Red Hat Storage Day Seattle: Stretching A Gluster Cluster for Resilient Message Persistence and File Storage at McMaster University

Oct 18, 2016Download as PPTX, PDF1 like414 views

One of four Canadian universities ranked in the top 100 worldwide, McMaster University began a project in 2012 to replace its legacy systems with Oracle PeopleSoft for financials, human resources, and other functions. The new Mosaic system runs on Red Hat Enterprise Linux using Red Hat Gluster Storage for shared file storage and messaging where needed. The physical infrastructure is designed for high availability even if one data center room fails, using load balancing, database replication, and clustered resources.

• One of four Canadian Universities ranked among the top
100 in the world (Shanghai Jiaotong University Academic
Ranking of World Universities, August 2016)
• Founded in 1887 in Toronto
• Moved to Hamilton in 1930
• Over 32,000 Full-time Staff, Faculty & Students

Systems Renewal - Mosaic Administrative
Systems
In 2012, McMaster began a multi year ERP project to replace its
Legacy Systems
Oracle PeopleSoft Financials, Human Resources, Campus
Solutions, Data Warehouse and related modules was selected for
Implementation

McMaster and Red Hat
▪Academic “Departmental” license for RHEL (since
converted to full site license)
▪Instance based Gluster, JBoss Fuse, EUS (and
possibly Ceph in our future…)

Mosaic and Red Hat Gluster Storage
▪Deployment of most PeopleSoft components on
RHEL with Gluster for shared/replicated storage
where required
▪Original middleware components RHEL and JBoss
Fuse with Gluster for shared/replicated message
persistence storage

Physical Technical Infrastructure
• Mosaic should continue to function even with the loss of one room (at reduced capacity)
• Message Oriented Middleware (JBoss Fuse with ActiveMQ)
• PeopleSoft file storage (report repository and user attachments)

 User access/presentation layer
• Citrix Netscaler active/passive VPX instances
 Stack components (web/app)
• Largely independent (with help of load balancer/ADC)
 Databases
• Data guard for replication to other site
• Clusterware/RAC for fault tolerance and load distribution
 User application file storage (attachments&reports)
 MOM message persistence
How Close were we?

What about a [supportable] stretch replica cluster?
• All nodes must be on the same logical network – no storage traffic routing
• Latency must be low
• Maximum two storage sites ( not including a possible quorum device at a third neutral site ) **
• Red Hat Architecture Review

**Speaking of quorum devices…
▪Current Installation vulnerable to spit-brain
❑ </knock_wood>this hasn’t been an issue…
▪New version supports a third node in the trusted pool
that doesn’t contribute a storage brick
❑ Need to make sure quorum ratio is set
▪Upstream Gluster supports an arbiter node
❑ File metadata stored with the arbiter, but not file contents
❑ Roadmap for support in RHGS product?

Replicated file storage and MOM message persistence Sandbox
Environment
• Three RHEL7.2 VMs
• gl10.nie.vm (arbiter) ; gl11.nie.vm, gl12.nie.vm storage bricks and ActiveMQ brokers (and test clients)

This document discusses the evolution of storage from traditional appliances to software-defined storage. It notes that many IT decision makers find current storage capabilities inadequate and unable to handle emerging workloads. Traditional appliances face issues like vendor lock-in, lack of flexibility, and high costs. Public cloud storage is more scalable but still has complexity and limitations. The document then introduces software-defined storage as an open solution with standardized platforms that addresses these issues through increased cost efficiency, provisioning speed, and deployment options with less vendor lock-in and skill requirements. It describes Red Hat's portfolio of Ceph and Gluster open source software-defined storage solutions and their target use cases.

Red Hat Storage Day Seattle: Stabilizing Petabyte Ceph Cluster in OpenStack C...Red_Hat_Storage

Cisco uses Ceph for storage in its OpenStack cloud platform. The initial Ceph cluster design used HDDs which caused stability issues as the cluster grew to petabytes in size. Improvements included throttling client IO, upgrading Ceph versions, moving MON metadata to SSDs, and retrofitting journals to NVMe SSDs. These steps stabilized performance and reduced recovery times. Lessons included having clear stability goals and automating testing to prevent technical debt from shortcuts.

