SlideShare a Scribd company logo

What is in All of Those SSTable Files Not Just the Data One but All the Rest Too! (John Schulz, The Pythian Group) | Cassandra Summit 2016

DataStax, profile picture
DataStax

The document provides an overview of SSTable file structure in Apache Cassandra, detailing essential components such as data files, indices, and metadata. It includes information on key concepts, a sample data table, and how to manage Cassandra clusters using the Cassandra Cluster Manager (CCM). Additionally, it describes the contents of individual SSTable files and their significance within the database system.

Software
1 of 43
Downloaded 75 times
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
John H Schulz What is in all of those sstable files? Not just the data one, but all the rest too!
1 Bio and My Employer 2 Some background 3 The Test Table and data 4 The actual file contents 5 Conclusions and Other Stuff 2© DataStax, All Rights Reserved.
Necessary Stuff Nap time if you wish
BIO Likes databases more than cheese sandwiches with peppered bacon and onions Open source databases since 2003 Relational databases since 1984 Owned one of the first Apple IIs in the neighborhood Lives in Michigan near the Detroit Zoo
About Pythian 11,400 Pythian currently manages more than 11,400 systems. 400+ Pythian currently employs more than 400 people in 200 cities in 35 countries 1997 Pythian was founded in 1997 Global Leader In IT Transformation And Operational Excellence Unparalleled Expertise • Top 5% in databases, applications, infrastructure, Big Data, Cloud, Data Science, and DevOps Unmatched Certifications • 9 Oracle ACEs, 4 Oracle ACE Directors, 1 Oracle ACE Associate • 6 Microsoft MVPs, 1 Microsoft Certified Master • 5 Google Platform Qualified Developers • 1 Cloudera Champion of Big Data • 1 Mongo DB Certified DBA Associate Level • DataStax Certified Partner, 1 MVP, 3 Architect, 2 Certified administrators Broad Technical Experience • Oracle, Microsoft, MySQL, Oracle EBS, Hadoop, Cassandra, MongoDB, virtualization, configuration management, monitoring, trending, and more.
Background
Background Some Terminology Cassandra data directory structure
Some Terminology Partition Key The column(s) which are hashed to create a token mapped by the partitioner Partition The rows and columns which belong to a specific Partition Key Clustering Key The column(s) which, with the Partition Key define a row Column A specific entity with a type (can be user defined) a value, version, time to live and a name
More on columns A single column can stand on its own in an SSTable file What is the minimum information needed to find and use a column Column Name Data type Cluster column values Partition column values Version (usually a timestamp) Time to Live Value
The Cassandra data directory structure data key-spaces Tables backups snapshots SSTable Files
Example partial directory tree from Cassandra 2.1 [root@Snoopy-3 node1]# ls -lR data data: total 12 drwxr-xr-x. 3 root root 4096 Aug 19 13:59 stuff drwxr-xr-x. 19 root root 4096 Aug 19 13:58 system drwxr-xr-x. 4 root root 4096 Aug 19 13:58 system_traces data/stuff: total 4 drwxr-xr-x. 2 root root 4096 Aug 19 14:00 simplefields-bdd61590663611e69c3e1d84c92693ab data/stuff/simplefields-bdd61590663611e69c3e1d84c92693ab: total 36
