Beyond the File System Designing Large Scale File Storage and Serving Cal Henderson.

Beyond the File System

Designing Large Scale File Storage and Serving

Cal Henderson

Web Builder 2.0 2

Hello!

Web Builder 2.0 3

Big file systems?

• Too vague!

• What is a file system?

• What constitutes big?

• Some requirements would be nice

Web Builder 2.0 4

ScalableLooking at storage and serving infrastructures1

Web Builder 2.0 5

ReliableLooking at redundancy, failure rates, on the fly changes2

Web Builder 2.0 6

CheapLooking at upfront costs, TCO and lifetimes3

Web Builder 2.0 7

Four buckets

Storage

Serving

BCP

Cost

Web Builder 2.0 8

Storage

Web Builder 2.0 9

The storage stack

File system

Block protocol

RAID

Hardware

ext, reiserFS, NTFS

SCSI, SATA, FC

Mirrors, Stripes

Disks and stuff

File protocol NFS, CIFS, SMB

Web Builder 2.0 10

Hardware overview

The storage scale

Internal DAS SAN NAS

Lower Higher

Web Builder 2.0 11

Internal storage

• A disk in a computer– SCSI, IDE, SATA

• 4 disks in 1U is common

• 8 for half depth boxes

Web Builder 2.0 12

DAS

Direct attached storage

Disk shelf, connected by SCSI/SATA

HP MSA30 – 14 disks in 3U

Web Builder 2.0 13

SAN

• Storage Area Network

• Dumb disk shelves

• Clients connect via a ‘fabric’

• Fibre Channel, iSCSI, Infiniband– Low level protocols

Web Builder 2.0 14

NAS

• Network Attached Storage

• Intelligent disk shelf

• Clients connect via a network

• NFS, SMB, CIFS– High level protocols

Web Builder 2.0 15

Of course, it’s more confusing than that

Web Builder 2.0 16

Meet the LUN

• Logical Unit Number

• A slice of storage space

• Originally for addressing a single drive:– c1t2d3– Controller, Target, Disk (Slice)

• Now means a virtual partition/volume– LVM, Logical Volume Management

Web Builder 2.0 17

NAS vs SAN

With SAN, a single host (initiator) owns a single LUN/volume

With NAS, multiple hosts own a single LUN/volume

NAS head – NAS access to a SAN

Web Builder 2.0 18

SAN Advantages

Virtualization within a SAN offers some nice features:

• Real-time LUN replication

• Transparent backup

• SAN booting for host replacement

Web Builder 2.0 19

Some Practical Examples

• There are a lot of vendors

• Configurations vary

• Prices vary wildly

• Let’s look at a couple– Ones I happen to have experience with– Not an endorsement ;)

Web Builder 2.0 20

NetApp Filers

Heads and shelves, up to 500TB in 260U

FC SAN with 1 or 2 NAS heads

Web Builder 2.0 21

Isilon IQ

• 2U Nodes, 3-96 nodes/cluster, 6-600 TB

• FC/InfiniBand SAN with NAS head on each node

Web Builder 2.0 22

Scaling

Vertical vs Horizontal

Web Builder 2.0 23

Vertical scaling

• Get a bigger box

• Bigger disk(s)

• More disks

• Limited by current tech – size of each disk and total number in appliance

Web Builder 2.0 24

Horizontal scaling

• Buy more boxes

• Add more servers/appliances

• Scales forever*

*sort of

Web Builder 2.0 25

Storage scaling approaches

• Four common models:

• Huge FS

• Physical nodes

• Virtual nodes

• Chunked space

Web Builder 2.0 26

Huge FS

• Create one giant volume with growing space– Sun’s ZFS– Isilon IQ

• Expandable on-the-fly?

• Upper limits– Always limited somewhere

Web Builder 2.0 27

Huge FS

• Pluses– Simple from the application side– Logically simple– Low administrative overhead

• Minuses– All your eggs in one basket– Hard to expand– Has an upper limit

Web Builder 2.0 28

Physical nodes

• Application handles distribution to multiple physical nodes– Disks, Boxes, Appliances, whatever

• One ‘volume’ per node

• Each node acts by itself

• Expandable on-the-fly – add more nodes

• Scales forever

Web Builder 2.0 29

Physical Nodes

• Pluses– Limitless expansion– Easy to expand– Unlikely to all fail at once

• Minuses– Many ‘mounts’ to manage– More administration

Web Builder 2.0 30

Virtual nodes

• Application handles distribution to multiple virtual volumes, contained on multiple physical nodes

• Multiple volumes per node

• Flexible

• Expandable on-the-fly – add more nodes

• Scales forever

Web Builder 2.0 31

Virtual Nodes

• Pluses– Limitless expansion– Easy to expand– Unlikely to all fail at once– Addressing is logical, not physical– Flexible volume sizing, consolidation