Red Hat Storage Day Seattle: Persistent Storage for Containerized ApplicationsRed_Hat_Storage

This document discusses persistent storage solutions for containerized applications. It describes how containers provide benefits like faster development and deployment cycles compared to virtual machines. However, most applications still require persistent storage for data. The document outlines requirements for container storage solutions, such as scalability, resilience, flexibility and being software-defined. It presents Red Hat Storage as a solution, highlighting features like replication, erasure coding and snapshots. Red Hat Storage can provide persistent storage to containers using technologies like Ceph, Amazon EBS, NFS and GlusterFS.

Red Hat Storage Day Seattle: Supermicro Solutions for Red Hat Ceph and Red Ha...Red_Hat_Storage

Red Hat Storage Day Dallas - Gluster Storage in Containerized Application Red_Hat_Storage

The document discusses using Gluster Storage to provide storage for containerized applications in a Kubernetes cluster. It outlines the challenges of replatforming an ecommerce site to use open source technologies, applying RAS(S) principles, and having a scalable and fault-tolerant solution. The plan is to use Docker containers, Kubernetes for orchestration, and GlusterFS storage. GlusterFS provides highly available, replicated storage across all Kubernetes nodes to support the storage needs of containerized applications.

Red Hat Storage Day Boston - Persistent Storage for Containers Red_Hat_Storage

Persistent storage is important for containerized applications. Red Hat provides container-ready storage using Red Hat Gluster Storage which provides scalable, distributed file storage for containers. It allows storage and containers to coexist on the same hardware, improving utilization and lowering costs. Red Hat Gluster Storage is optimized to provide container-native storage on OpenShift for workloads like databases to get the benefits of containers while ensuring persistent storage.

Red Hat Storage Day Seattle: Why Software-Defined Storage MattersRed_Hat_Storage

The document discusses the benefits of software-defined storage over traditional storage approaches. It argues that software-defined storage uses standard hardware and open source software, providing flexibility, scalability, and lower costs compared to proprietary appliances or public cloud storage. It also describes Red Hat's portfolio of software-defined storage solutions, including Ceph and Gluster, which leverage open source technologies to power a variety of enterprise workloads.

Red Hat Storage Day Dallas - Defiance of the Appliance Red_Hat_Storage

The document discusses the challenges with traditional enterprise storage and the benefits of software-defined storage using Red Hat Gluster Storage and Ceph. It highlights how software-defined storage provides near linear performance scaling, lower total cost of ownership, open source innovation, container-native storage, and freedom from vendor lock-in compared to traditional proprietary storage systems.

Red Hat Storage Day New York - Persistent Storage for ContainersRed_Hat_Storage

Red Hat Gluster Storage provides persistent container storage for OpenShift. It has evolved from container-ready (running outside containers) to container-native (running inside containers). The current and upcoming versions provide dynamic storage provisioning without admin intervention, improved usability, and support for database workloads through non-shared storage. A demo shows deploying Gluster Storage containers in OpenShift and creating a new persistent volume claim for an application.

Red Hat Storage Day Boston - Supermicro Super StorageRed_Hat_Storage

The document discusses Supermicro's evolution from server and storage innovation to total solution innovation. It provides examples of their all-flash storage servers and Red Hat Ceph reference architectures using Supermicro hardware. The document also discusses optimizing hardware configurations for different workloads and summarizes Supermicro's portfolio of Ceph-ready nodes and turnkey storage solutions.

Red Hat Storage Day Atlanta - Persistent Storage for Linux Containers Red_Hat_Storage

This document discusses persistent storage options for Linux containers. It notes that while some containerized applications are stateless, most require persistence for storing application and configuration data. It evaluates options like NFS, GlusterFS, Ceph RBD, and block storage, noting that persistent storage needs to be scalable, resilient, flexible, software-defined, and open. It provides examples of using Gluster and Ceph storage with containers. The document concludes that most containerized apps will need persistent storage and that software-defined storage allows for hyperconverged applications and storage on premises or in hybrid clouds.

