Client-Based Access Control Management for XML documents Luc ...

Client-Based Access Control Management for XML documents Luc ...

1/25 Flash Device Support for Database Management CIDR 2011 Philippe Bonnet, ITU Copenhagen, Denmark Luc Bouganim, INRIA, Paris Rocquencourt, France This work is partially supported by the Danish Strategic Research Council. Outline 2/25 Note: These slides are an extended version of the slides shown at CIDR 2011 Motivation Flash device behavior The Good, the Bad and the FTL Minimal FTL Bimodal FTL

Example: Hash join on Bimodal FTL Conclusion DBMS on (or using) flash devices 3/25 NAND flash performance is impressive Flash devices is part of the memory hierarchy Replace or complement hard disks DBMS design = 3 decades of optimization based on the (initial) hard disk behavior Revisit the DBMS design wrt. flash device behavior? Need to understand the behavior of flash devices 4/25 Some examples of behavior (Samsung) SR, SW and RR have similar (good) performance RW, not shown, are much more expensive, 10-30ms Response time (s)s)

IO size (KB) 5/25 Some examples of behavior (Samsung) 100 Response time (ms) Response time (ms) 100 10 1 100 200 300 IO number 400 Random Writes (16KB)

Out of the box 1 rt Avg(rt) Avg(rt) o-o-b rt Avg(rt) 0.1 10 0.1 500 100 200 300 IO number 400 500

Random Writes (16 KB) After filling the device Average performance can vary of an order of magnitude depending on the device state 6/25 Some examples of behavior (Intel X25-E) SR, SW and RW have similar performance. RR are more costly! RW (16 KB) performance varies from 100 s)s to 100 ms!! (x 1000) 7/25 Some examples of behavior (Fusion IO) Capacity vs Performance tradeoff Sensitivity to device state Response time (s)s)

IO Size = 4KB Low level formatted Fully written Flash device behavior (1) 8/25 Understanding flash behavior [uFLIP, CIDR 2009] Flash devices (e.g., SSDs) do not behave as flash chips Flash devices performance is difficult to measure (device state) Need for an adequate methodology We proposed a wide benchmark to cover current and future devices. We also observed a common behavior and deduced design hints Not true anymore on recent devices! Making assumptions about flash behavior Consider the behavior of flash chips (embedded context) Consider the behavior of a given device or of a class of devices Flash device behavior (2) What is actually the behavior of flash devices?

Update in place are inefficient? Random writes are slower than sequential ones? Better not filling the whole device if we want good performance? Behavior varies across devices and firmware updates Should we continue running after the flash technology? In this talk, we propose another way to include flash devices in the DBMS landscape 9/25 10/25 The Good Flash devices performance is impressive! A single flash chip offers great performance e.g., 40 MB/s Read, 10 MB/s Write Random access is as fast as sequential access Low energy consumption A flash device contains many (e.g., 16, 32) flash chips and provides inter-chips parallelism

Flash devices include some (power-failure resistant) cache e.g., 16-32 MB of RAM 11/25 The Bad Flash chips have severe constraints! C1: Write granularity: Writes must be performed at flash page granularity (e.g. 4 KB) C2: Must erase a block (e.g., 64 pages) before rewriting a page C3: Writes must be sequential within a flash block C4: Limited lifetime (from 104 up to 106 erase operations) Write granularity: a page (4 KB) Writes must be sequential within the block

(64 pages) Erase granularity: a block (256 KB) 12/25 And The FTL The Flash Translation Layer (FTL) emulates a classical block device, handling flash constraints Distribute erase across flash (wear leveling) Address C4 (limited lifetime) Make out-of-place updates (using reserved flash blocks) Address C2 (erase before write) and C1 (writes smaller than a pageupdates) Maintain a logical to physical address mapping Necessary for out-of-place updates and wear leveling, address C3 (seq. writes)

A garbage collector is necessary! Logical to physical mapping Page Mapping: Logical @ Mapping table (900 MB for a 1 TB flash) 13/25 Physical @ Problem Page Logical @ Block Block Mapping: Page Block

Search for the correct page Mapping Problem table (12 MB for a 1 TB flash) Physical @ Beside these two extremes, many techniques were designed, using temporal/spatial locality, caching, detecting hotness of data, distinguishing RW and SW, grouping blocks, etc. FTL is a complex piece of software, generally kept secret by flash device manufacturers 14/25 FTL designers vs DBMS designers goals Flash device designers goals:

Hide the flash device constraints (usability) Improve the performance for most common workloads Make the device auto-adaptive Mask design decision to protect their advantage (black box approach) DBMS designers goals: Have a model for IO performance (and behavior) Predictable Clear distinction between efficient and inefficient IO patterns To design the storage model and query processing/optimization strategies Reach best performance, even at the price of higher complexity (having a full control on actual IOs) These goals are conflicting! 15/25 Minimal FTL: Take the FTL out of equation! FTL provides only wear leveling, using block mapping to address