• Minuses– Many ‘mounts’ to manage– More administration

Web Builder 2.0 32

Chunked space

• Storage layer writes parts of files to different physical nodes

• A higher-level RAID striping

• High performance for large files– read multiple parts simultaneously

Web Builder 2.0 33

Chunked space

• Pluses– High performance– Limitless size

• Minuses– Conceptually complex– Can be hard to expand on the fly– Can’t manually poke it

Web Builder 2.0 34

Real Life

Case Studies

Web Builder 2.0 35

GFS – Google File System

• Developed by … Google

• Proprietary

• Everything we know about it is based on talks they’ve given

• Designed to store huge files for fast access

Web Builder 2.0 36

GFS – Google File System

• Single ‘Master’ node holds metadata– SPF – Shadow master allows warm swap

• Grid of ‘chunkservers’– 64bit filenames– 64 MB file chunks

Web Builder 2.0 37

GFS – Google File System

1(a) 2(a)

1(b)

Master

Web Builder 2.0 38

GFS – Google File System

• Client reads metadata from master then file parts from multiple chunkservers

• Designed for big files (>100MB)

• Master server allocates access leases

• Replication is automatic and self repairing– Synchronously for atomicity

Web Builder 2.0 39

GFS – Google File System

• Reading is fast (parallelizable)– But requires a lease

• Master server is required for all reads and writes

Web Builder 2.0 40

MogileFS – OMG Files

• Developed by Danga / SixApart

• Open source

• Designed for scalable web app storage

Web Builder 2.0 41

MogileFS – OMG Files

• Single metadata store (MySQL)– MySQL Cluster avoids SPF

• Multiple ‘tracker’ nodes locate files

• Multiple ‘storage’ nodes store files

Web Builder 2.0 42

MogileFS – OMG Files

Tracker

Tracker

MySQL

Web Builder 2.0 43

MogileFS – OMG Files

• Replication of file ‘classes’ happens transparently

• Storage nodes are not mirrored – replication is piecemeal

• Reading and writing go through trackers, but are performed directly upon storage nodes

Web Builder 2.0 44

Flickr File System

• Developed by Flickr

• Proprietary

• Designed for very large scalable web app storage

Web Builder 2.0 45

Flickr File System

• No metadata store– Deal with it yourself

• Multiple ‘StorageMaster’ nodes

• Multiple storage nodes with virtual volumes

Web Builder 2.0 46

Flickr File System

SM

SM

SM

Web Builder 2.0 47

Flickr File System

• Metadata stored by app– Just a virtual volume number– App chooses a path

• Virtual nodes are mirrored– Locally and remotely

• Reading is done directly from nodes

Web Builder 2.0 48

Flickr File System

• StorageMaster nodes only used for write operations

• Reading and writing can scale separately

Web Builder 2.0 49

Serving

Web Builder 2.0 50

Serving files

Serving files is easy!

ApacheDisk

Web Builder 2.0 51

Serving files

Scaling is harder

ApacheDisk

ApacheDisk

ApacheDisk

Web Builder 2.0 52

Serving files

• This doesn’t scale well

• Primary storage is expensive– And takes a lot of space

• In many systems, we only access a small number of files most of the time

Web Builder 2.0 53

Caching

• Insert caches between the storage and serving nodes

• Cache frequently accessed content to reduce reads on the storage nodes

• Software (Squid, mod_cache)

• Hardware (Netcache, Cacheflow)

Web Builder 2.0 54

Why it works

• Keep a smaller working set

• Use faster hardware– Lots of RAM– SCSI– Outer edge of disks (ZCAV)

• Use more duplicates– Cheaper, since they’re smaller

Web Builder 2.0 55

Two models

• Layer 4– ‘Simple’ balanced cache– Objects in multiple caches– Good for few objects requested many times

• Layer 7– URL balances cache– Objects in a single cache– Good for many objects requested a few times

Web Builder 2.0 56

Replacement policies

• LRU – Least recently used

• GDSF – Greedy dual size frequency

• LFUDA – Least frequently used with dynamic aging

• All have advantages and disadvantages

• Performance varies greatly with each

Web Builder 2.0 57

Cache Churn

• How long do objects typically stay in cache?

• If it gets too short, we’re doing badly– But it depends on your traffic profile

• Make the cached object store larger

Web Builder 2.0 58

Problems

• Caching has some problems:

– Invalidation is hard– Replacement is dumb (even LFUDA)

• Avoiding caching makes your life (somewhat) easier

Web Builder 2.0 59

CDN – Content Delivery Network

• Akamai, Savvis, Mirror Image Internet, etc

• Caches operated by other people– Already in-place– In lots of places

• GSLB/DNS balancing

Web Builder 2.0 60

Edge networks

Origin

Web Builder 2.0 61

Edge networks

Origin

Cache

Cache

Cache

CacheCache

Cache

CacheCache

Web Builder 2.0 62

CDN Models

• Simple model– You push content to them, they serve it

• Reverse proxy model– You publish content on an origin, they proxy

and cache it

Web Builder 2.0 63

CDN Invalidation

• You don’t control the caches– Just like those awful ISP ones

• Once something is cached by a CDN, assume it can never change– Nothing can be deleted– Nothing can be modified

Web Builder 2.0 64

Versioning

• When you start to cache things, you need to care about versioning

– Invalidation & Expiry– Naming & Sync

Web Builder 2.0 65

Cache Invalidation

• If you control the caches, invalidation is possible

• But remember ISP and client caches

• Remove deleted content explicitly– Avoid users finding old content– Save cache space

Web Builder 2.0 66

Cache versioning

• Simple rule of thumb:– If an item is modified, change its name (URL)

• This can be independent of the file system!

