Download presentation
Presentation is loading. Please wait.
Published byMagdalen Dean Modified over 8 years ago
1
Model-driven Data Layout Selection for Improving Read Performance Jialin Liu 1, Bin Dong 2, Surendra Byna 2, Kesheng Wu 2, Yong Chen 1 Texas Tech University 1 Lawrence Berkeley National Lab (LBNL) 2 HPDIC’14
2
Outline Motivation I/O Performance Model Implementation Evaluation Related Work Conclusion 2 HPDIC’14
3
Trend of Scientific Discovery: Data-intensive Scientific simulations tends to be data intensive VPIC, plasma physics, 30TB in a single run LHC, high energy physics, 1PB/yr, Hunt for the GOD Particle 3 Source: LBNL, CERN HPDIC’14
4
Read/Write Pattern Mismatching The read pattern is usually different with the write pattern Scientists choose optimal layouts for their application 4 Longitude Latitude Depth Longitude Latitude Depth Write PatternRead Pattern HPDIC’14
5
Various Data Layouts to Bridge the Gap Read access patterns to a single file may vary over time Scientists design various layouts: dimension transposing, hilbert curve, z-curve, etc. Z-curve Transposing 0 4 1 5 2 6 3 7 8 12 9 13 10 14 11 15 0 4 1 5 2 6 3 7 8 12 9 13 10 14 11 15 0 4 1 5 2 6 3 7 8 12 9 13 10 14 11 15 0 4 1 5 2 6 3 7 8 12 9 13 10 14 11 15 Hilbert-curve Original 5 HPDIC’14
6
Huge Parameter Space I/O (client)Layout (server) Independent/ Collective I/O Num(proc), num(aggregator) Access pattern: 1D, planes, sub- cube Buffer size, Request size Parallelism Chunk size Z-curve, Hilbert curve, Transposing Stripe count, Stripe size Block size, RAID level Data1 I/O Request1 Better Layouts? I/O Request2 6 Data2 Data3 Data4 Data5 Data6 HPDIC’14
7
Our Contributions 7 HPDIC’14 Multiple data layouts A new disk-level I/O representation A new parallel I/O performance model Automatic layout selection strategy in SDS
8
Challenges 8 Transparently direct the I/O to better organization Considering different I/O methods, e.g., independent and collective Aware of various data organization Low overhead HPDIC’14
9
Example I/O Architecture 9 Diagram of I/O Architecture on Hopper (NERSC) HPDIC’14 /scratch2, 1PB, Lustre, 13 lsi, 6 ost per oss, 2 oss per disk controller 156 OSTS, RAID6 (8+2) Block size is 512B, read/write in 4K
10
Disk-level I/O Representation 10 Abstract Two Level Data Striping HPDIC’14 HDF5 as data format; MPI-IO as middle layer; Luster as file system. Lustre splits the data into "stripes" and stores evenly on OSTs RAID 6 splits an OST stripe into smaller equal sized chunks of data and stores it on multiple disks
11
I/O Pattern Formalization 11 HPDIC’14 Given a logical I/O request in HDF5 format, e.g., I/O 1
12
BD&BG Estimation 12 HPDIC’14 Step 1: Map I/O to each strip Step 2: Compute BD Step 3: Compute BG
13
Model Design 13 Initial Observation HPDIC’14 The p-values of BG and BD are 0.000435 and 0.169462 BG is more significant than BD.
14
Model Design 14 Base linear I/O Model HPDIC’14 Parameter Space Reduce
15
Model Design 15 HPDIC’14 Refining for Independent I/O ost 0 ost 1 ost 2 p0 p1 pm Worst Case: I/O cost is proportional to Nproc Best Case P 6 P 5 P 4 P 3 P 2 P 1 P 0 1 1 2 2 3 3 4 4 5 5 6 6 7 7 Worst Case Block Order
16
Model Design 16 HPDIC’14 Process Aggregator ost 0 ost 1 ost 2 ost 3 Aggregated Requests Comm Phase< 20% Refining for Collective I/O
17
Implementation in SDS 17 HPDIC’14 The SDS Metadata Manager manages the metadata. The Layout Recommender direct the I/O to the best layout. The server periodically reads the frequently read patterns and contacts the Layout Recommender to identify the best data layouts, to partially re- organize.
18
Experimental Setup 18 Hopper, NERSC, Cray XE6 o Peak performance of 1.28 Petaflops/sec, 153,216 compute cores, 212 Terabytes of memory, and 2 Petabytes of disk Dataset o 3-D, 128G each file, 100 OST, 1M Library o HDF5 Test Layouts o 3 transposed datasets, various stripe size Test I/O pattern o 1D, 2D plane, 3D sub-cube, I/O size, processes HPDIC’14
19
Results 19 HPDIC’14 Synthetic benchmark, 3D array with [1024, 1024, 65536] size in [x, y, z] dimensions. Store data in [x, y, z], [x, z, y], and [z, x, y] dimensions that represent three different data layouts Independent I/O, 100 training sets, 8 tests for each pattern with 512 processes The average speedup for 1D patterns is 128.42X, that for 2D patterns is 68.46X, and that for 3D patterns is 9.26X.
20
Results 20 HPDIC’14 Average residual error is 40.7 Coefficient of determination R 2 is 0.9951
21
Results 21 HPDIC’14 Average residual error is 3.95 Coefficient of determination R 2 is 0.9969
22
Results 22 HPDIC’14
23
Related Work 23 HPDIC’14 Data Layout/Organization Optimization CHARISMA(Purakayastha), EDO(Tian), PDLA(Yin), Smart- IO(Tian), etc No optimal layout for all I/O requests, even for same access pattern I/O Performance Model HPC-IO(Shan), Queue-Raid(Smirni), Cost-IO(Song), Auto- Tune(You), Machine-L(Wang, Yu), Disk-sim(Bucy) Desire new workload representation to be aware various layouts.
24
Conclusion and Future Work 24 Conclusion Multiple layouts co-existing can benefit the I/O performance The model automatically select the best layout The model can be applied easily into the SDS Future Work Partial Data Replication with layout cover set HPDIC’14
25
Model-driven Data Layout Selection 25 Thanks HPDIC’14 http://discl.cs.ttu.edu/doku.php
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.