C4 (limited lifetime) Pros Maximal performance for SR, RR, SW Semi-Random Writes Maximal control for the DBMS Constrained Patterns only (C1, C2, C3) Cons All complexity is handled by the DBMS All IOs must follow C1-C3 The whole DBMS must be rewritten The flash device is dedicated Minimal flash device

DBMS Block mapping, Wear Leveling (C4) Flash chips (C1) Write granularity (C2) Erase before write (C3) Sequential writes within a block (C4) Limited lifetime 16/25 Semi-random writes (uFLIP [CIDR09]) Inter-blocks : Random Intra-block : Sequential Example with 3 blocks of 10 pages: IO address

time Bimodal FTL: a simple idea 17/25 Bimodal Flash Devices: Provide a tunnel for those IOs that respect constraints C1-C3 ensuring maximal performance Manage other unconstrained IOs in best effort Minimize interferences between these two modes of operation Pros DBMS Flexible Maximal performance and control for the DBMS for constrained IOs No behavior guarantees for unconstrained IOs. Bimodal flash device

Cons unconstrained patterns constr. patterns (C1, C2, C3) Update mgt, Garb. Coll. C2,Wear C3) Leveling Block(C1, map., (C4) Flash chips (C1) Write granularity (C2) Erase before write (C3) Sequential writes within a block (C4) Limited lifetime 18/25

Bimodal FTL: easy to implement Constrained IOs lead to optimal blocks Flag = Optimal Page 0 Page 1 Page 2 Page 3 Page 4 Page 5 Flag = Non-Optimal CurPos=6 CurPos=6 Optimal blocks can be trivially mapped using a small map table in safe cache detected using a flag and cursor in safe cache

Page 0 Page 1 Page 1 Page 1 Page 0 Page 2 16 MB for a 1TB device No interferences! No change to the block device interface: Need to expose two constants: block size and page size 19/25 Bimodal FTL: better than Minimal + FTL Free Non-optimal block can become optimal (thanks to GC)

(CurPos = 0) TRIM TRIM Write at @ CurPos++ Write at @ CurPos Non optimal Optimal Write at @ CurPos++ Flag = Non-Optimal CurPos=6 Page 0 Page 1 Page 1 Page 1 Page 0

Page 2 Garbage collector actions Flag = Optimal CurPos=3 Page 0 Page 1 Page 2 Bimodal FTL does not exist yet! A simple test Device must support TRIM operation Only recent SSDs P1 Results on Intel X25-M

P2 P3 20/25 Impact on DBMS Design 21/25 Using bimodal flash devices, we have a solid basis for designing efficient DBMS on flash: What IOs should be constrained? i.e., what part of the DBMS should be redesigned? How to enforce these constraints? Revisit literature: Solutions based on flash chip behavior enforce C1-C3 constraints Solutions based on existing classes of devices might not. Example: Hash Join on HDD

One pass partitioning 22/25 Multi-pass partitioning (2 passes) Tradeoff: IOSize vs Memory consumption IOSize should be as large as possible, e.g., 256KB 1 MB To minimize IO cost when writing or reading partitions IOSize should be as small as possible To minimize memory consumption: One pass partitioning needs 2 x IOSize x NbPartitions in RAM Insufficient memory multi-pass performance degrades! 23/25 Hash join on SSD and on bimodal SSD With non bimodal SSDs

No behavior guarantees but Choosing IOSize = Block size (128 256 KB) should bring good performance With bimodal SSDs Maximal performance are guaranteed (constrained patterns) Use semi-random writes IOSize can be reduced up to page size (2 4 KB) with no penalty Memory savings Performance improvement Conclusion 24/25 Adding bimodality is necessary to support efficiently DBMS on flash devices DBMS designer retains control over IO performance DBMS leverages performance potential of flash chips Adding bimodality to FTL does not hinder competition between flash device manufacturers, they can

bring down the cost of constrained IO patterns (e.g., using parallelism) bring down the cost of unconstrained IO patterns without jeopardizing DBMS design This study is very preliminary many issues to explore More complex storage systems (e.g., RAID, ASM, etc) What abstraction for flash device? Memory abstraction (block device interface) Network abstraction (two systems collaborating) 25/25 More information Bimodal Flash devices: P. Bonnet, L. Bouganim : Flash Device Support for Database Management. 5th Biennial Conference on Innovative Data Systems Research (CIDR), January 2010. Benchmark: L. Bouganim, B. Jnsson, P. Bonnet. uFLIP: Understanding Flash IO Patterns, 4th Biennial Conference on Innovative Data Systems Research (CIDR), (Best paper award), January 2009

Energy consumption: M. Bjrling, P. Bonnet, L. Bouganim, Bjrn r Jnsson, uFLIP: Understanding the Energy Consumption of Flash Devices, IEEE Data Engineering Bulletin, vol. 33, n4, December 2010. Demonstration: M. Bjrling, L. Le Folgoc, A. Mseddi, P. Bonnet, L. Bouganim, Bjrn r Jnsson, Performing Sound Flash Device Measurements: The uFLIP Experience, 29th ACM International Conference on Management of Data (ACM SIGMOD), June. 2010. Web Sites:, , Authors: [email protected] , [email protected]

Recently Viewed Presentations