Web Builder 2.0 67

Virtual versioning

• Database indicates version 3 of file

• Web app writes version number into URL

• Request comes through cache and is cached with the versioned URL

• mod_rewrite converts versioned URL to path

Version 3

example.com/foo_3.jpg

Cached: foo_3.jpg

foo_3.jpg -> foo.jpg

Web Builder 2.0 68

Authentication

• Authentication inline layer– Apache / perlbal

• Authentication sideline– ICP (CARP/HTCP)

• Authentication by URL– FlickrFS

Web Builder 2.0 69

Auth layer

• Authenticator sits between client and storage

• Typically built into the cache software

Cache

Authenticator

Origin

Web Builder 2.0 70

Auth sideline

• Authenticator sits beside the cache

• Lightweight protocol used for authenticator

Cache

Authenticator

Origin

Web Builder 2.0 71

Auth by URL

• Someone else performs authentication and gives URLs to client (typically the web app)

• URLs hold the ‘keys’ for accessing files

Cache OriginWeb Server

Web Builder 2.0 72

BCP

Web Builder 2.0 73

Business Continuity Planning

• How can I deal with the unexpected?– The core of BCP

• Redundancy

• Replication

Web Builder 2.0 74

Reality

• On a long enough timescale, anything that can fail, will fail

• Of course, everything can fail

• True reliability comes only through redundancy

Web Builder 2.0 75

Reality

• Define your own SLAs

• How long can you afford to be down?

• How manual is the recovery process?

• How far can you roll back?

• How many node x boxes can fail at once?

Web Builder 2.0 76

Failure scenarios

• Disk failure

• Storage array failure

• Storage head failure

• Fabric failure

• Metadata node failure

• Power outage

• Routing outage

Web Builder 2.0 77

Reliable by design

• RAID avoids disk failures, but not head or fabric failures

• Duplicated nodes avoid host and fabric failures, but not routing or power failures

• Dual-colo avoids routing and power failures, but my need duplication too

Web Builder 2.0 78

Tend to all points in the stack

• Going dual-colo: great

• Taking a whole colo offline because of a single failed disk: bad

• We need a combination of these

Web Builder 2.0 79

Recovery times

• BCP is not just about continuing when things fail

• How can we restore after they come back?

• Host and colo level syncing– replication queuing

• Host and colo level rebuilding

Web Builder 2.0 80

Reliable Reads & Writes

• Reliable reads are easy– 2 or more copies of files

• Reliable writes are harder– Write 2 copies at once– But what do we do when we can’t write to

one?

Web Builder 2.0 81

Dual writes

• Queue up data to be written– Where?– Needs itself to be reliable

• Queue up journal of changes– And then read data from the disk whose write

succeeded

• Duplicate whole volume after failure– Slow!

Web Builder 2.0 82

Cost

Web Builder 2.0 83

Judging cost

• Per GB?

• Per GB upfront and per year

• Not as simple as you’d hope– How about an example

Web Builder 2.0 84

Hardware costs

Cost of hardware

Usable GB

Single Cost

Web Builder 2.0 85

Power costs

Cost of power per year

Usable GB

Recurring Cost

Web Builder 2.0 86

Power costs

Power installation cost

Usable GB

Single Cost

Web Builder 2.0 87

Space costs

Cost per U

Usable GB

[ ]U’s needed (inc network)x

Recurring Cost

Web Builder 2.0 88

Network costs

Cost of network gear

Usable GB

Single Cost

Web Builder 2.0 89

Misc costs

Support contracts + spare disks

Usable GB

+ bus adaptors + cables[ ]Single & Recurring Costs

Web Builder 2.0 90

Human costs

Admin cost per node

Node countx

Recurring Cost

Usable GB

[ ]

Web Builder 2.0 91

TCO

• Total cost of ownership in two parts– Upfront– Ongoing

• Architecture plays a huge part in costing– Don’t get tied to hardware– Allow heterogeneity– Move with the market

(fin)

Web Builder 2.0 93

Photo credits

• flickr.com/photos/ebright/260823954/• flickr.com/photos/thomashawk/243477905/• flickr.com/photos/tom-carden/116315962/• flickr.com/photos/sillydog/287354869/• flickr.com/photos/foreversouls/131972916/• flickr.com/photos/julianb/324897/• flickr.com/photos/primejunta/140957047/• flickr.com/photos/whatknot/28973703/• flickr.com/photos/dcjohn/85504455/

Web Builder 2.0 94

You can find these slides online:

iamcal.com/talks/