SSTable files in 2.1.9 8 Aug 19 14:00 stuff-simplefields-ka-1-CRC.db 1044 Aug 19 14:00 stuff-simplefields-ka-1-Data.db 9 Aug 19 14:00 stuff-simplefields-ka-1-Digest.sha1 24 Aug 19 14:00 stuff-simplefields-ka-1-Filter.db 130 Aug 19 14:00 stuff-simplefields-ka-1-Index.db 4460 Aug 19 14:00 stuff-simplefields-ka-1-Statistics.db 116 Aug 19 14:00 stuff-simplefields-ka-1-Summary.db 79 Aug 19 14:00 stuff-simplefields-ka-1-TOC.txt
The SSTable files Cassandra 3.0 8 Aug 19 10:42 mb-1-big-CRC.db 481 Aug 19 10:42 mb-1-big-Data.db 10 Aug 19 10:42 mb-1-big-Digest.crc32 24 Aug 19 10:42 mb-1-big-Filter.db 84 Aug 19 10:42 mb-1-big-Index.db 4799 Aug 19 10:42 mb-1-big-Statistics.db 80 Aug 19 10:42 mb-1-big-Summary.db 80 Aug 19 10:42 mb-1-big-TOC.txt
Parts of an SSTable file name Pre 2.2 stuff-simplefields-ia-1-Data.db 2.2 going forward mb-1-big-Data.db Version Counter What it is Keyspace table name Version Counter What it is
Data File Size Changed!!! Cassandra 2.1 1044 Aug 19 14:00 stuff-simplefields-ka-1-Data.db Cassandra 3.0 481 Aug 19 10:42 mb-1-big-Data.db
A relational view of how the files are connected © DataStax, All Rights Reserved. 16 DATA Summary Index Filter Yes/No Statistics Digest Compression TOC CRC METADATA
Metadata • Partitioner • Commit log location • Field names and types • Bloom filter chance • Counter information • Repair time • Cardinality estimate • Checksums • Compression Algorithm • Filenames © DataStax, All Rights Reserved. 17
The Test Table and Data
My test table CREATE TABLE stuff.simplefields ( field1 text, field2 text, field3 text, field4 timestamp, field5 decimal, PRIMARY KEY ((field1, field2), field3) ) WITH CLUSTERING ORDER BY (field3 ASC) AND compression = {};
Some test data to put in it insert into simpleFields (field1, field2, field3, field4, field5) values('FV1', 'FV2' , 'FV3', '2016-08-11 11:37:25.976', '127.00'); insert into simpleFields (field1, field2, field3, field4, field5) values('GV1', 'GV2' , 'GV3', '2016-08-11 11:37:25.976', '127.10'); insert into simpleFields (field1, field2, field3, field4, field5) values('AV1', 'AV2' , 'AV3', '2016-08-11 11:37:25.976', '127.20'); insert into simpleFields (field1, field2, field3, field4, field5) values('BV1', 'BV2' , 'BV3', '2016-08-11 11:37:25.976', '127.30'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV3', '2016-08-11 11:37:25.976', '127.40'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV4', '2016-08-11 11:37:25.976', '127.50'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV5', '2016-08-11 11:37:25.976', '127.60'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV6', '2016-08-11 11:37:25.976', '127.70');
My Data in the Table field1 | field2 | field3 | field4 | field5 --------+--------+--------+-------------------------+-------- CV1 | CV2 | CV3 | 2016-08-11 11:37:25.976 | 127.40 CV1 | CV2 | CV4 | 2016-08-11 11:37:25.976 | 127.50 CV1 | CV2 | CV5 | 2016-08-11 11:37:25.976 | 127.60 CV1 | CV2 | CV6 | 2016-08-11 11:37:25.976 | 127.70 FV1 | FV2 | FV3 | 2016-08-11 11:37:25.976 | 127.00 GV1 | GV2 | GV3 | 2016-08-11 11:37:25.976 | 127.10 BV1 | BV2 | BV3 | 2016-08-11 11:37:25.976 | 127.30 AV1 | AV2 | AV3 | 2016-08-11 11:37:25.976 | 127.20
How I got my clusters going Using CCM – Cassandra Cluster Manager https://github.com/pcmanus/ccm Example of creating a cluster: ccm create --vnodes --nodes=1 –v 2.1.9--start ver-219 ccm node1 nodetool status Datacenter: datacenter1 ======================= Status=Up/Down |/ State=Normal/Leaving/Joining/Moving -- Address Load Tokens Owns Host ID Rack UN 127.0.0.1 196.38 KB 256 ? 6ecec1cd-424d-4173-a83f-aefe485f4e2c rack1 © DataStax, All Rights Reserved. 22
