Pivot3 Storage and Compute STAC. Technology Overview

Transcription

1 Pivot3 Storage and Compute STAC Technology Overview April 2011

2 Pivot3 STAC Technology Overview Table of Contents Introduction... 3 Pivot3 STAC Architecture Virtual Server Environments Pivot3 RAIGE Operating System Virtual Server Software Pivot3 Appliances Pivot3 vbank Appliances Pivot3 CloudBank Appliances Pivot3 DataBank Appliances Summary , Pivot3, Inc., All rights reserved. 2

3 Introduction Industry transitions are often created when elegant software makes it practical to use IA 64 based commodity hardware for tasks that previously relied on expensive proprietary hardware. These inflection points drive widespread deployment of new technology into more cost sensitive markets. Examples of companies that used software to drive technology transitions include Microsoft, VMware and Google. Microsoft, for example, introduced IA 64 servers to business critical computing once the Windows NT operating system became a viable application platform. This effectively ended the era of mini computers. VMware similarly shook up the mainframe business by developing virtualization software for IA 64 servers. Most recently, Google and Amazon have developed software to support cloud hosting of server and storage resources using tens of thousands of inexpensive IA 64 motherboards. In each case, the disruptive software had to anticipate that commodity hardware components would fail and that the performance of general purpose hardware would be more difficult to manage than application specific hardware. Offsetting these challenges was the disruptive cost base of the IA64 platforms and the relentless performance improvements offered up by Moore s Law. The Pivot3 Storage and Compute STAC platform is the latest example where new software for IA 64 servers creates an inflection point for users. The platform consolidates server and storage hardware using a single collapsed software stack to deliver a highly available on demand infrastructure for capacity intensive or compute intensive applications. This document provides a technical overview of the Pivot3 STAC platform and describes the innovations inherent in the design , Pivot3, Inc., All rights reserved. 3

4 Pivot3 STAC Architecture The Pivot3 STAC architecture combines high performance compute and storage resources in scale out appliances that are stacked together. It is the first and only scale out architecture that deploys a storage area network (SAN) first and then layers in local virtual servers. The combined server/storage platform is both a high availability SAN storage solution with no single point of failure and a high performance, high availability server solution that takes advantage of commodity hardware components. There are four major components in Pivot3 STAC are as follows: Virtual Server Environments Pivot3 RAIGE OS Virtual Server Softwware Pivot3 Appliance Hardware The following chapters describe each of these four architectural components in more detail , Pivot3, Inc., All rights reserved. 4

5 1. Virtual Server Environments The Pivot3 STAC simultaneously provides virtualized servers and scale out storage resources in a consolidated platform that is ideal for capacity intensive environments. There is no application integration required and no certification hurdles for new applications. Remote management and monitoring of virtual server applications and operating systems is supported as with any remote server system. VMware and Xen Virtualization Platforms Pivot3 scale out storage platforms that can support many, one or no virtual server environments based on the virtualization and compute resources available in the STAC appliances. vbank : Scale out SSD/disk storage appliance with multiple ESXi virtual machines (VMs) CloudBank : Scale out disk storage appliance with a single Xen VM DataBank : Scale out disk storage appliance vbanks provide VMware environments where multiple virtual servers can be managed through VMware vcenter. Failover and live migration of virtual servers are controlled by VMware s High Availability (HA) and vmotion features. CloudBanks provide Xen environments where a single virtual server is managed by Pivot3. Failover for virtual servers is controlled by the Pivot3 RAIGE OS. DataBanks provide scale out storage that is accessed by external servers. DataBanks can also be added to Cloud Bank STACs where less compute resources are required in a STAC. ISV Ecosystem Pivot3 maintains a strong ISV ecosystem around the STAC architecture. While the STAC architecture is a general purpose platform, several application solution areas have been targeted for deep application integration and support: data protection, digital surveillance, and rich media. For each solution area listed below, the Pivot3 Application Lab ensures interoperability and characterizes performance of independent software vendor (ISV) applications. Data Protection The Pivot3 STAC provides excellent storage performance for data protection applications. Part of, or all of, the Pivot3 platform can be used as a backup, recovery, archive or deduplication target by running ISV applications on virtual servers in the Pivot3 STAC. Virtual servers can also host replication or snapshot software applications for local or remote disaster recovery and archiving purposes. For a list of data protection partners, please see protection. Digital Surveillance The Pivot3 STAC is ideally suited for the write intensive applications common in video surveillance environments, such as video management, access control, or video analytics. Different applications can be run on each appliance and still have access to the shared storage of the underlying Pivot3 shared storage. For a list of digital surveillance partners, see surveillance. Rich Media The capacity scaling of the Pivot3 STAC architecture is well suited to rich media environments while the versatility of both block and file access can simplify installation support. Either the Windows Storage Server or Linux file systems (such as Samba) can run on virtual servers in the Pivot3 STAC. For a list of rich , Pivot3, Inc., All rights reserved. 5

