Protecting Mission Critical Bus App
0 likes341 views
This document discusses how the software solution Neverfail Continuous Availability can be used along with server virtualization to provide high availability and disaster recovery for mission-critical multi-tier applications. It outlines common deployment scenarios organizations face with virtualized applications and how Neverfail allows protection of applications spanning physical and virtual servers across local and remote sites. Key benefits of Neverfail include protection from various failure types, flexibility to use physical or virtual secondary servers, and embedded replication for disaster recovery without requiring other solutions.
Protecting Mission Critical Bus App
- 1. Eliminating End User and Application Downtime Protecting Mission-Critical Multi-Tier Business Applications with Neverfail and Virtualization
- 2. The rapid adoption of server virtualization has enabled many organizations to significantly reduce their IT capital expense, operating costs as well as adapt to changing business requirements more easily and improve overall server availability. However, when it comes to extending the benefits of server virtualization to their missioncritical business applications, many organizations often struggle with how to re-architect their environment and strike the right balance between performance, scalability, manageability and most importantly, high availability and disaster recovery. This white paper outlines many of the common deployment scenarios that such organizations face and describes how Neverfail Continuous Availability and virtualization can be used to deliver a more complete, consistent and cost effective availability solution for mission-critical, business applications. Protecting Mission-Critical Applications in a Mixed Physical,Virtual World Most modern business applications incorporate multiple tiers of interdependent servers (Web Front End, Application, Database/ Storage etc.) each of which have differing roles and therefore different performance, management and availability requirements. However, to ensure the continuous availability of the entire business application, each of these systems must be managed and protected in a consistent manner in order to minimize the risk of business application downtime. Most modern business applications incorporate multiple tiers of interdependent servers (Web Front End, Application, Database/Storage etc.) each of which have differing roles and therefore different performance, management and availability requirements. Introduction Web Tier Exchange Messaging System Virtualized Application Tier Database Tier Exchange CAS MBX Server SQL Server File Server VM VMVM Exchange CAS Exchange OWA Anti-Spam Server BES Server Anti-Virus Server Exchange Hub VM VM VM VM Physical Web Front End SharePoint Farm Virtualized Application Tier Database Tier IIS+MOSS SQL Server SQL Server VM VMVM IIS+MOSS IIS+MOSS Central Admin Query Server Exec Calc Server Index Server VM VM VM VM Physical Web Front End Exchange Messaging System Physical Application Tier Database Tier Exchange CAS Exchange OWA Exchange Hub Blackberry Server Anti-Virus Server Exchange MBX Exchange MBX SQL Server Web Front End SharePoint Farm Virtualized Application Tier Database Tier SQL Server SQL Server Central Admin IIS+MOSS IIS+MOSS IIS+MOSS Query Server Excel Calc Server Index Server When organizations look to virtualize their mission-critical business applications, careful consideration must be given to ensure that nothing impacts performance, scalability and manageability and that nothing compromises the continuous availability of the entire business application. Such concerns have led many organizations to part-virtualize their server farms, often leaving the Database tier completely untouched as well as many other high performance application servers. Web Tier Exchange Messaging System Virtualized Application Tier Database Tier Exchange CAS MBX Server SQL Server File Server VM VMVM Exchange CAS Exchange OWA Anti-Spam Server BES Server Anti-Virus Server Exchange Hub VM VM VM VM Physical Web Front End SharePoint Farm Virtualized Application Tier Database Tier IIS+MOSS SQL Server SQL Server VM VMVM IIS+MOSS IIS+MOSS Central Admin Query Server Exec Calc Server Index Server VM VM VM VM Physical Web Front End Exchange Messaging System Physical Application Tier Database Tier Exchange CAS Exchange OWA Exchange Hub Blackberry Server Anti-Virus Server Exchange MBX Exchange MBX SQL Server Web Front End SharePoint Farm Virtualized Application Tier Database Tier SQL Server SQL Server Central Admin IIS+MOSS IIS+MOSS IIS+MOSS Query Server Excel Calc Server Index Server Figure 1: Exchange & SharePoint: Mission-critical, multi-tier business applications Figure 2: Typical deployment scenarios: Part-virtualized, multi-tier, business applications