Clusters currently on my Machine ver-12a ver-219 ver227 ver_33 ver-12a-multi ver-219-single ver-227-single ver_20_11a ver-221 ver-301 test ver-2011-single ver_225 ver-308 ver-120-single ver-213-multi ver_225_6 ver-308-single © DataStax, All Rights Reserved. 23
A peek inside the files
Quick overview of each File CRC – A single CRC of the Data file Compression – The name of the compression engine used Data – The SSTable data stored in sorted order by Partition/Cluster Key Digest – One or more CRCs one per 2GB segment of the Data file Filter – The bloom filter – folded hashes of all key values in the Data file Index – A list of all partition/Cluster Keys Statistics – Hodgepodge of metadata about the SSTable Summary – A list of some of the Partition/Cluster keys – Think second level index TOC – Text list of the files
A peek inside the less interesting files hexdump mb-1-big-CRC.db 0000000 0100 0000 c7ff 8b91 0000008 cat mb-1-big-Digest.crc32 4291269003 cat mb-1-big-TOC.txt Filter.db CRC.db Data.db Digest.crc32 Summary.db TOC.txt Statistics.db Index.db
Cassandra 2.1 Index 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 0 0 0000010 0 0 0 0 0 0 0 0 0 0 0 f 0 003 F V 0000020 1 0 0 003 F V 2 0 0 0 0 0 0 0 001 340 0000030 0 0 0 0 0 f 0 003 G V 1 0 0 003 G V 0000040 2 0 0 0 0 0 0 0 002 m 0 0 0 0 0 f 0000050 0 003 B V 1 0 0 003 B V 2 0 0 0 0 0 0000060 0 0 002 372 0 0 0 0 0 f 0 003 A V 1 0 0000070 0 003 A V 2 0 0 0 0 0 0 0 003 207 0 0 0000080 0 0
Cassandra 3.0 Index hexdump -c mb-1-big-Index.db 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 0 0 0000010 0 f 0 003 F V 1 0 0 003 F V 2 0 200 312 0000020 0 0 f 0 003 G V 1 0 0 003 G V 2 0 201 0000030 017 0 0 f 0 003 B V 1 0 0 003 B V 2 0 0000040 201 U 0 0 f 0 003 A V 1 0 0 003 A V 2 0000050 0 201 233 0
Cassandra 2.1 Summary 0000000 0 0 0 200 0 0 0 001 0 0 0 0 0 0 0 030 0000010 0 0 0 200 0 0 0 001 004 0 0 0 0 003 C V 0000020 1 0 0 003 C V 2 0 0 0 0 0 0 0 0 0 0000030 0 0 0 f 0 003 C V 1 0 0 003 C V 2 0 0000040 0 0 0 f 0 003 A V 1 0 0 003 A V 2 0 0000050 0 004 m m a p 0 0 0 001 0 0 0 0 0 0 0000060 0 0 0 004 m m a p 0 0 0 001 0 0 0 0 0000070 0 0 0 0
Cassandra 3.0 Summary hexdump -c mb-1-big-Summary.db 0000000 0 0 0 200 0 0 0 001 0 0 0 0 0 0 0 030 0000010 0 0 0 200 0 0 0 001 004 0 0 0 0 003 C V 0000020 1 0 0 003 C V 2 0 0 0 0 0 0 0 0 0 0000030 0 0 0 f 0 003 C V 1 0 0 003 C V 2 0 0000040 0 0 0 f 0 003 A V 1 0 0 003 A V 2 0
Cassandra 3.0 Statistics Part Beginning hexdump -c mb-1-big-Statistics.db 0000000 0 0 0 004 0 0 0 0 0 0 0 $ 0 0 0 001 0000010 0 0 0 Y 0 0 0 002 0 0 0 y 0 0 0 003 0000020 0 0 021 250 0 + o r g . a p a c h e 0000030 . c a s s a n d r a . d h t . M 0000040 u r m u r 3 P a r t i t i o n e 0000050 r ? 204 z 341 G 256 024 { 0 0 0 034 377 377 377 0000060 376 r 031 001 005 256 346 275 001 302 263 326 005 334 361 364 0000070 t 354 241 224 n 324 234 337 001 0 0 0 227 0 0 0 0000080 0 0 0 0 001 0 0 0 0 0 0 0 0 0 0 0