6 media partners, see media , Pivot3, Inc., All rights reserved. 6

7 2. Pivot3 RAIGE Operating System The Pivot3 RAIGE (RAID Across General Ethernet) Operating System runs on each Pivot3 appliance in a Pivot3 STAC to deliver scale out storage across the Pivot3 appliances. RAIGE OS provides logical volume management, distributed data protection and automatic load balancing across appliances for ease of management, high availability and high performance. Logical Volume Management The RAIGE OS virtualizes physical disks and appliances in a Pivot3 STAC so that capacity can be managed logically beyond the physical limits of each appliance. Reliability also improves because volume access is not disrupted by physical hardware component failures that can include an entire appliance. Pivot3 appliances are discovered and managed by RAIGE Director Software that runs on any PC or server connected to the STAC. Physical appliances located on the same local subnet can be selected and assigned to one or more Pivot3 STACs. Multiple Pivot3 STACs can be managed with one instance of the RAIGE Director Software. Appliances are assigned to a STAC using the RAIGE Director Software , Pivot3, Inc., All rights reserved. 7

8 Capacity Management The aggregate capacity of the underlying appliances can then be parceled into logical volumes using the RAIGE Director Software. Attributes for each logical volume, such as RAID protection, name, rebuild priority, and access control are set by volume without requiring knowledge of the underlying physical hardware. Capacity can be physically and logically added to both the STAC and to existing logical volumes. Capacity expansion is dynamic and does not interfere with data storage or retrieval processes. The aggregate capacity of the Pivot3 appliances is presented as a multi ported iscsi target. Logical volumes are created from the Pivot3 STAC Initiator Management Access to volumes is managed by a list of iscsi initiators that are allowed to login to specific volumes. iscsi initiator logins can either be un authenticated or authenticated via MD5 based CHAP and an initiator shared secret. For initiators that support mutual CHAP logins, the STAC can be configured to authenticate volumes to the initiator via MD5 based CHAP and an STAC wide shared secret. Initiator names/logins have an Access Control List (ACL) that allows the administrator to define Read Write, Read Only or no access to volumes. STAC Management Pivot3 provides a software utility called the RAIGE Connection Manager to automatically configure and maintain network connections between servers and volumes since there can be many connections in a large Pivot3 STAC. Distributed Data Protection The key elements of the Pivot3 RAIGE OS are distributed RAID algorithms, parallel disk write cache, Unified Flash and Disk Caching, virtual global sparing, parallel rebuilding of failed drives, priority rebuilding of volumes, continuous background verification and predictive sparing , Pivot3, Inc., All rights reserved. 8

9 RAID Algorithms The RAIGE OS distributes data and parity across Pivot3 appliances so that data is efficiently protected against component failures. There is no need to create physical disk or RAID sets as you would with a traditional RAID system. Rather, disks are treated as raw capacity and the RAID function is implemented at the volume level. The normally burdensome management tasks of defining RAID groups and partitioning volumes, which are associated with traditional RAID devices, are not necessary with a Pivot3 STAC. RAID Protection Four RAID protection levels are provided to meet the data protection goals of each application: RAID 1e Enhanced network mirroring protects against either one disk drive failure or one appliance failure. The e indicates enhanced RAID 1 for STACs of three or more appliances where data is protected even if an entire appliance fails. Data is protected by striping an exact copy of the primary data across drives in another appliance in the STAC. Capacity required for RAID 1e protection is roughly one half of the appliances in a STAC since two copies of the data are distributed in the STAC. Precise usable capacity is included in the product specification sheets. RAID 5e Enhanced distributed parity protects against either one disk drive failure or one appliance failure. The e indicates enhanced RAID 5 for STACs of three or more appliances where data is also protected even if an entire appliance fails. Data is striped across each appliance in the STAC and protected by one level of parity. Distributed parity is interleaved across each appliance in the STAC so that data stored in each appliance is protected by parity in another appliance. Capacity required for RAID 5e protection is roughly one appliance per STAC since one copy of parity is distributed in the STAC. For precise usable capacity, please refer to the product specification sheets , Pivot3, Inc., All rights reserved. 9