- 3. Neverfail Continuous Availability for Mixed Physical, Virtual Environments The VMware vSphere and Microsoft Hyper-V platforms incorporate many platform-level technologies (VMware HA, FT, VMotion, Storage vMotion, SRM, Live/Quick Migration etc.) that can significantly improve overall system resilience and server availability. However, when it comes to protecting your mission-critical business applications, in a mixed physical, virtual environment, across both local and remote sites (for disaster recovery) these platform-level technologies do not provide complete protection for your mission-critical business applications. In addition, unlike Neverfail, these platform-level availability technologies do not protect your mission-critical business applications from the most common causes of downtime, such as application failure, service failure, performance degradation, configuration drift, user error etc. As illustrated in the above graphic, taking a“Silo-based”approach to availability and disaster recovery creates significant additional complexity, as it requires that each solution to be installed, configured, managed and maintained in different ways. Unfortunately, this complexity will invariably result in an increased risk of downtime for the mission-critical business application that you are trying to protect! Load Balancing Business Service Web Server Web Server ????? Application Server Application Server MS Clustering Database Mirroring Database Server Database Server Business Service Web Server Web Server Application Server Application Server Database Server Database Server Neverfail Approach to AvailabilitySilo-based Approach to Availability CompleteProtectionforBusinessService Figure 3: Conventional, silo-based approach to availability vs. Neverfail Continuous Availability When organizations look to virtualize their mission-critical business applications, careful consideration must be given to ensure that nothing impacts performance, scalability and manageability and that nothing compromises the continuous availability of the entire business application. Such concerns have led many organizations to part-virtualize their server farms, often leaving the Database tier completely untouched as well as many other high performance application servers.
- 4. As can be seen from the above table, when it comes to providing complete protection against downtime for mission-critical business applications in a mixed physical, virtual environment, Neverfail’s Continuous Availability Suite is the only solution that covers all of the common high availability and disaster recovery deployment scenarios. Neverfail Mixed Physical, Virtual Deployment Scenarios Neverfail provides the flexibility and freedom of choice to implement both local and remote site high availability solutions that meet your technical and business requirements, whatever the scenario. If local high availability is required and the primary production servers are physical machines, Neverfail gives you the option to choose physical or virtual servers as the secondary server for the high availability pair. Availability / Protection Microsoft Hyper-V Microsoft DPM VMware HA/FT VMware SRM Neverfail v6.2 Local High Availability √ X √ X √ Remote Site Protection X √ X √ √ Local & Remote Site Availability X X X X √ Virtual Machine Level Protection X √ X √ √ Physical & Virtual Server Protection X √ X X √ Storage System Failure Protection X √ X √ √ Network Failure Protection X X X √ √ Operating System Failure Protection X X Basic X √ Application Aware Failure Protection X X X X √ Configuration Drift Protection X √ X √ √ Automatic &Manual Failover/Failback X X X X √ Embedded Replication (for DR) X √ X X √ Embedded WAN Optimization X √ X X √ Figure 4: Neverfail Continuous Availability vs. VMware and Microsoft Hyper-V availability features Neverfail provides the flexibility and freedom of choice to implement both local and remote site high availability solutions that meet your technical and business requirements, whatever the scenario. If local high availability is required and the primary production servers are physical machines, Neverfail gives you the option to choose physical or virtual servers as the secondary server for the high availability pair.
- 5. Neverfail All Virtual Deployment Scenarios Where virtualization is used to implement a local virtual cluster for server high availability, using Neverfail to replicate and maintain mission critical applications at a remote location, can bring immense benefits in terms of disaster availability. It is likely that only a few mission critical applications require Continuous Availability. By focusing on these applications, the business can be protected without requiring complex infrastructure upgrades such as stretched or replicated SANs which would be more suited to full site protection. The Virtualization of secondary servers (as detailed above) is a popular choice for local high availability, for the following reasons: • There is no application disruption as the primary applications remains untouched • The consolidation of the secondary HA server (via virtualization) reduces the total cost of HA • The secondary servers facilitate the testing of virtualized mission-critical business applications • The Neverfail vXtender licensing lowers the cost of protecting the virtual secondary servers Production Site Primary Physical Server Neverfail Channel Secondary Physical Server LAN Zero downtime failover Production Site Primary Physical Server Neverfail Channel Secondary Virtual Server LAN Zero downtime failover VM Figure 5: Neverfail protecting a physical server with a secondary physical server (P2P local HA) Figure 6: Neverfail protecting multiple servers in a mixed physical, virtual environment (P2V HA) Where virtualization is used to implement a local virtual cluster for server high availability, using Neverfail to replicate and maintain mission critical applications at a remote location, can bring immense benefits in terms of disaster availability.
