On the Trade-Offs among Performance, Energy, and Endurance in a Versatile Hybrid Drive

There are trade-offs among performance, energy, and device endurance for storage systems. Designs optimized for one dimension or workload often suffer in another. Therefore, it is important to study the trade-offs to enable adaptation to workloads and dimensions. As Flash SSD has emerged, hybrid drives have been studied more closely. However, hybrids are mainly designed for high throughput, efficient energy consumption, or improving endurance—leaving quantitative study on the trade-offs unexplored. Past endurance studies also lack a concrete model to help study the trade-offs. Last, previous designs are often based on inflexible policies that cannot adapt easily to changing conditions. We designed and developed GreenDM, a versatile hybrid drive that combines Flash-based SSDs with traditional HDDs. The SSD can be used as cache or as primary storage for hot data. We present our endurance model together with GreenDM to study these trade-offs. GreenDM presents a block interface and requires no modifications to existing software. GreenDM offers tunable parameters to enable the system to adapt to many workloads. We have designed, developed, and carefully evaluated GreenDM with a variety of workloads using commodity SSD and HDD drives. We demonstrate the importance of versatility to enable adaptation to various workloads and dimensions.

[1]  Richard F. Freitas,et al.  Storage class memory: Technology, systems and applications , 2009, 2010 IEEE Hot Chips 22 Symposium (HCS).

[2]  Nikolai Joukov,et al.  Operating system profiling via latency analysis , 2006, OSDI '06.

[3]  Zhichao Li,et al.  Model discovery for energy-aware computing systems: An experimental evaluation , 2011, 2011 International Green Computing Conference and Workshops.

[4]  Mahesh Balakrishnan,et al.  Extending SSD Lifetimes with Disk-Based Write Caches , 2010, FAST.

[5]  Yuanyuan Zhou,et al.  Hibernator: helping disk arrays sleep through the winter , 2005, SOSP '05.

[6]  Kang G. Shin,et al.  What does control theory bring to systems research? , 2009, OPSR.

[7]  Zhichao Li,et al.  Power consumption in enterprise-scale backup storage systems , 2012, FAST.

[8]  Yao Sun,et al.  PEARL: Performance, Energy, and Reliability Balanced Dynamic Data Redistribution for Next Generation Disk Arrays , 2008, 2008 IEEE International Symposium on Modeling, Analysis and Simulation of Computers and Telecommunication Systems.

[9]  Dirk Grunwald,et al.  Massive Arrays of Idle Disks For Storage Archives , 2002, ACM/IEEE SC 2002 Conference (SC'02).

[10]  Mircea R. Stan,et al.  How I Learned to Stop Worrying and Love Flash Endurance , 2010, HotStorage.

[11]  Eduardo Pinheiro,et al.  Failure Trends in a Large Disk Drive Population , 2007, FAST.

[12]  Jin Qian,et al.  PARAID: A gear-shifting power-aware RAID , 2007, TOS.

[13]  Steven Swanson,et al.  Reliably Erasing Data from Flash-Based Solid State Drives , 2011, FAST.

[14]  Hyojun Kim,et al.  Evaluating Phase Change Memory for Enterprise Storage Systems: A Study of Caching and Tiering Approaches , 2014, TOS.

[15]  Nikolai Joukov,et al.  Auto-pilot: A Platform for System Software Benchmarking , 2005, USENIX Annual Technical Conference, FREENIX Track.

[16]  Peter Desnoyers,et al.  What Systems Researchers Need to Know about NAND Flash , 2013, HotStorage.

[17]  Erez Zadok,et al.  Benchmarking File System Benchmarking: It *IS* Rocket Science , 2011, HotOS.

[18]  Akshat Verma,et al.  SRCMap: Energy Proportional Storage Using Dynamic Consolidation , 2010, FAST.

[19]  Zhichao Li,et al.  On the energy consumption and performance of systems software , 2011, SYSTOR '11.

[20]  Mahmut T. Kandemir,et al.  Revisiting widely held SSD expectations and rethinking system-level implications , 2013, SIGMETRICS '13.

[21]  Mahmut T. Kandemir,et al.  Scheduler-based DRAM energy management , 2002, DAC '02.

[22]  Rini T. Kaushik,et al.  GreenHDFS: towards an energy-conserving, storage-efficient, hybrid Hadoop compute cluster , 2010 .

[23]  John D. Strunk.  Hybrid aggregates: combining SSDs and HDDs in a single storage pool , 2012, OPSR.

[24]  Anand Sivasubramaniam,et al.  HybridStore: A Cost-Efficient, High-Performance Storage System Combining SSDs and HDDs , 2011, 2011 IEEE 19th Annual International Symposium on Modelling, Analysis, and Simulation of Computer and Telecommunication Systems.

[25]  Kimiko O. Bowman,et al.  Gamma Distribution , 2011, International Encyclopedia of Statistical Science.

[26]  Erez Zadok,et al.  vATM: vSphere Adaptive Task Management , 2012 .

[27]  J. Koomey.  Worldwide electricity used in data centers , 2008 .

[28]  Nikolai Joukov,et al.  GreenFS: making enterprise computers greener by protecting them better , 2008, Eurosys '08.

[29]  Tian Luo,et al.  hStorage-DB: Heterogeneity-aware Data Management to Exploit the Full Capability of Hybrid Storage Systems , 2012, Proc. VLDB Endow..

[30]  Erez Zadok,et al.  Energy and performance evaluation of lossless file data compression on server systems , 2009, SYSTOR '09.

[31]  Frank Mueller,et al.  Feedback EDF scheduling exploiting dynamic voltage scaling , 2004, Proceedings. RTAS 2004. 10th IEEE Real-Time and Embedded Technology and Applications Symposium, 2004..

[32]  Luiz André Barroso,et al.  The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines , 2009, The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines.

[33]  Ethan L. Miller,et al.  Pergamum: Replacing Tape with Energy Efficient, Reliable, Disk-Based Archival Storage , 2008, FAST.

[34]  Feng Chen,et al.  Hystor: making the best use of solid state drives in high performance storage systems , 2011, ICS '11.

[35]  Zhichao Li,et al.  On the Importance of Evaluating Storage Systems' $Costs , 2014, HotStorage.

[36]  Sungjin Lee,et al.  Lifetime management of flash-based SSDs using recovery-aware dynamic throttling , 2012, FAST.

[37]  Margo I. Seltzer,et al.  Operating system benchmarking in the wake of lmbench: a case study of the performance of NetBSD on the Intel x86 architecture , 1997, SIGMETRICS '97.

[38]  Scott A. Brandt,et al.  A Hybrid Disk-Aware Spin-Down Algorithm with I/O Subsystem Support , 2007, 2007 IEEE International Performance, Computing, and Communications Conference.

[39]  Himabindu Pucha,et al.  Cost Effective Storage using Extent Based Dynamic Tiering , 2011, FAST.

[40]  Arkady Kanevsky,et al.  Are disks the dominant contributor for storage failures?: A comprehensive study of storage subsystem failure characteristics , 2008, TOS.

[41]  Mahmut T. Kandemir,et al.  DRPM: dynamic speed control for power management in server class disks , 2003, 30th Annual International Symposium on Computer Architecture, 2003. Proceedings..

[42]  Amar Phanishayee,et al.  FAWN: a fast array of wimpy nodes , 2009, SOSP '09.

[43]  Navendu Jain,et al.  Managing cost, performance, and reliability tradeoffs for energy-aware server provisioning , 2011, 2011 Proceedings IEEE INFOCOM.