RED HAT ENTERPRISE LINUX OPENSTACK PLATFORM Stanislav Polasek ELOS Technologies stanislav.polasek@elostech.cz
ELOS Technologies Automatizace infrastruktury datových center Centrální správa konfigurací LifeCycle Automation Migrace do veřejných cloudů Hybridní cloud modely Partneři Red Hat Premium Partner (RHCA/RHCE) PuppetLabs Solution Provider (Puppet Architect/Developer) AWS Consulting Partner (AWS Solution Architect)
Knowledge Factory Edukační centrum pro nové technologie Automatizace infrastruktury Docker OpenStack PuppetLabs www.knowledgefactory.com
What is OpenStack?
Cloud Infrastructure for Cloud Workloads Modular architecture Designed to easily scale out Based on (growing) set of core services
Capabilities OpenStack meets the needs of new scale-out applications Brings public cloud-like capabilities into your datacenter Provides massive on-demand (scale-out) capacity 1,000's 10,000's 100k's of VMs Removes vendor lock-in Community development = higher feature velocity Open source provides high-degree of flexibility to customize and interoperate Features and functions you need, faster to market over proprietary software Greater automation, resource provisioning, and scaling
Workload Type TRADITIONAL: SCALE UP CLOUD: SCALE OUT Big stateful VM Small stateless VMs 1 Application 1 VM 1 Application Many VMs Lifecycle in years Lifecycle hours to months Scale up (VM gets bigger) Scale out (add VMs) MIXED/HYBRID (OpenStack) Not designed to tolerate failure of If a VM dies, application kills it VM, so you need features that and creates a new one, app stays keep VMs up up Application SLA requires Application SLA requires enterprise virtualization features adding/removing VM instances to (migration, HA, etc.) to keep application cloud to maintain applications available application availability Combination of traditional scaleup and cloud scale-out workloads. For example: Database may be hosted on traditional workloads, web front-end and logic layers on cloud workloads.
Free software Free Software is gratis, right? We ll save on licensing costs!
Real good reasons for free software o Adaptability - you can adapt or complete the software o Security - you can control what s in it o Sharing - so you can split costs with others o Standard - so you can be sure it interoperates o Hybridation - your providers work the same way o Etc...
Isn't free software a good enough reason? Free software still has costs o Maintenance (subscription or internalized) o Expertise (experts are not cheap) o Knowledge (your teams have to learn new skills)
Why should my enterprise pick OpenStack?
OpenStack is an agility enabler
Agility Enabler Enable business units to be more reactive in a faster market Enable development teams to be more productive and autonomous Enable (cloud ready) applications to be more scalable Enable more creativity Enable finer grain cost analysis and responsibility identification Be quicker, better, smarter to success!
So, I just have to install OpenStack and I am done?
OpenStack is just a technology, a tool to enable the transformation OpenStack is an agility enabler For your business units For your development teams For your business departments (B to C) But only if you change the ways they interact with the data center
Transforming the interactions: BUs Business units before Business units transformed Tell IT what they need Iterate on their needs Wait for approval Own the application they use Wait for IT to complete the need Quickly launch MVPs to test Complain when it fails Operate their applications Rely on IT for application SLA Rely on IT for data center SLA Are restrained by global policies that should not apply to them Can adapt policies to match their market needs
Transforming the interactions: Developers Developers before Developers transformed - Devops Receive lengthy specification and write code to match their understanding of the problem Closely interact with stakeholders Fight for hardware allocation Are responsible for hw allocation Transfer operational burden as quickly as possible Same team operate the applications they developed Are not responsible of SLA Build SLA into the application Rely on global data store to be globally available Build hybridation into the application Iterate quickly to validate/adapt
Transforming the interactions : B to C. Business department before Business department transformed Tell IT what they need Allocate their own hardware Subcontract wildly to agencies Provide access to agencies Complain when it fails Are responsible of their operations Don t understand delays Fight with their own delays Always buy advertising space and create unrealistic hurdles Find someone else to hurry for their crazy plans Get tired of waiting and go AWS Are finally traceable in their spend
Cloud infrastructures are just one more abstraction enabling the handling of massive number of nodes as if they were a simple entity (aka pet vs. cattle)
To benefit from the abstraction you have to transform your tools Central DBs Vertical Scaling Filesystems FIFO Specialized Unique Sequential Distributed DB (nosql) Horizontal Scaling Blocks and Objects Bus Commoditized Pattern based Parallelized OpenStack is made for the abstraction to happen
It is not meant to handle your old workloads OpenStack is made for the abstraction to happen Migrations still have a cost Without adaptation it does not bring any benefits You need to accelerate, not just create more work Lean enablement cannot be done with bloated software
Start using your cloud with news apps or components Convert your apps calmly, within their normal life cycles You can mix and match computing models to proceed in small shorter steps Not all need to be cloud at once then you should start seeing the real benefits of OpenStack
Make hybrid apps Enable front-ends to scale on multiple site Enable data to be maintained on multiple locations which in turn will provide the benefits of real hybrid deployments that OpenStack enables
Change the way your IT teams work: Each group should be responsible of delivering its resources globally Stop dividing those who operate from those who design Stop defining silos in terms of responsibility but in term of APIs and measurable objectives to maintain Stop the pre-control craziness trust is the first enabler of delegation which leads to responsibility, autonomy and creativity Measure end goals, not milestones this is the only thing that matters
Why OpenStack now? TTM is key in all highly competitive sectors Further reduction of TTM cannot be obtained without Agility/DevOps model Influencing Open Source environment is critical to get your vertical s point of view known Transformation takes time, better start early This is the best way toward hybridation
Why Red Hat?
