The document discusses designing large scale file storage and serving systems. It covers key requirements like scalability, reliability, and cost-effectiveness. It then describes various approaches to storage, such as network attached storage (NAS), storage area networks (SAN), and caching. Specific examples like Google File System, MogileFS, and Flickr's file system are examined. Business continuity planning (BCP) through redundancy and replication is also discussed.

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

2. Hello!

3. Big file systems? Too vague! What is a file system? What constitutes big? Some requirements would be nice

4. Scalable Looking at storage and serving infrastructures 1

5. Reliable Looking at redundancy, failure rates, on the fly changes 2

6. Cheap Looking at upfront costs, TCO and lifetimes 3

7. Four buckets Storage Serving BCP Cost

8. Storage

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

10. Hardware overview The storage scale NAS SAN DAS Internal Higher Lower

11. Internal storage A disk in a computer SCSI, IDE, SATA 4 disks in 1U is common 8 for half depth boxes

12. DAS Direct attached storage Disk shelf, connected by SCSI/SATA HP MSA30 – 14 disks in 3U

13. SAN Storage Area Network Dumb disk shelves Clients connect via a ‘fabric’ Fibre Channel, iSCSI, Infiniband Low level protocols

14. NAS Network Attached Storage Intelligent disk shelf Clients connect via a network NFS, SMB, CIFS High level protocols

15. Of course, it’s more confusing than that

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

17. NAS vs SAN With a 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

18. SAN Advantages Virtualization within a SAN offers some nice features: Real-time LUN replication Transparent backup SAN booting for host replacement

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 ;)

20. NetApp Filers Heads and shelves, up to 500TB in 6 Cabs FC SAN with 1 or 2 NAS heads

21. Isilon IQ 2U Nodes, 3-96 nodes/cluster, 6-600 TB FC/InfiniBand SAN with NAS head on each node

22. Scaling Vertical vs Horizontal

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

24. Horizontal scaling Buy more boxes Add more servers/appliances Scales forever* *sort of

25. Storage scaling approaches Four common models: Huge FS Physical nodes Virtual nodes Chunked space

26. Huge FS Create one giant volume with growing space Sun’s ZFS Isilon IQ Expandable on-the-fly? Upper limits Always limited somewhere

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

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

29. Physical Nodes Pluses Limitless expansion Easy to expand Unlikely to all fail at once Minuses Many ‘mounts’ to manage More administration

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

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

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

33. Chunked space Pluses High performance Limitless size Minuses Conceptually complex Can be hard to expand on the fly Can’t manually poke it

34. Real Life Case Studies

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

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

37. GFS – Google File System 1(a) 2(a) 1(b) Master

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

39. GFS – Google File System Reading is fast (parallelizable) But requires a lease Master server is required for all reads and writes

40. MogileFS – OMG Files Developed by Danga / SixApart Open source Designed for scalable web app storage

41. MogileFS – OMG Files Single metadata store (MySQL) MySQL Cluster avoids SPF Multiple ‘tracker’ nodes locate files Multiple ‘storage’ nodes store files

42. MogileFS – OMG Files Tracker Tracker MySQL

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

44. Flickr File System Developed by Flickr Proprietary Designed for very large scalable web app storage

45. Flickr File System No metadata store Deal with it yourself Multiple ‘StorageMaster’ nodes Multiple storage nodes with virtual volumes

46. Flickr File System SM SM SM

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

48. Flickr File System StorageMaster nodes only used for write operations Reading and writing can scale separately

49. Amazon S3 A big disk in the sky Multiple ‘buckets’ Files have user-defined keys Data + metadata

50. Amazon S3 Servers Amazon

51. Amazon S3 Servers Amazon Users

52. The cost Fixed price, by the GB Store: $0.15 per GB per month Serve: $0.20 per GB

53. The cost S3

54. The cost S3 Regular Bandwidth

55. End costs ~$2k to store 1TB for a year ~$63 a month for 1Mb ~$65k a month for 1Gb

56. Serving

57. Serving files Serving files is easy! Apache Disk

58. Serving files Scaling is harder Apache Disk Apache Disk Apache Disk

59. 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

60. 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)

61. 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

62. 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

63. 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

64. 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

65. Problems Caching has some problems: Invalidation is hard Replacement is dumb (even LFUDA) Avoiding caching makes your life (somewhat) easier

66. 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

67. Edge networks Origin

68. Edge networks Origin Cache Cache Cache Cache Cache Cache Cache Cache

69. 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

70. 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

71. Versioning When you start to cache things, you need to care about versioning Invalidation & Expiry Naming & Sync

72. 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

73. Cache versioning Simple rule of thumb: If an item is modified, change its name (URL) This can be independent of the file system!

74. 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

75. Authentication Authentication inline layer Apache / perlbal Authentication sideline ICP (CARP/HTCP) Authentication by URL FlickrFS

76. Auth layer Authenticator sits between client and storage Typically built into the cache software Cache Authenticator Origin

77. Auth sideline Authenticator sits beside the cache Lightweight protocol used for authenticator Cache Authenticator Origin

78. Auth by URL Someone else performs authentication and gives URLs to client (typically the web app) URLs hold the ‘keys’ for accessing files Cache Origin Web Server

79. BCP

80. Business Continuity Planning How can I deal with the unexpected? The core of BCP Redundancy Replication

81. Reality On a long enough timescale, anything that can fail, will fail Of course, everything can fail True reliability comes only through redundancy

82. 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 boxes can fail at once?

83. Failure scenarios Disk failure Storage array failure Storage head failure Fabric failure Metadata node failure Power outage Routing outage

84. 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 may need duplication too

85. 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

86. 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

87. 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?

88. 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!

89. Cost

90. Judging cost Per GB? Per GB upfront and per year Not as simple as you’d hope How about an example

91. Hardware costs Cost of hardware Usable GB Single Cost

92. Power costs Cost of power per year Usable GB Recurring Cost

93. Power costs Power installation cost Usable GB Single Cost

94. Space costs Cost per U Usable GB [ ] U’s needed (inc network) x Recurring Cost

95. Network costs Cost of network gear Usable GB Single Cost

96. Misc costs Support contracts + spare disks Usable GB + bus adaptors + cables [ ] Single & Recurring Costs

97. Human costs Admin cost per node Node count x Recurring Cost Usable GB [ ]

98. 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

99. (fin)

100. 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/

101. You can find these slides online: iamcal.com/talks/