a large fraction of it turns cold (rarely accessed). For such data there more economical layouts e.g., erasure and/or network codes, and media like tape, optical and others like those employed by AWS Glacier. In this presentation we discuss the requirements for cold archival object stores and how external devices can be integrated into Swift. With such an integration Swift becomes a single platform that can manage the entire lifecycle of objects providing a means for customers to implement on-premises lifecycle mgmt. that was previously available only on AWS S3-Glacier. We discuss how middleware and other mechanisms in Swift can be used to provide automated tiering and alternatives on how tiering policies can be declared and scalably implemented.

Using the mechanism employed for lifecycle management we also show how ingested objects can be automatically deposited in different storage classes (sequential devices like tape, random access devices like disks) transparent to the application at ingest time based on operator-specified policies. This complements the existing “Application directed placement of objects” use case of Storage Policies feature in Swift with a new “Operator-driven application-transparent placement of objects into different Storage Classes” use case. We show why the latter use case is necessary for operating object repositories over multiple generations of devices and media.

FlexPod with Red Hat Enterprise Linux OpenStack Platform 6

Presenters: Dave Cain and Eric Railine, NetApp

Abstract: While enterprise customers are considering open source solutions such as OpenStack, many don’t want to give up the level of support or integration they’ve come to depend on. They also don’t want to deploy OpenStack on an infrastructure that is difficult to scale up or automate. So what’s the solution? Deploying Red Hat Enterprise Linux OpenStack Platform 6 on FlexPod provides a fully supported cloud platform that can scale.

Key Takeaways: In this session, you will learn about Red Hat Enterprise Linux OpenStack Platform 6 running on a FlexPod Datacenter solution: a pre-validated, scale-out infrastructure built on Cisco UCS, Cisco Nexus, and NetApp storage with Clustered Data ONTAP 8.3.