- 6. Neverfail Local and Remote Site Availability Deployment Scenarios With Neverfail, once a local high availability pair has been established, adding a remote site availability or disaster recover site becomes simple. It’s just a matter of implementing the appropriate environment at the remote DR site, using Neverfail Continuous Availability to synchronize the data and application configuration. Production Site Primary Virtual Server Neverfail Channel Secondary Virtual Server LAN Zero downtime failover VM Primary Site Physical Primary Server Neverfail Channel Virtual Secondary Server VM LAN Recovery Site Tertiary Server VM LAN WAN Zero downtime failover Neverfail Channel Zero downtime failover Neverfail Physical to Virtual Migration Scenarios Neverfail also reduces the effort associated with migrating from a physical server environment to a virtual server environment as a by-product of the local high availability and replication capabilities, inherent within the Neverfail product. After installing the secondary virtual servers and synchronizing the data, switching over to the secondary machines and making them the production servers is as easy as pressing a button. This process can be performed in the local environment, the remote environment or both. Once it is determined that there is no application performance degradation, the physical machine can be shut down and removed altogether, or re-purposed as needed. Figure 7: Neverfail protecting multiple servers in a purely virtual environment (V2V HA) Figure 8: Protecting mixed physical, virtual servers with local & remote site availability (P2V2V) With Neverfail, once a local high availability pair has been established, adding a remote site availability or disaster recover site becomes simple. It’s just a matter of implementing the appropriate environment at the remote DR site, using Neverfail Continuous Availability to synchronize the data and application configuration.
- 7. The Importance of Fast, User Transparent, Automatic Failover & Failback In order to meet the aggressive Recovery Time Objectives (RTO) that mission-critical business applications require, Neverfail pro- vides fast, user transparent, automatic failover and failback for the entire business application. Neverfail’s ability to perform fast, user transparent, automatic failover of mission-critical business applications, in mixed physical and virtual environments, is what sets the Neverfail solution apart from other Data Protection and Recovery solutions such as VMware Site Recovery Manager (SRM) & Microsoft Data Protection Manager (DPM). For example, The VMware SRM product is primarily designed to automate the site disaster recovery process (“Run book automation”) for up to 1,000 virtual machines, a job which it does extremely well. However, when it comes to providing Continuous Availability for mission-critical business applications when a disaster strikes, the process of bringing the business application back up and running on the remote site could take anywhere between 20 minutes to several hours. By comparison, with Neverfail’s Continuous Availability solution the same mission-critical business application (taking the same environment and configuration) would be up and running again in a matter of seconds or at most, a few minutes. The Importance of Heterogeneous Replication Neverfail Continuous Availability incorporates heterogeneous replication as standard. Unlike VMware SRM, which requires SAN based replication in order to provide site protection, Neverfail’s replication technology can work with Direct Attach Storage (DAS) or Storage Area Network (SAN) attached storage and enables any system to be replicated to any other system, regardless of the make or model of the underlying storage. This“open”approach to system replication provides organizations with the flexibility and freedom of choice to configure system replication in the most cost effective manner (e.g. high end EMC to low end NetApp) which can deliver tremendous cost savings. The Importance of WAN Optimization for Remote Site Availability and DR When it comes to designing a remote site availability or disaster recovery solution, the number of options available will often be constrained by network bandwidth and/or the physical distance between sites, servers or virtualization hosts. After all, not every- one can afford the high costs associated with deploying a low latency network (required for SAN based synchronous replication) or justify moving large amounts of traffic over wide area network (WAN) links. Neverfail Continuous Availability incorporates WAN optimization which makes implementing a remote site availability or disaster recovery solution for business applications affordable and easy to implement by overcoming the common issues of replicating over long distances and/or limited bandwidth. Figure 9: Server failover, showing the failover progress, replication state and actual Recovery Point In order to meet the aggressive Recovery Time Objectives (RTO) that mission-critical business applications require, Neverfail provides fast, user transparent, automatic failover and failback for the entire business application.