10 RAID 6e Enhanced distributed dual parity protects against either three drive failures or one appliance failure plus one drive failure. The e indicates enhanced RAID 6 for STACs of three or more appliances where data is also protected even if an entire appliance fails. Data is striped across each appliance in the STAC and protected by two levels of parity. The first level of parity, network parity, is striped across each appliance in the STAC, much like the parity in RAID 5e. The next level of parity, disk parity, is striped across the disks within an appliance. Capacity required for RAID 6e protection is roughly one appliance per STAC and one drive per appliance to store two different parities in the STAC. Precise usable capacity is included in the product specification sheets. RAID 6x Enhanced distributed triple parity protects against either five drive failures or one appliance failure plus two drive failures. Like RAID 5e and 6e, RAID 6x provides fault tolerance even if an entire appliance fails for STACs of three or more appliances. Data is striped across each appliance in the STAC and protected by three levels of parity. The first level of parity, network parity, is striped across each appliance in the STAC, much like the parity in RAID 5e. The next two levels of parity, dual disk parity, are striped across the disks within an appliance. Capacity required for RAID 6x protection is roughly one appliance per STAC and two drives per appliance. Precise usable capacity is included in the product specification sheets , Pivot3, Inc., All rights reserved. 10

11 RAID Efficiency One important factor to consider when selecting an appropriate RAID level is balancing between fault tolerance and usable capacity in a STAC. Simply put, higher protection levels result in lower usable capacity. Pivot3 s innovative implementation of distributed RAID delivers up to 90% usable capacity and protects against more drive failures than scale out approaches that offer simple mirroring. The higher usable capacity similarly delivers improved power, cooling and space efficiency which is especially important for large scale deployments. Disk Groups The RAIGE OS maintains logical Disk Groups that further minimize the effect of drive failures on the overall STAC. Disk Groups consist of one drive per appliance and are automatically created and maintained by the RAIGE OS. By organizing the placement of parity and mirror data within one Disk Group, the impact of a drive failure is limited to its Disk Group. Disk Groups effectively increase the number of simultaneous drive failures that each Pivot3 STAC can sustain without data loss since drive failures outside of a Disk Group do not affect other Disk Groups. Online Management: Changing and Mixing RAID Protection in a STAC Since RAID protection is set by volume, different RAID levels may co exist on a STAC and RAID levels can be changed while the system is running, without disruption to data reads or writes. For example, a RAID 6e volume can be changed to a RAID 5e volume to free up space while access to the volume continues uninterrupted. Dynamic Expansion: Adding Physical Capacity to an Existing STAC Capacity can be physically added to an existing Pivot3 STAC by connecting new physical appliances to the SAN and then configuring them into the existing Pivot3 STAC using the RAIGE Director Software. The additional physical capacity is dynamically added to the Pivot3 STAC and data is restriped across the appliances so that capacity is automatically provisioned and performance is automatically load balanced across all controllers and iscsi connections. The complexity normally associated with managing meta LUNs for volumes larger than the domain of one physical conventional RAID controller is consequently eliminated. Allocate on write The RAIGE OS uses an allocate on write method so that a configured volume can be written to immediately and does not require disk formatting time, which for large conventional arrays may take over 24 hours. Scale out Unified Disk and Flash Write Caches Pivot3 RAIGE OS uses a patented scale out write cache approach that aggregates available disk and solid state flash resources across appliances in a STAC. Volumes benefit from write cache resources that scale as appliances are added to a STAC. Disk writes are automatically load balanced across all available cache resources in a STAC. Unified Disk Cache The Pivot3 Unified Disk Write Cache is specific to DataBank and CloudBank appliances and provides cost effective performance for streaming applications that are characterized by long block random write mostly operations. The Pivot3 Disk Write Cache protects in flight data against power loss and aggregates long block write IO s to maximize performance. The Disk Write Cache maximizes ingest rates while eliminating the need for expensive battery backed RAM. This patented approach takes advantage of the massive network bandwidth available in a Pivot3 STAC and delivers high performance by distributing many parallel writes to the SATA disks within and across appliances , Pivot3, Inc., All rights reserved. 11