Cassandra 3.0 Statistics Tail end 00001230 70 65 29 01 28 6f 72 67 2e 61 70 61 63 68 65 2e |pe).(org.apache.| 00001240 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d 61 72 |cassandra.db.mar| 00001250 73 68 61 6c 2e 55 54 46 38 54 79 70 65 00 02 06 |shal.UTF8Type...| 00001260 66 69 65 6c 64 34 28 6f 72 67 2e 61 70 61 63 68 |field4(org.apach| 00001270 65 2e 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d |e.cassandra.db.m| 00001280 61 72 73 68 61 6c 2e 55 54 46 38 54 79 70 65 06 |arshal.UTF8Type.| 00001290 66 69 65 6c 64 35 28 6f 72 67 2e 61 70 61 63 68 |field5(org.apach| 000012a0 65 2e 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d |e.cassandra.db.m| 000012b0 61 72 73 68 61 6c 2e 55 54 46 38 54 79 70 65 |arshal.UTF8Type|
Cassandra 2.1 First Part © DataStax, All Rights Reserved. 33 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 177 377 0000010 377 377 200 0 0 0 0 0 0 0 0 t 0 003 C V 0000020 3 0 0 0 0 0 0 005 : p w w [ 341 0 0 0000030 0 0 0 017 0 003 C V 3 0 0 006 f i e l 0000040 d 4 0 0 0 005 : p w w [ 341 0 0 0 027 0000050 2 0 1 6 - 0 8 - 1 1 1 1 : 3 7 0000060 : 2 5 . 9 7 6 0 017 0 003 C V 3 0 0 0000070 006 f i e l d 5 0 0 0 005 : p w w [ 0000080 341 0 0 0 006 1 2 7 . 4 0 0 t 0 003 C 0000090 V 4 0 0 0 0 0 0 005 : p w w n / 0 00000a0 0 0 0 0 017 0 003 C V 4 0 0 006 f i e 00000b0 l d 4 0 0 0 005 : p w w n / 0 0 0 00000c0 027 2 0 1 6 - 0 8 - 1 1 1 1 : 3 00000d0 7 : 2 5 . 9 7 6 0 017 0 003 C V 4 0 00000e0 0 006 f i e l d 5 0 0 0 005 : p w w 00000f0 n / 0 0 0 006 1 2 7 . 5 0 0 t 0 003 0000100 C V 5 0 0 0 0 0 0 005 : p w w 200 261 0000110 0 0 0 0 0 017 0 003 C V 5 0 0 006 f i 0000120 e l d 4 0 0 0 005 : p w w 200 261 0 0 0000130 0 027 2 0 1 6 - 0 8 - 1 1 1 1 :
Cassandra 2.1 Second Part © DataStax, All Rights Reserved. 34 0000140 3 7 : 2 5 . 9 7 6 0 017 0 003 C V 5 0000150 0 0 006 f i e l d 5 0 0 0 005 : p w 0000160 w 200 261 0 0 0 006 1 2 7 . 6 0 0 t 0 0000170 003 C V 6 0 0 0 0 0 0 005 : p w w 213 0000180 a 0 0 0 0 0 017 0 003 C V 6 0 0 006 f 0000190 i e l d 4 0 0 0 005 : p w w 213 a 0 00001a0 0 0 027 2 0 1 6 - 0 8 - 1 1 1 1 00001b0 : 3 7 : 2 5 . 9 7 6 0 017 0 003 C V 00001c0 6 0 0 006 f i e l d 5 0 0 0 005 : p 00001d0 w w 213 a 0 0 0 006 1 2 7 . 7 0 0 0 00001e0 0 f 0 003 F V 1 0 0 003 F V 2 0 177 377 00001f0 377 377 200 0 0 0 0 0 0 0 0 t 0 003 F V 0000200 3 0 0 0 0 0 0 005 : p w v 271 d 0 0 0000210 0 0 0 017 0 003 F V 3 0 0 006 f i e l 0000220 d 4 0 0 0 005 : p w v 271 d 0 0 0 027 0000230 2 0 1 6 - 0 8 - 1 1 1 1 : 3 7 0000240 : 2 5 . 9 7 6 0 017 0 003 F V 3 0 0 0000250 006 f i e l d 5 0 0 0 005 : p w v 271 0000260 d 0 0 0 006 1 2 7 . 0 0 0 0 0 f 0
Cassandra 2.1 Third Part © DataStax, All Rights Reserved. 35 0000270 003 G V 1 0 0 003 G V 2 0 177 377 377 377 200 0000280 0 0 0 0 0 0 0 0 t 0 003 G V 3 0 0 0000290 0 0 0 0 005 : p w w : 255 0 0 0 0 0 00002a0 017 0 003 G V 3 0 0 006 f i e l d 4 0 00002b0 0 0 005 : p w w : 255 0 0 0 027 2 0 1 00002c0 6 - 0 8 - 1 1 1 1 : 3 7 : 2 5 00002d0 . 9 7 6 0 017 0 003 G V 3 0 0 006 f i 00002e0 e l d 5 0 0 0 005 : p w w : 255 0 0 00002f0 0 006 1 2 7 . 1 0 0 0 0 f 0 003 B V 0000300 1 0 0 003 B V 2 0 177 377 377 377 200 0 0 0 0000310 0 0 0 0 0 t 0 003 B V 3 0 0 0 0 0 0000320 0 005 : p w w Q 234 0 0 0 0 0 017 0 003 0000330 B V 3 0 0 006 f i e l d 4 0 0 0 005 0000340 : p w w Q 234 0 0 0 027 2 0 1 6 - 0 0000350 8 - 1 1 1 1 : 3 7 : 2 5 . 9 7 0000360 6 0 017 0 003 B V 3 0 0 006 f i e l d 0000370 5 0 0 0 005 : p w w Q 234 0 0 0 006 1 0000380 2 7 . 3 0 0 0 0 f 0 003 A V 1 0 0 0000390 003 A V 2 0 177 377 377 377 200 0 0 0 0 0 0
