System Availability and Data Protection of Infortrend s ESVA Storage Solution White paper Abstract This white paper analyzes system availability and data protection on Infortrend s ESVA storage systems. ESVA systems feature 99.999% availability and include excellent data protection, including RAID protection and local and remote replication capabilities, to ensure that users can enjoy extremely reliable and stable storage operations.
Table of Contents ESVA (Enterprise Scalable Virtualized Architecture) Storage Systems... 3 Combination of Features Contribute to ESVA s Excellent Reliability... 4 ESVA s Advanced Hardware Design Delivers High System Availability... 5 Component Redundancy... 5 Five Nines System Availability... 5 ESVA Data Protection Features... 7 RAID Protection... 7 Local and Remote Replication Capabilities... 8 Appendix I: Detailed Analysis of ESVA s Five Nines System Availability... 10 Appendix II: Local and Remote Replication Specifications on ESVA... 12 2
ESVA (Enterprise Scalable Virtualized Architecture) Storage Systems The Infortrend ESVA (Enterprise Scalable Virtualized Architecture) Series is a leading storage solution designed for mid-range enterprise SAN. At affordable prices, it meets mission-critical storage demands for performance, scalability and reliability with advanced hardware designs and comprehensive data services. On the innovative Enterprise Scalable Virtualized Architecture, various features, including storage virtualization, thin provisioning, distributed load balancing, automated storage tiering, automatic data migration, prioritized volume accessibility, snapshot and replication, are consolidated to realize optimal business benefits. With ESVA systems, users can optimize ROI, simplify storage infrastructure, and maximize application productivity. For more information about Infortrend s ESVA storage systems, please visit our website: http://esva.infortrend.com/ 3
Combination of Features Contribute to ESVA s Excellent Reliability Infortrend s ESVA systems are designed to be extremely reliable and secure. This is achieved with advanced hardware design that ensures high system availability and data protection features that help to guard data stored in ESVA systems. The following figure highlights the four major elements that contribute to ESVA s reliability: Figure 1: Key Elements Contributing to ESVA System Reliability 4
ESVA s Advanced Hardware Design Delivers High System Availability Component Redundancy The first step in achieving a reliable and stable storage system is designing the system s hardware in such a way that the system is nearly always available to users. Design features that are important in this respect are high-quality, redundant hardware components and the availability of the overall ESVA system. Redundant hardware components on ESVA systems include the RAID controllers, power supplies and cooling modules. The main benefit of redundant components is that a system will not shut down as a result of a single component failure. If a single component fails, the system can still operate normally. With redundant controllers, fault-tolerant data paths can be created so that data transmission to and from host applications can continue even in the event of a failure of one of the controllers. Five Nines System Availability System availability is an indication of normal operating time vs. non-operating time (due to system failure or other factors). Several statistics are important to determine system availability, including: MTBF (Mean Time Between Failure): the average time between two failures of a component or system. MTTR (Mean Time To Recovery): the average time needed to repair a component or system after a failure so that it operates normally again. MTTF (Mean Time To Failure): average time between the resumption of normal operations to the next failure. Availability is calculated as follows: MTTF/MTBF. 5
The table below shows often-used availability statistics (expressed in percentages) and the corresponding downtime of a system (per year). Obviously, each minute of downtime can have far-reaching consequences for enterprises. The differences between the percentages shown below can thus be very meaningful for companies that are evaluating storage systems for purchase. Availability Downtime Per Year 98% 7.30 days 99% 3.65 days 99.5% 1.83 days 99.9% 8.76 hours 99.99% 52.56 minutes 99.999% 5.26 minutes 99.9999% 31.54 seconds Table 1: Widely Used Availability Indicators and Corresponding Downtime Infortrend s ESVA systems feature an availability of 99.999% ( five nines ), meaning that on average an ESVA system will only suffer approximately five minutes of downtime per year 1. In terms of hardware, ESVA systems are extremely reliable and allow users to enjoy stable system operations. 1 6 For detailed calculations at the basis of ESVA systems availability, see the Appendix.