OpenStack: Framework for the Cloud Needs to access x86 hardware resources Needs an operating environment, hypervisor, services Leverages existing code libraries for functionality
The Importance of Integration with Linux Red Hat Supported Guests A typical OpenStack cloud is made up of at least 9 core services + plugins to interact with 3rd party systems OpenStack KVM RHEL Hardware These services run on top of a Linux distribution with a complex set of user space integration dependencies OpenStack cannot be productized as a stand alone layer A supported, stable platform requires integration and testing of each of the components If your Windows virtual machine hosted by a KVM hypervisor running on an IBM blade, connecting to an EMC storage array through an Emulex HBA has issues with storage corruption, who do you call?
Red Hat Enterprise Linux Optimized Enablers for OpenStack Red Hat Supported Guests OpenStack KVM RHEL Hardware Virtualization guest performance, reliability and Windows Security - SELinux enforcing guest isolation Network SDN/OVS performance optimized Storage vendor plugins, performance, thin provisioning Ecosystem certification of hardware, storage and networks Linux Kernel Security Enhanced Linux (SELinux) KVM Device Drivers Network Stack
RHEL OpenStack Platform 7
RHEL OpenStack Platform 7 Hypervisor Support Red Hat Enterprise Virtualization Hypervisor *Red Hat Enterprise Linux KVM VMware vsphere *vcenter Driver Lightweight / small footprint Less overhead Smaller attack surface Cost effective Closer to operating system DNA Provides massive scale-out capabilities Maximum benefit with virtualized Linux Co-exist with existing infrastructure assets Provides a seamless path to future migration to OpenStack Uses NSX1 plugin for Neutron 1 NSX is only supported in production environments, per VMware's support requirements *ESXi driver not supported
RHEL OpenStack Platform 7 Virtual Guest Support Red Hat Enterprise Linux 3 Red Hat Enterprise Linux 4 Red Hat Enterprise Linux 5 Red Hat Enterprise Linux 6 Red Hat Enterprise Linux 7 Red Hat Enterprise Linux Atomic Host *32 and 64 bit for all versions of RHEL Windows XP SP3+1 Windows 73 Windows 83 Microsoft SVVP Certified Windows Server 2003 SP2+3 Windows Server 20083 Windows Server 2008 R22 Windows Server 20122 SUSE Linux Enterprise Server 10 SUSE Linux Enterprise Server 11 *32 and 64 bit for all versions 1 32 bit only 64 bit only 3 32 and 64 bit 2
Largest OpenStack Partner Ecosystem Over 350+ members since launch in April 2013 Over 900 certified solutions in partner Marketplace Over 4,000 RHEL certified compute servers Over 13,000 applications available on RHEL OEMs, IHVs, and ISVs System Integrators Channel Partners Cloud Service Providers Managed Service Providers
RHEL OpenStack Platform Director Intuitive graphical installer, driven by an API backend Ensures a production-ready environment with Automated Health Checks (AHC) during and after installation Enables high availability (HA) across controller and compute nodes (including networking in active-active ) 1 Automatically Utilizes Fencing as containment mechanism Includes Red Hat Ceph Storage client and server deployment 1 with integrated director configuration support for storage backends Optional partner integration/configuration support NetApp Data ONTAP (incl. 7-mode) Cisco Nexus 1000v Red Hat Ceph Storage entitlement and subscription sold separately
Director: Deployment Consistency Best practices & reference architectures Automation and repeatability RHEL OpenStack platform director was created based on expertise from the field Lessons learned from previous deployment tools Reference architectures with certified hardware make deployment easier Hardware performance and validation testing AHC (Automated Health Checks) Hardware detection Performance information Black sheep detection Automated tests during and after deployment Find and troubleshoot problems faster
Director: Lifecycle Management Deployment is just the first step, most of the action happens afterwards Add and remove capacity Deploy critical updates Upgrade to new OpenStack versions Automation is a must API first, used by both the CLI and GUI and allow for better integration with external tools Automated hardware detection and performance tests Automated functional tests to validate the deployment as early as possible Orchestrated upgrades, easier to keep closer to the newest features
Director: Operational Visibility Important for troubleshooting and system status Ensure correct node behavior Tracking resource capacity Monitoring and alerts for operational failures Operational tools Log aggregation and search Core service and infrastructure availability monitoring Performance monitoring
Community Contributions...
Red Hat Community Contributions Top Contributor to Kilo Release Overall commits per company (aggregated) Red Hat total community contributions (projects) 4000 20.09% Red Hat 3500 HP 3000 IBM 2500 Mirantis 2000 Rackspace 1500 1000 500 0 Yahoo! OpenStack Foundation NEC 10.37% 18.78% 82.91% 11.31% 22.46% 24.23% 29.99% 4.36% Cisco Nova Sahara TripleO-heat-temps Source: http://activity.openstack.org/dash/browser/scm-companies.html?release=kilo http://stackalytics.com/?release=kilo&company=red%20hat 25.62% 8.82% Horizon Heat Cinder Swift Neutron Glance Ceilometer Ironic
Red Hat's OpenStack Leadership Why these statistics matter? With Red Hat's near 20 year history in open source, we have the experience and resources to: Support production-ready customers globally Drive new features Influence strategy and direction of project Enable partner collaboration Wide ranging participation in OpenStack projects, contrasts with most vendors who are more narrowly focused All of these efforts allows us to create a production-ready distribution with ecosystem, enterprise lifecycle, and support that customers expect from Red Hat
Integrated Cloud Solutions...
Red Hat Cloud Infrastructure Delivering an open private cloud
Open Hybrid Cloud CloudForms adds heterogeneous capacity
Red Hat Cloud Suite for Applications Integrated DevOps Platform for the enterprise IaaS PaaS Management