Cassandra 3.0 Part A © DataStax, All Rights Reserved. 36 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 177 377 0000010 377 377 200 0 0 0 0 0 0 0 $ 0 003 C V 3 0000020 $ 032 270 221 b 027 2 0 1 6 - 0 8 - 1 1 0000030 1 1 : 3 7 : 2 5 . 9 7 6 b 006 1 0000040 2 7 . 4 0 $ 0 003 C V 4 % + 300 D a 0000050 b 027 2 0 1 6 - 0 8 - 1 1 1 1 : 0000060 3 7 : 2 5 . 9 7 6 b 006 1 2 7 . 5 0000070 0 $ 0 003 C V 5 % , 300 N 234 b 027 2 0 0000080 1 6 - 0 8 - 1 1 1 1 : 3 7 : 2 0000090 5 . 9 7 6 b 006 1 2 7 . 6 0 $ 0 003 00000a0 C V 6 % , 300 [ ` b 027 2 0 1 6 - 0 00000b0 8 - 1 1 1 1 : 3 7 : 2 5 . 9 7 00000c0 6 b 006 1 2 7 . 7 0 001 0 f 0 003 F V 00000d0 1 0 0 003 F V 2 0 177 377 377 377 200 0 0 0 00000e0 0 0 0 0 $ 0 003 F V 3 # 032 0 b 027 2 00000f0 0 1 6 - 0 8 - 1 1 1 1 : 3 7 : 0000100 2 5 . 9 7 6 b 006 1 2 7 . 0 0 001 0 0000110 f 0 003 G V 1 0 0 003 G V 2 0 177 377 377
Cassandra 3.0 Part B © DataStax, All Rights Reserved. 37 0000120 377 200 0 0 0 0 0 0 0 $ 0 003 G V 3 $ 0000130 032 221 346 b 027 2 0 1 6 - 0 8 - 1 1 0000140 1 1 : 3 7 : 2 5 . 9 7 6 b 006 1 2 0000150 7 . 1 0 001 0 f 0 003 B V 1 0 0 003 B 0000160 V 2 0 177 377 377 377 200 0 0 0 0 0 0 0 $ 0000170 0 003 B V 3 $ 032 255 036 b 027 2 0 1 6 - 0000180 0 8 - 1 1 1 1 : 3 7 : 2 5 . 9 0000190 7 6 b 006 1 2 7 . 3 0 001 0 f 0 003 A 00001a0 V 1 0 0 003 A V 2 0 177 377 377 377 200 0 0 00001b0 0 0 0 0 0 $ 0 003 A V 3 $ 032 240 > b 00001c0 027 2 0 1 6 - 0 8 - 1 1 1 1 : 3 00001d0 7 : 2 5 . 9 7 6 b 006 1 2 7 . 2 0 00001e0 001
Bloom Filter contents 2.1 and 3.0 hexdump stuff-simplefields-ka-1-Filter.db 0000000 0000 0500 0000 0200 0032 2090 3002 0009 0000010 1803 0110 0022 40c0 hexdump mb-1-big-Filter.db 0000000 0000 0500 0000 0200 1808 5068 8822 0019 0000010 1012 1100 0002 4180 © DataStax, All Rights Reserved. 38
Conclusions References and questions
Conclusions • Cassandra 3.0 SSTable format is considerably more compact • In Cassandra 3.0 field names are no longer stored in the data file field Ids are used instead • TTL and Timestamps are grouped • TTL and Timestamp Differences are stored as deltas • Sorted String Tables are sorted…. By Partition Key Token Value • The Statistics file is full of a bunch of different stuff • Data, Summary, Filter and Statistics have important useful stuff © DataStax, All Rights Reserved. 40
References http://distributeddatastore.blogspot.com/2013/08/cassandra-sstable- storage-format.html https://github.com/scylladb/scylla/wiki/Storage-On-Disk-File-Formats http://thelastpickle.com/blog/2016/03/04/introductiont-to-the-apache- cassandra-3-storage-engine.html © DataStax, All Rights Reserved. 41
QUESTIONS? © DataStax, All Rights Reserved. 42
In version "ka", contents of Statistics component file is split in to 3 types of metadata called validation, compaction and stats. Validation metadata is used to validate SSTable which includes partitioner and bloom filter fp chance fields. Compaction metadata includes ancestors information which is also available in older formats and a new field called cardinality estimator. Cardinality estimator is used to efficiently pre-allocate bloom filter space in a merged compaction file by estimating how much the input SStables overlap. Stats metadata contains rest of the information available in old formats and two additional fields. First one is a flag to track the presence of local/remote counter shards and the other one is for storing the repair time. © DataStax, All Rights Reserved. 43