ESVA Data Protection Features Infortrend s ESVA storage systems include a number of features that ensure excellent data protection. These features include RAID protection and data services, such as local and remote replication capabilities. RAID Protection RAID (Redundant Array of Independent Disks) technology is at the core of ESVA systems. RAID involves the use of two or more disk drives to improve performance, error recovery, and fault tolerance. Mainstream RAID levels are described below: RAID 0 (Disk striping): delivers maximum performance, without any data redundancy. RAID 1 (Disk mirroring): data is mirrored, or copied, on two different disks, offering maximized data redundancy and excellent reliability. RAID 3 (Disk striping with dedicated parity): delivers excellent performance, while one dedicated parity drive offers data redundancy. RAID 5 (Disk striping with interspersed parity): delivers excellent performance; data redundancy is same as RAID 3, with the difference that parity data is interspersed among all member drives. RAID 6 (Disk striping with two interspersed parity disks): a RAID 6 configuration is essentially an extension of a RAID 5 configuration with a second independent distributed parity scheme that offers even better data redundancy. Infortrend ESVA systems are by default configured to operate with RAID 5, with excellent data redundancy to ensure that data will not be lost due to failure of individual disk drives. If users feel that the data protection offered by RAID 5 is not enough, they can choose to configure their storage system with RAID 6. This RAID level offers an extra parity scheme to ensure that data is even better protected. 7
Local and Remote Replication Capabilities In addition, ESVA systems offer data services including local and remote replication capabilities that offer a number of ways in which to improve data protection and avoid data loss. Local replication capabilities include snapshot and volume copy/mirror, while remote replication can be done in either synchronous or asynchronous mode. Snapshots allow users to perform instant, low-impact backups with space-efficient differential copies. In the instant of snapshot creation, a point-in-time data image is taken without disrupting online applications. Based on the image, data changes will be copied to the snapshot volume when new writes occur. With copy-on-write design, snapshots protect data from accidental modifications, deletions and corruptions with minimal capacity requirements and performance overhead. By accessing a snapshot copy as the desired recovery point, users can immediately restore system availability from data disruption. Figure 2: ESVA Snapshot Feature Volume mirror/copy allows users to non-disruptively create full data copies in a single storage pool. The copies created with volume mirror can be constantly synced with the source data or split as independent copies. When a volume fatally fails, these copies can be constantly synced and can be leveraged for quick recovery without any data loss. Split copies can be used to enhance operational resilience by allowing shared access of critical information among production and batch applications, such as testing, data mining, scientific analysis and tape backup. As business needs require, the split copies can be quickly re-synced with source. 8
Users can also create copies independent from source with volume copy technology. These copies are ideal for time-critical applications since they are immediately available for use even before the copying process is fully completed. Similar to split copies created by volume mirror, the editing of independent copies does not affect the service level of primary applications or their data integrity. Remote replication capability allows users to create full data copies across storage pools, either in synchronous or asynchronous mode. If source data fails due to system malfunctions or disasters, users can leverage the disk-based remote copy to restart service within only minutes. If the source needs to resume its role, it can be quickly synced with the remote copy for only differentials. To further ensure the integrity of the remote data, remote replication allows users to protect the remote copy with snapshot technology. Granular snapshot images can help restore the corrupt remote copy in seconds when it needs to resume business service. Figure 3: Full data copies in a single storage pool (local replication) and across storage pools (remote replication) 9
Appendix I: Detailed Analysis of ESVA s Five Nines System Availability ESVA system availability is based on individual component availability and the way in which individual components are linked together. The availability of individual components is calculated based on component specifications of component suppliers or Infortrend. From these individual availability figures, the availability of the overall ESVA system can be calculated. Analyzed components to determine system availability include: Ctrl PCBA: Printed Circuit Board Assembly located within the controller Host Board: Host board located within the controller BP: Backplane board within system enclosure; separates the front and rear sections of the enclosure FAN: Cooling fan modules; protect the system from overheating PSU: Power supply units; provides constant power to the subsystem The table below shows availability of individual components and overall system availability. Detailed formulas are provided in the notes that follow the table. Table 2: ESVA System Availability Notes: Individual component availability formula: MTTF/MTBF x 100 10
Redundant availability (parallel-coupled) formula: (1 - [(1-A#1) x (1-A#2)]) x 100 o A#1=Availability of Component #1 of a redundant pair o A#2=Availability of Component #2 of a redundant pair o Availability in decimal numbers Overall ESVA system availability: Ctrl PCBA availability x Host Board availability x BP availability x FAN availability x PSU availability x 100 o Except for BP (single component availability), other availability numbers in this formula are for parallel-coupled redundant components o Component availability in decimal numbers 11
Appendix II: Local and Remote Replication Specifications on ESVA The table below shows local and remote replication specifications for different ESVA models. Table 3: Local and Remote Replication Specifications on ESVA Systems 12