- 8. WANSmart replication combines the proven benefits of data compression with Neverfail’s advanced de-duplication technology. When combined, these two technologies accelerate the replication of business application(s) across the WAN, significantly reduc- ing the bandwidth requirements (by a factor of up to 30) as well as enabling significantly better Recovery Time (RTO) and Recovery Point Objectives (RPO) to be achieved. Web Front End Virtualized Application Tier Database Tier SQL Server SQL Server VM Web Front End SharePoint Farm (Passive) Virtualized Application Tier Database Tier IIS+MOSSIIS+MOSS IIS+MOSSIIS+MOSS VM VM VM VM Central Admin Query Server Exec Calc Server Index Server VM VM VM VM SQL Server SQL Server VM VM Web Front End SharePoint Farm (Passive) Virtualized Application Tier Database Tier IIS+MOSSIIS+MOSS IIS+MOSSIIS+MOSS VM VM VM VM Central Admin Query Server Exec Calc Server Index Server VM VM VM VM SQL Server SQL Server VM VM WANSmart TM Cloud VM SharePoint Farm (Active) IIS+MOSSIIS+MOSS IIS+MOSSIIS+MOSS VM VM VM VM Central Admin Query Server Exec Calc Server Index Server VM VM VM VM VM Neverfail Continuous Availability Director Application Modules Scope 3rd Party Adapters AMF Monitoring Data Continuity WANSmart Figure 10: Using Neverfail WAN optimization (WANSmart) to improve RTO & RPO 1 5432 Data Changes Intercepted Compression & De-duplication Transmission Across WAN Decompression & Reconstitution Data Changes Applied ACTIVE SERVER PASSIVE SERVER WAN Duplicates Found VM Restore Duplicates To reduce RTO’s further, WANSmart also performs the necessary DNS updates automatically, which ensures that the mission- critical business application WAN failover is as seamless as possible and that the highest level of continuous availability is achieved for the business users. Figure 11: WANSmart de-duplication and compression reduce bandwidth requirements by up to 30x WANSmart replication combines the proven benefits of data compression with Neverfail’s advanced de-duplication technology. When combined, these two technologies accelerate the replication of business application(s) across the WAN, significantly reducing the bandwidth requirements (by a factor of up to 30) as well as enabling significantly better Recovery Time (RTO) and Recovery Point Objectives (RPO) to be achieved.
- 9. Neverfail Continuous Availability - Built for Mission Critical Applications Neverfail’s Continuous Availability Suite is unique in that it proactively manages and protects the entire business application ecosystem whether built upon physical servers, virtual servers or a combination of both. It can protect mission-critical business applications against any type of failure, such as the physical server hardware, storage, network, operating system, service, application, configuration drift, human error etc. as well as a total disaster, such as a complete site failure. If a problem occurs, Neverfail ensures your applications are protected. Cloud VM Neverfail Continuous Availability Director Application Modules Scope 3rd Party Adapters Application-Aware ManagementFramework(AMF) Monitoring Data Continuity WANSmart Figure 12: The Neverfail Continuous Availability Suite Application-Aware, Centralized Management and Control The Continuous Availability Director management console provides organizations with an enterprise-wide, business-centric view of mission-critical business applications and IT services. The view spans both physical and virtual environments, as well as providing a consolidated view of local and remote systems. It has a flexible approach, which allows logical grouping of application, database, messaging and other servers. Groupings provide a way of visualizing interdependencies across servers to provide a central console where important events, alerts and overall health of the business application can be viewed. Web Front End Virtualized Application Tier Database Tier SQL Server SQL Server VM Cloud VM SharePoint Farm IIS+MOSSIIS+MOSS IIS+MOSSIIS+MOSS VM VM VM VM Central Admin Query Server Exec Calc Server Index Server VM VM VM VM VM Neverfail Continuous Availability Director Application Modules Scope 3rd Party Adapters AMF Monitoring Data Continuity WANSmart Web Front End Web Applications Virtualized Application Tier Database Tier IIS+MOSS SQL Server SQL Server SQL Server VMVM IIS+MOSS IIS+MOSS File ServerCommerce CustomBiztalk Physical VM IIS+MOSS Web Front End Virtualized Application Tier Database Tier SQL Server SQL Server Messaging Systems Symantec A/V Blackberry Server Domino Server Exchange Server Outlook Web Access VM VM VM VM VM Figure 13: Continuous Availability Director: Protecting multiple, mission-critical business applications The Continuous Availability Director management console provides organizations with an enterprise-wide, business-centric view of mission-critical business applications and IT services.