More Related Content

PPTX
Deletes Without Tombstones or TTLs (Eric Stevens, ProtectWise) | Cassandra Su...
DataStax
 
PPTX
Hecuba2: Cassandra Operations Made Easy (Radovan Zvoncek, Spotify) | C* Summi...
DataStax
 
PDF
Cassandra Community Webinar | In Case of Emergency Break Glass
DataStax
 
PDF
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
DataStax
 
PPTX
Lessons Learned From Running 1800 Clusters (Brooke Jensen, Instaclustr) | Cas...
DataStax
 
PDF
Maximum Overdrive: Tuning the Spark Cassandra Connector (Russell Spitzer, Dat...
DataStax
 
PDF
Cassandra Fundamentals - C* 2.0
Russell Spitzer
 
PPTX
Large partition in Cassandra
Shogo Hoshii
 
Deletes Without Tombstones or TTLs (Eric Stevens, ProtectWise) | Cassandra Su...
DataStax
 
Hecuba2: Cassandra Operations Made Easy (Radovan Zvoncek, Spotify) | C* Summi...
DataStax
 
Cassandra Community Webinar | In Case of Emergency Break Glass
DataStax
 
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
DataStax
 
Lessons Learned From Running 1800 Clusters (Brooke Jensen, Instaclustr) | Cas...
DataStax
 
Maximum Overdrive: Tuning the Spark Cassandra Connector (Russell Spitzer, Dat...
DataStax
 
Cassandra Fundamentals - C* 2.0
Russell Spitzer
 
Large partition in Cassandra
Shogo Hoshii
 

What's hot (20)

PDF
Lessons from Cassandra & Spark (Matthias Niehoff & Stephan Kepser, codecentri...
DataStax
 
PDF
A Cassandra + Solr + Spark Love Triangle Using DataStax Enterprise
Patrick McFadin
 
PPTX
Cassandra Tools and Distributed Administration (Jeffrey Berger, Knewton) | C*...
DataStax
 
PDF
Successful Architectures for Fast Data
Patrick McFadin
 
PDF
Advanced Apache Cassandra Operations with JMX
zznate
 
PDF
Storing time series data with Apache Cassandra
Patrick McFadin
 
PPTX
Real time data pipeline with spark streaming and cassandra with mesos
Rahul Kumar
 
PPTX
Maximum Overdrive: Tuning the Spark Cassandra Connector
Russell Spitzer
 
PDF
How We Used Cassandra/Solr to Build Real-Time Analytics Platform
DataStax Academy
 