Red Hat Storage Day LA - Performance and Sizing Software Defined Storage Red_Hat_Storage

This document summarizes a presentation given by Kyle Bader of Red Hat on software defined storage and performance testing of MySQL on Red Hat Ceph Storage compared to AWS EBS. Some key points: - Performance testing showed Red Hat Ceph Storage could provide over 78 IOPS/GB for MySQL workloads, meeting and exceeding the 30 IOPS/GB target of AWS EBS provisioned IOPS. - The price per IOP of Red Hat Ceph Storage on a Supermicro cluster was $0.78, well below the $2.50 target cost of AWS EBS provisioned IOPS storage. - Different hardware configurations, especially core-to-flash ratios, impacted Ceph Storage performance

Red Hat Storage Day Atlanta - Red Hat Gluster Storage vs. Traditional Storage...Red_Hat_Storage

Red Hat Gluster Storage allows organizations to repurpose existing industry-standard servers as storage servers rather than purchasing new hardware from storage vendors. It also allows storage clusters to be grown incrementally and for storage innovations to be delivered independently of hardware upgrades. In contrast, traditional storage appliances limit organizations to the hardware, increments of growth, and feature timelines defined by the vendor. Red Hat Gluster Storage also provides near-linear performance scaling and 3-year total cost of ownership that is hundreds of thousands of dollars lower than traditional storage vendors for a 1PB throughput-optimized configuration.

Red Hat Storage Day Atlanta - Why Software Defined Storage MattersRed_Hat_Storage

This document summarizes an agenda for a Red Hat Storage Day event in Atlanta in August 2016. The agenda includes presentations on software defined storage, Red Hat Ceph Storage on Intel, Red Hat Gluster Storage vs traditional storage appliances, and storage for containerized applications. It also lists a cocktail reception following the presentations. Additional sections provide background on trends driving adoption of software defined storage solutions and an overview of Red Hat's storage portfolio including Ceph and Gluster open source software solutions.

Red Hat Storage Day New York - Intel Unlocking Big Data Infrastructure Effici...Red_Hat_Storage

This document discusses using Ceph storage with Apache Hadoop to provide a scalable and efficient storage solution for big data workloads. It outlines the challenges of scaling Hadoop storage independently from compute resources using the native Hadoop Distributed File System. The solution presented is to use the open source Ceph storage system instead of direct-attached storage. This allows Hadoop compute and storage resources to scale independently and provides a centralized storage platform for all enterprise data workloads. Performance tests showed the Ceph and Hadoop configuration providing up to a 60% improvement in I/O performance when using Intel caching software and SSDs.

Red Hat Storage Day LA - Why Software-Defined Storage Matters and Web-Scale O...Red_Hat_Storage

This document contains an agenda for Red Hat Storage Day being held in Los Angeles in August 2016. The agenda includes presentations and sessions on topics like why software defined storage matters, designing Ceph clusters on Intel hardware, use cases for software defined storage, solutions from SuperMicro, persistent storage for Linux containers, performance and sizing considerations for software defined storage clusters, and web-scale object storage with Ceph. There will also be a Q&A session and cocktail reception.

Red Hat Storage Day New York - Red Hat Gluster Storage: Historical Tick Data ...Red_Hat_Storage

Red Hat Gluster Storage is a software-defined, distributed, scale-out file storage solution that is cost-efficient, high performing at scale, and easy to deploy, manage and scale in public, private and hybrid cloud environments. It offers mature NFS, SMB and HDFS interfaces for enterprise applications such as analytics, media streaming, active archives and enterprise virtualization. The document discusses using Red Hat Gluster Storage for historical tick data repositories, including its architecture, benefits over traditional storage solutions, and analytics workflows.

Red Hat Storage Day Dallas - Storage for OpenShift Containers Red_Hat_Storage

This document discusses using Red Hat Gluster Storage for persistent storage of OpenShift containers. It describes how containers improve software development and management. Containers provide more efficient use of resources than virtual machines. Red Hat Gluster Storage provides scalable, distributed storage optimized for container environments. It can be deployed on-premises or in the cloud and integrated with OpenShift to offer storage as a service for containerized applications.

Red Hat Storage for Mere MortalsRed_Hat_Storage

Red Hat Storage, based on the upstream GlusterFS project, was developed as a distributed file system in the oil-and-gas, high-performance compute arena. With Red Hat Storage, you can easily set up flexible distributed storage using commodity x86 hardware. Simple, inexpensive internal or JBOD storage can be linked across multiple physical servers and presented as a single storage namespace. This storage can be used for log files, web content, virtual machine, home directory, and other storage use cases. In this session, we’ll demonstrate how to: Install Red Hat Gluster Storage Configure disks Link the storage nodes Define storage bricks Present storage to clients We'll talk about tips and tricks, best practices, backup and recovery, and other storage-related topics.