12 On each physical disk, cache zones spread across the sectors of each disk are used for the intermediate caching of write data. Depending on the position of the disk head on the platter when a write request occurs, the cached data is saved in the nearest cache zone, greatly reducing head seek latencies. Host acknowledgements allow the host server to move on to the next activity and the cached data is moved to its final placement on the media as a background task. Disk Write Cache Zones Unified Flash Cache Pivot3 vbank appliances add support for solid state drive (SSD) Unified Flash Cache to improve performance for short block random write performance characteristics of latency sensitive database applications. The Unified Flash Cache is similar in implementation to the disk write cache in that many writes are processed in parallel within and across appliances. The Unified Flash Cache takes advantage of two parallel 25 GB SSDs in each appliance to reduce latency for more random workloads. Write Process When an application writes data to a Pivot3 STAC, the following processes take place to complete the write request: When RAIGE Director creates a logical volume, RAIGE OS creates a logical block address (LBA) map. The map assigns a unique LBA owner for each block that is being written. When RAIGE OS receives a write request, the request is routed to the correct LBA owner who becomes the master for that write. The master for the write simultaneously places the data in its cache Disk Write Cache or Unified Flash Cache and sends the data to another appliance in the STAC to be placed in the second cache. Only after the data is safely stored in both caches, RAIGE OS sends an acknowledgement for the write request to the application. As a background task, the data is moved from the cache to its proper LBA location where it remains and can be read or overwritten at any time. RAIGE OS also calculates appropriate network and disk parities and writes them across appliances. For RAID 1e and 5e, each appliance writes to one disk. For RAID 6e, the data and/or parities are written to two disks in each appliance and for RAID 6x, three disks in each appliance. These multiple writes protect the data from power loss or disk failure. As additional appliances are added to a STAC, or if an appliance fails, the LBA map is updated and the new map holds the new unique locations of each data block , Pivot3, Inc., All rights reserved. 12

13 Virtual Global Sparing Virtual drive sparing is used to automate and speed drive rebuilding if a drive fails in a Pivot3 STAC. The capacity of one logical spare drive is reserved across all of the drives in the STAC and removed from usable capacity. In the event of a drive failure, the rebuild process begins immediately using the previously reserved capacity. Unlike spare drives in conventional systems that standby during normal operation, virtual global spare drives in a Pivot3 STAC contribute to the overall performance of the RAID system during normal operation. Virtual RAID controller sparing is effectively supported since data is protected in the case of an appliance failure. Parallel Rebuild Innovation Conventional RAID systems are constrained by the physical relationship between RAID groups and their member disk drives. As a result, sparing and rebuilds are similarly constrained to physical drives. This becomes an important limitation as drive capacities grow to beyond 1 TB and rebuilding times for single drives increase. Pivot3 STACs provide extremely fast parallel rebuilds of failed drives because of the distributed nature of data allocation and sparing. Many drives contribute to the rebuild process and the recovered data is written to all drives resulting in a massively parallel activity. Only sectors of a failed disk that actually have data allocated and written need to be rebuilt which further speeds rebuild times in lesser utilized STACs. Priority Rebuilds by Volume Rebuilds are performed by volume. All volumes in the STAC are allocated to all of the drives in the STAC. An added benefit is the ability to designate a priority level for each volume so that higher priority volumes are rebuilt first. Rebuilding any specific volume may require rebuilding only a small portion of a drive. With conventional systems, rebuilding happens at the disk level which generally means volumes are only protected once the entire disk is fully rebuilt. Background Verification The Pivot3 STAC continuously performs background disk verification. Each disk is completely scanned to identify disks that are beginning to fail and to detect and repair bad blocks on the media. This is another process that benefits from the massive available bandwidth of the STAC and the processing power available in the Pivot3 appliances. Predictive Sparing Pivot3 Predictive Sparing is a background routine that continuously monitors disk drives to identify drives that could negatively impact overall system performance or that have a high likelihood of failure. Since drive deterioration is often characterized by gradual performance degradation, Predictive Sparing is an important method of gracefully removing suspect drives from the STAC, both to optimize performance and to keep the Pivot3 STAC in a fully protected state. Predictive Sparing includes two factory set thresholds for each drive. If the performance parameters for a drive exceed the first threshold, the low performing drive is automatically spared and the drive is subsequently and gracefully removed from the STAC. If the second higher threshold is exceeded Predictive Sparing fails the drive immediately and automatically initiates drive sparing, except in the case where failing the drive would cause data loss , Pivot3, Inc., All rights reserved. 13