PDF
Spark Cassandra Connector: Past, Present, and Future
Russell Spitzer
 
PDF
Cassandra EU - Data model on fire
Patrick McFadin
 
PDF
Owning time series with team apache Strata San Jose 2015
Patrick McFadin
 
PPTX
Using Spark to Load Oracle Data into Cassandra
Jim Hatcher
 
PDF
Introduction to cassandra 2014
Patrick McFadin
 
PDF
Cassandra and Spark: Optimizing for Data Locality-(Russell Spitzer, DataStax)
Spark Summit
 
PDF
Spark cassandra connector.API, Best Practices and Use-Cases
Duyhai Doan
 
PDF
Cassandra Community Webinar: Apache Cassandra Internals
DataStax
 
PDF
Understanding Cassandra internals to solve real-world problems
Acunu
 
PDF
Zero to Streaming: Spark and Cassandra
Russell Spitzer
 
PDF
Beyond the Query: A Cassandra + Solr + Spark Love Triangle Using Datastax Ent...
DataStax Academy
 
Lessons from Cassandra & Spark (Matthias Niehoff & Stephan Kepser, codecentri...
DataStax
 
A Cassandra + Solr + Spark Love Triangle Using DataStax Enterprise
Patrick McFadin
 
Cassandra Tools and Distributed Administration (Jeffrey Berger, Knewton) | C*...
DataStax
 
Successful Architectures for Fast Data
Patrick McFadin
 
Advanced Apache Cassandra Operations with JMX
zznate
 
Storing time series data with Apache Cassandra
Patrick McFadin
 
Real time data pipeline with spark streaming and cassandra with mesos
Rahul Kumar
 
Maximum Overdrive: Tuning the Spark Cassandra Connector
Russell Spitzer
 
How We Used Cassandra/Solr to Build Real-Time Analytics Platform
DataStax Academy
 
Spark Cassandra Connector: Past, Present, and Future
Russell Spitzer
 
Cassandra EU - Data model on fire
Patrick McFadin
 
Owning time series with team apache Strata San Jose 2015
Patrick McFadin
 
Using Spark to Load Oracle Data into Cassandra
Jim Hatcher
 
Introduction to cassandra 2014
Patrick McFadin
 
Cassandra and Spark: Optimizing for Data Locality-(Russell Spitzer, DataStax)
Spark Summit
 
Spark cassandra connector.API, Best Practices and Use-Cases
Duyhai Doan
 
Cassandra Community Webinar: Apache Cassandra Internals
DataStax
 
Understanding Cassandra internals to solve real-world problems
Acunu
 
Zero to Streaming: Spark and Cassandra
Russell Spitzer
 
Beyond the Query: A Cassandra + Solr + Spark Love Triangle Using Datastax Ent...
DataStax Academy
 
Ad

Viewers also liked (20)

PPTX
Webinar: Transforming Customer Experience Through an Always-On Data Platform
DataStax
 
PPTX
What We Learned About Cassandra While Building go90 (Christopher Webster & Th...
DataStax
 
PDF
Troubleshooting Cassandra (J.B. Langston, DataStax) | C* Summit 2016
DataStax
 
PDF
Data Modeling a Scheduling App (Adam Hutson, DataScale) | Cassandra Summit 2016
DataStax
 
PPTX
The Missing Manual for Leveled Compaction Strategy (Wei Deng & Ryan Svihla, D...
DataStax
 
PPTX
Webinar - Macy’s: Why Your Database Decision Directly Impacts Customer Experi...
DataStax
 
PDF
Can My Inventory Survive Eventual Consistency?
DataStax
 
PPTX
Bloor Research & DataStax: How graph databases solve previously unsolvable bu...
DataStax
 
PDF
Building Killr Applications with DSE
DataStax
 
PPTX
Using Spark to Load Oracle Data into Cassandra (Jim Hatcher, IHS Markit) | C*...
DataStax
 
PPTX
Webinar: Fighting Fraud with Graph Databases
DataStax
 
PDF
Tuning Speculative Retries to Fight Latency (Michael Figuiere, Minh Do, Netfl...
DataStax
 
PPTX
There are More Clouds! Azure and Cassandra (Carlos Rolo, Pythian) | C* Summit...
DataStax
 
PPTX
Give sense to your Big Data w/ Apache TinkerPop™ & property graph databases
DataStax
 
PDF
Using Approximate Data for Small, Insightful Analytics (Ben Kornmeier, Protec...
DataStax
 