Red Hat Storage Day New York - Penguin Computing Spotlight: Delivering Open S...Red_Hat_Storage

This document discusses Penguin Computing's open solutions utilizing Red Hat Storage. It describes Penguin Computing as providing compute, storage, and networking solutions using open technologies. It then discusses various Penguin Computing solutions like the Tundra Extreme Scale open compute platform, Arc5ca Ethernet switches, and FrostByte HS storage appliances. The document also summarizes Red Hat Gluster Storage benefits for financial data analytics like deeper analysis, lower costs, and better performance compared to traditional storage solutions.

Red Hat Storage Day Boston - OpenStack + Ceph StorageRed_Hat_Storage

- Red Hat OpenStack Platform delivers an integrated and production-ready OpenStack cloud platform that combines Red Hat's hardened OpenStack infrastructure which is co-engineered with Red Hat Enterprise Linux. - Ceph is an open-source, massively scalable software-defined storage that provides a single, efficient, and unified storage platform on clustered commodity hardware. Ceph is flexible and can provide block, object, and file-level storage for OpenStack. - Architectures using OpenStack and Ceph include hyperconverged infrastructure which co-locates compute and storage on the same machines, and multi-site configurations with replicated Ceph storage across sites for disaster recovery.

Why Software-Defined Storage MattersColleen Corrice

Red Hat Storage Day Boston - Red Hat Gluster Storage vs. Traditional Storage ...Red_Hat_Storage

Red Hat Gluster Storage provides a software-defined storage solution that is more cost efficient and flexible than traditional storage appliances. It leverages standard x86 hardware and has open source architecture with no vendor lock-in. A comparison shows Gluster Storage outperforms EMC Isilon on factors like cost, scalability, data protection methods, access protocols, and management capabilities. Gluster Storage is positioned to go beyond traditional storage by supporting containers, disaster recovery in cloud environments, and its roadmap includes additional advanced features.

Implementation of Dense Storage Utilizing HDDs with SSDs and PCIe Flash Acc...Red_Hat_Storage

Red Hat Storage Day New York - QCT: Avoid the mess, deploy with a validated s...Red_Hat_Storage

This document provides an overview of QCT's validated Red Hat Ceph and Gluster storage solutions. QCT offers pre-configured and optimized storage appliances built with Red Hat Ceph and Gluster storage software. Their QxStor solutions include different configurations optimized for throughput, capacity, or IOPS. The document discusses QCT's testing results showing the performance and scalability of their Ceph and Gluster solutions. It also describes QCT's center of excellence where they collaborate with partners to test and develop new solutions.

Red hat Storage Day LA - Designing Ceph Clusters Using Intel-Based HardwareRed_Hat_Storage

This document discusses how data growth driven by mobile, social media, IoT, and big data/cloud is requiring a fundamental shift in storage cost structures from scale-up to scale-out architectures. It provides an overview of key storage technologies and workloads driving public cloud storage, and how Ceph can help deliver on the promise of the cloud by providing next generation storage architectures with flash to enable new capabilities in small footprints. It also illustrates the wide performance range Ceph can provide for different workloads and hardware configurations.

Why Software-Defined Storage MattersRed_Hat_Storage

Architecting Ceph SolutionsRed_Hat_Storage

Azure fb-google Web ServicesShreya Srivastava

Microsoft Azure is a cloud computing service that provides infrastructure, platform and software services through global data centers. It supports virtual machines, web apps, storage, databases, analytics and more. Azure uses a specialized operating system called Microsoft Azure to manage computing resources across its global fabric layer. Facebook's data center fabric provides scalable networking infrastructure to support increasing traffic and new products. It uses ECMP routing and multi-speed links for load balancing. The fabric is designed as a non-oversubscribed environment and uses automation tools to manage topology changes. Google's first data centers used donated hardware from Sun, Intel and IBM. It has numerous centers worldwide with large facilities in the US, Europe and Asia. Google developed software for

Waters Grid & HPC Coursejimliddle

More Related Content

What's hot (20)