14 Automatic Load Balancing Dynamic load balancing of bandwidth and capacity across initiators, network ports, appliance controllers and disk drives is managed by the RAIGE OS with no administrative intervention. Since data is equally distributed across the Pivot3 STAC, changes to either the physical infrastructure or logical entities can be quickly accommodated to eliminate disk, controller and network hot spots. For write operations, the dedicated IA 64 processors in each appliance have ample processing power for both RAID operations and TCP offload processing. For read operations, Pivot3 appliances return data to the application servers in parallel, providing load balanced performance across all appliances and all drives. The parallel architecture and load balancing of the RAIGE OS allow Pivot3 STACs to effectively aggregate many 1Gbps Ethernet ports and quickly surpass the bandwidth in proprietary 4Gbps Fibre Channel systems. Load balancing of capacity and performance extends to physical reconfiguration of each Pivot3 STAC. Following additions or removals of physical appliances to an existing STAC, the RAIGE OS restripes data across the new physical appliance count and automatically optimizes the load across the new physical network connections. VMware vmotion Support VMware vmotion technology allows a running VM to be dynamically migrated from one physical server to another. VMware vmotion allows the administrator to: Continuously and automatically optimize VMs within resource pools Perform hardware maintenance without scheduling downtime and disrupting business operations Proactively move VMs away from failing or underperforming servers Pivot3 STACs support vmotion by providing multi initiator support so that multiple external physical servers can access the same virtual machine volume files concurrently , Pivot3, Inc., All rights reserved. 14

15 3. Virtual Server Software Each Pivot3 appliance runs a virtualization software layer that allows storage and server operating systems to run simultaneously on the same appliance. Virtualization Software Pivot3 supports two powerful virtualization software products VMware s vsphere Hypervisor (ESXi) and the Xen hypervisor provided by Xen.org. Pivot3 tightly integrates VMware vsphere Hypervisor (ESXi) as the virtualization layer in its multi VM appliances, vbanks. Each vbank appliance can host multiple VMs to enable effective server consolidation. There is a broad range of supported guest operating systems, including Windows, Linux, Netware and Solaris. Pivot3 vbank STAC with Four Appliances , Pivot3, Inc., All rights reserved. 15

16 Pivot3 also integrates the Xen hypervisor in its single VM appliances, CloudBanks. Supported server operating systems for CloudBanks include Microsoft Windows Server, RedHat Enterprise Linux, CentOS and SUSE Linux operating systems. Pivot3 RAIGE Scale out SAN Windows or Linux VM Xen Virtualization Layer Intel 64 Platform Pivot3 CloudBank Appliance Virtual Machine I/O iscsi I/O in a Scale out Application Platform virtual machine (VM) begins with a single iscsi session between the iscsi initiator in the host and each Pivot3 logical volume accessible by the host. The iscsi sessions utilize a purely virtual network interface card (NIC) that interfaces directly to RAIGE OS running on the Pivot3 appliance hosting the VM. This network path is immune to failure points common in physical networks (cables, switches). A host I/O destined for a Pivot3 logical volume is sent by the iscsi initiator to the iscsi target for that logical volume through the virtual NIC. The RAIGE OS examines the logical block address associated with the command to determine which Pivot3 appliance the data should be written to or read from. If the I/O can be processed on the local appliance, the request is serviced immediately and the data never traverses a physical network cable. If the data is associated with another appliance in the STAC, the RAIGE OS on the local appliance will read or write the data to the appropriate Pivot3 appliance. This I/O takes place utilizing the fault tolerant and load balanced storage networks forming the backbone of the Pivot3 STAC. Once the data transfer is complete, status for the I/O is returned to the host over the virtual NIC. Direct Disk Access (DDA) Pivot3 s patent pending Direct Disk Access pipelines I/Os to significantly improve performance in virtualized environment. Traditionally, when creating VMs, a hypervisor has full control of the platform hardware and uses emulation of the I/O devices. This emulation layer enables the hypervisor to provide a consistent view of a hardware device and to isolate and contain device accesses to only those VMs that are assigned to the device by the hypervisor. However, it also slows down the I/O performance, creating as much as 30% performance overhead. With the Direct Disk Access approach, VMware vsphere Hypervisor (ESXi) assigns the disk controller directly to the RAIGE OS while still providing the isolation capability. This allows the RAIGE OS to access disk devices without loss of native performance, thus eliminating virtualization overhead. The Pivot3 approach could be considered the inverse of virtual storage appliance (VSA) solutions where a software solution collects residual storage on locally RAIDed virtual servers or VMs and presents this as shared storage. In the Pivot3 case, the virtual SAN is created prior to booting any VMs so that performance is not inhibited by virtual disk services within each appliance and capacity efficiency grows as the STAC size increases , Pivot3, Inc., All rights reserved. 16

