Download presentation
Presentation is loading. Please wait.
Published byFranklin Mills Modified over 8 years ago
1
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Unistore: A Unified Storage Architecture for Cloud Computing Project Members: Wei Xie, Dr. Jiang Zhou, Dr. Yong Chen Presented by Wei Xie Data-Intensive Scalable Computing Laboratory(DISCL) Computer Science Department Texas Tech University
2
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Unistore: review of project plan Workloads Access patterns Devices Bandwidth Throughput Block erasure Concurrency Wear-leveling Characterization Component I/O Pattern Random/Sequential Read/write Hot/cold API Write_to_SSD Read_from_SSD Write_to_HDD Data Placement Component Placement Algorithm Modified Consistent Hash Based on sheepdog distributed store for virtual machine Optimization for heterogeneous storage (SSDs and HDDs) Optimization for heterogeneous workload
3
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Planned Schedule 2015, Q1: investigation and survey about Unistore 2015, Q2: characterization component development of Unistore 2015, Q3: metadata management of Unistore 2015, Q4: data distribution management of Unistore 2016, Q1: VM image store and loading 2016, Q2: advanced functions of Unistore 2016, Q3: performance optimization of Unistore 2016, Q4: module integration and system benchmarking finishedon-goingto-be-done
4
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Sheepdog: component Cluster manager QEMU block driver Object storage Gateway Object manager
5
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Sheepdog: object storage object typeWhat it contains data objectActual vdi data vdi object Metadata of image(name, size, data object ids, and etc) vmstate objectState info(used for snapshot) vdi attr objectsExtended attributes struct sd_inode { char name[SD_MAX_VDI_LEN]; /* the name of this VDI*/ uint64_t ctime; /* creation time of this VDI */ uint64_t vdi_size; /* the size of VDI */ uint64_t vm_state_size; /* the size of vm state (used for live snapshot) */ uint8_t nr_copies; /* the number of object redundancy */ uint32_t data_vdi_id[MAX_DATA_OBJS]; /* the data object IDs this VDI contains*/ …… };
6
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Sheepdog: gateway Responsible for where to store objects, or data placement Consistent hashing Add/remove node not significantly change mapping I/O load balance How to make the consistent hashing support heterogonous device? Two hash rings for HDD and SSD, respectively Write_to_HDD Write_to_SDD write
7
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Deployment 3 CentOS 6.5 virtual machines on iMac workstation Sheepdog built on the 3 virtual machine and form a cluster Use corosync to manage the cluster (can switch to zookeeper if necessary) Will migrate to a real Linux cluster later (for testing)
8
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Benchmarking Planned testing tools: fio, dd for generating synthetic workload Real workload benchmark iostat Comparison with other product GlusterFS Ceph
9
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Workload characterization Hot/cold data detection and separation Multiple bloom filter [1] Temporal locality [2] Spatial locality analysis (is it more random or sequential?) I/O size, write/read ratio, inter-arrival time, queue depth, latency, and IOPS. Online I/O trace collection for off-line analysis Online analysis [3] Characterization Component 9 [1] Park, Dongchul, and David HC Du. "Hot data identification for flash-based storage systems using multiple bloom filters." Mass Storage Systems and Technologies (MSST), 2011 IEEE 27th Symposium on. IEEE, 2011 [2] Wei Xie, Yong Chen, and Philip Roth, A Low-cost Adaptive Data Separation Method for the Flash Translation Layer of Solid State Drives, to be published. [3] Easy and Efficient Disk I/O Workload Characterization in VMware ESX Server
10
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Initial Characterization Result 10 Plan to implement online hot/cold data detection Hot data store on SSD, cold on HDD Initial result collected from OLTP I/O trace
11
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting 11
12
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Backup slides
13
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Sheepdog: strong consistency with epoch To keep replicated data consistent, epoch is used. Keep history of node member ship Check epoch before read/write epoch Node membership 1 A, B 2 A, B, C 3 A
14
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting Sheepdog: a read/write request default_write(oid, iocb){ … //check if epoch in iocb matches the system epoch if(iocb->epoch < sys_epoch()){ debug msg } write journal //write data to object file size=xpwrite(fd,iocb->buf, length, offset); } default_read(oid,iocb){ … //get the path of the object file to read get_store_path(oid, iocb->ec_index,path) //read from object file to iocb buffer default_read_from_path(oid,path,iocb) if(no object found && 0 epoch)<sys_epoch()) //if not able to read object, read from stale object read_from_stale_path(oid,path,iocb); }
15
April 9-10, 2015 Texas Tech University CAC@TTU Semiannual Meeting 15 Please take a moment to fill out your L.I.F.E. forms. http://iucrclife.chass.ncsu.edu/lifeforms/ What do you like about this project? What would you change? (Please include all relevant feedback.)
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.