17 Managing Virtual Servers Virtual servers running on a Pivot3 appliance have access to the entire shared capacity and bandwidth of the underlying iscsi storage STAC. Server instances can be started, stopped and managed as would any remote server. SNMP Support Pivot3 appliances can be monitored using Simple Network Management Protocol (SNMP). Community strings for SNMP are configured through the RAIGE Director Software and the SNMP MIB (management information base) is provided with the Pivot3 software. Because appliances cooperate within a STAC, SNMP agents can be set once at a STAC level and do not need to be set for each appliance. Pivot3 appliances run the Pivot3 SNMP agent and send out SNMP events to third party software applications that receive, or trap, the SNMP notifications. Since SNMP traffic is on the storage network, the server running third party application or an SNMP trap receiver needs to have access to both the storage network and the server network. Many network management applications use SMTP to forward traps as . The network management application will push an SMTP message to the corporate server, which will forward the message as an to a specified address. Data Replication Data replication is a key technology for disaster recovery (DR) and business continuity (BC). The Pivot3 Scale out Application Platform supports heterogeneous data replication by taking advantage of the embedded servers. Volume level and file level replication as well as continuous data protection (CDP) technologies are supported. By working closely with replication software providers, Pivot3 offers flexible and easy to manage replication solutions to protect against site failures. Replication software partners are listed on the Pivot3 web site at protection , Pivot3, Inc., All rights reserved. 17

18 Pivot3 Appliances Pivot3 vbank, CloudBank and DataBank appliances are the hardware building blocks of the STAC architecture. Each appliance is based on off the shelf server hardware. vbank: Scale out disk storage appliance with multiple ESXi VMs and Unified Flash Cache CloudBank: Scale out disk storage appliance with a single Xen VM DataBank: Scale out disk storage appliance 1. Pivot3 vbank Appliances Pivot3 RAIGE Scale out SAN storage Enterprise Compute Resources Intel Dual Quad core 5600 Series & 44 GB RAM for VMs VMware ESXi VMware failover (HA) and live migration (vmotion) Redundant Components Fans, power supplies, front loading disk drives Unified Flash Cache Solid state write cache for latency sensitive applications 6 Gigabit Ethernet NIC Ports 4 for VMs and 2 for iscsi Storage Capacity 12 or 24 TB hot swappable drives vbanks provide scale out IP storage as well as local VMware virtual server environments. Each vbank hosts multiple VMs that have access to shared SAN storage across all appliances in a Pivot3 vbank STAC. Storage volumes are created, protected and load balanced across the entire vbank STAC. Failover and live migration of VMs are controlled by VMware s HA and vmotion and managed through VMware vcenter. For the latest technical specifications, please see the Pivot3 vbank product specification sheet at Live Migration of VMs on vbanks VMware vsphere provides applications services that simplify the protection, load balancing and maintenance of multiple VMs. With VMware vmotion technology, a live running VM can be dynamically migrated from one physical server or vbank to another. VMware vmotion allows the administrator to: Continuously and automatically optimize VMs within resource pools such as vbank STACs Perform hardware maintenance without scheduling downtime and disrupting business operations Proactively move VMs away from failing or underperforming physical servers , Pivot3, Inc., All rights reserved. 18

19 Example: Live Migration within a vbank STAC In the above example, the administrator has detected that one of the appliances, vbank 1 with four VMs, is overloaded. Without interrupting applications and users on the VMs on this appliance, the administrator can quickly identify the best destination for a VM using vmotion s migration wizard and move a running VM, VM 4, to vbank 2. The administrator can even automate migrations to happen at pre defined times and without his presence, eliminating planned downtimes. VM Failover with Pivot3 vbank STACs VMware HA provides automated restart within minutes for all applications in the event of hardware or operating system failures. When a vbank appliance fails in a Pivot3 vbank STAC with three or more appliances, VMs on the failed vbank are automatically restarted on other available appliances in the Pivot3 STAC, protecting against unplanned downtime. Example: Appliance Failover within a vbank STAC VMware HA is easy to configure and requires minimal resources to provide failover protection. In this example, VMware HA provides the following: Automatic detection of server failures. VMware HA detects vbank failures and initiates the new virtual machine restart on a different vbank appliance in the resource pool (or the vbank STAC) without human intervention , Pivot3, Inc., All rights reserved. 19

20 Automatic detection of operating system failures. VMware HA detects operating system failures within VMs by monitoring heartbeat information. If a failure is detected, the affected VM is automatically restarted on the vbank. Smart failover of virtual machines to vbanks with best available resources. VMware HA automates the optimal placement of VMs restarted after server failure. Resource checks. VMware HA ensures capacity is always available in order to restart all VMs affected by a vbank failure. HA continuously monitors resource utilization and reserves spare capacity to be able to restart VMs. 2. Pivot3 CloudBank Appliances Pivot3 RAIGE Scale out SAN storage Enterprise Compute Resources Intel Dual Quad core 5600 Series & 8 GB RAM for VMs VMware ESXi VMware failover (HA) and live migration (vmotion) Redundant Components Fans, power supplies, front loading disk drives Unified Disk Cache Disk based cache for write intensive applications 6 Gigabit Ethernet NIC Ports 2 4 for VMs and 2 for iscsi Storage Capacity 4 24 TB storage with hot swappable drives CloudBanks provide scale out IP storage and local Xen virtual server environments. Each CloudBank delivers a single VM that has access to shared SAN storage across all appliances in a Pivot3 CloudBank STAC. Storage volumes are created, protected and load balanced across the entire CloudBank STAC. VM failover in CloudBanks is controlled by the Pivot3 RAIGE OS. For the latest technical specifications, please see the Pivot3 vbank product specification sheet at VM Failover with Pivot3 CloudBank STACs Pivot3 provides added server application reliability for CloudBank STACs of three or more appliances with VM recovery features that protect applications against unplanned server hardware outages. When a CloudBank appliance fails in a Pivot3 CloudBank STAC, the VM running on the failed CloudBank is automatically restarted on another available CloudBank in the Pivot3 STAC. This is possible since the virtual machine boot volumes are protected against an appliance failure and are available from the remaining appliances. The virtual machine dynamically re establishes virtual network and storage connections and will typically be running in less than two minutes , Pivot3, Inc., All rights reserved. 20

21 Integrated VM Failover Example: Pivot3 CloudBank STAC with 3 Appliances This self healing recovery mechanism reduces field support costs and does not require cluster server software, dedicated network connections or custom configurations. 3. Pivot3 DataBank Appliances Pivot3 RAIGE Scale out SAN storage 2 Gigabit Ethernet NIC Ports 2 for iscsi Redundant Components Fans, power supplies, front loading disk drives Unified Disk Cache Disk based cache for write intensive applications Storage Capacity TB storage with hot swappable drives DataBanks provide scale out IP storage where volumes are created, protected and load balanced across all appliances in a Pivot3 DataBank STAC. For the latest technical specifications, please see the Pivot3 DataBank product specification sheet at , Pivot3, Inc., All rights reserved. 21

22 Summary The Pivot3 STAC consolidates failover protected virtual servers and high availability shared storage on commodity hardware appliances. This unique architecture eliminates the cost, power, cooling, and rack space of physical servers, as well as dedicated failover hardware, while improving availability and simplifying management. By contrast, conventional storage arrays are a poor platform for integrating server virtualization technology for the simple reason that there is not enough compute horsepower available. Customers should look carefully at the predominant workloads of their environments and select virtualization platforms that best meet their requirements. The Pivot3 STAC can deliver the benefits of SAN storage with the server consolidation and server availability benefits commonly associated with server virtualization , Pivot3, Inc., All rights reserved. 22

23 Pivot3, Inc. 816 Congress Ave, Suite 970 Austin, TX Tel: Fax: Copyright 2011 Pivot3, Inc. All rights reserved. Specifications are subject to change without notice. Pivot3, RAIGE, STAC, vbank, CloudBank, DataBank, Unified Flash Cache, and DDA are trademarks or registered trademarks of Pivot3. All other trademarks are owned by their respective companies. TechOverview April 2011