PPTX
Webinar - DataStax Enterprise 5.1: 3X the operational analytics speed, help f...
DataStax
 
PDF
Cassandra Summit 2014: Reading Cassandra SSTables Directly for Offline Data A...
DataStax Academy
 
PDF
How Cassandra Deletes Data (Alain Rodriguez, The Last Pickle) | Cassandra Sum...
DataStax
 
PDF
Webinar - Bringing Game Changing Insights with Graph Databases
DataStax
 
PDF
2015 Internet Trends Report
IQbal KHan
 
Webinar: Transforming Customer Experience Through an Always-On Data Platform
DataStax
 
What We Learned About Cassandra While Building go90 (Christopher Webster & Th...
DataStax
 
Troubleshooting Cassandra (J.B. Langston, DataStax) | C* Summit 2016
DataStax
 
Data Modeling a Scheduling App (Adam Hutson, DataScale) | Cassandra Summit 2016
DataStax
 
The Missing Manual for Leveled Compaction Strategy (Wei Deng & Ryan Svihla, D...
DataStax
 
Webinar - Macy’s: Why Your Database Decision Directly Impacts Customer Experi...
DataStax
 
Can My Inventory Survive Eventual Consistency?
DataStax
 
Bloor Research & DataStax: How graph databases solve previously unsolvable bu...
DataStax
 
Building Killr Applications with DSE
DataStax
 
Using Spark to Load Oracle Data into Cassandra (Jim Hatcher, IHS Markit) | C*...
DataStax
 
Webinar: Fighting Fraud with Graph Databases
DataStax
 
Tuning Speculative Retries to Fight Latency (Michael Figuiere, Minh Do, Netfl...
DataStax
 
There are More Clouds! Azure and Cassandra (Carlos Rolo, Pythian) | C* Summit...
DataStax
 
Give sense to your Big Data w/ Apache TinkerPop™ & property graph databases
DataStax
 
Using Approximate Data for Small, Insightful Analytics (Ben Kornmeier, Protec...
DataStax
 
Webinar - DataStax Enterprise 5.1: 3X the operational analytics speed, help f...
DataStax
 
Cassandra Summit 2014: Reading Cassandra SSTables Directly for Offline Data A...
DataStax Academy
 
How Cassandra Deletes Data (Alain Rodriguez, The Last Pickle) | Cassandra Sum...
DataStax
 
Webinar - Bringing Game Changing Insights with Graph Databases
DataStax
 
2015 Internet Trends Report
IQbal KHan
 
Ad

Similar to What is in All of Those SSTable Files Not Just the Data One but All the Rest Too! (John Schulz, The Pythian Group) | Cassandra Summit 2016 (20)

PDF
SequoiaDB Distributed Relational Database
wangzhonnew
 
PDF
Distributed Computing for Everyone
Giovanna Roda
 
PDF
Backups
Svet Ivantchev
 
PDF
R data-import, data-export
FAO
 
PPT
Relational Algebra brief lecture notes for SQL.ppt
SubburamSivakumar1
 
PDF
Hailey_Database_Performance_Made_Easy_through_Graphics.pdf
cookie1969
 
PDF
AWS SSA Webinar 20 - Getting Started with Data Warehouses on AWS
Cobus Bernard
 
TXT
Redo logfile addition in oracle rac 12c
Debasish Nayak
 
DOCX
As400 load all subfile
aminem_mp
 
PDF
Aermap userguide under-revision
Ana Karina Calderón
 
PDF
Data Time Travel by Delta Time Machine
Databricks
 
PDF
Managing your Black Friday Logs NDC Oslo
David Pilato
 
PDF
Beyond php - it's not (just) about the code
Wim Godden
 
PDF
spark stream - kafka - the right way
Dori Waldman
 
PDF
Adaptive Query Processing on RAW Data
Manos Karpathiotakis
 
PDF
Managing your black friday logs - Code Europe
David Pilato
 
DOCX
Moving 12c database from NON-ASM to ASM
Monowar Mukul
 
PDF
Time series with Apache Cassandra - Long version
Patrick McFadin
 
PDF
Apache Spark Data intensive processing in practice
Marcin Szymaniuk
 
PDF
VirtaThon 2011 - Mining the AWR
Kristofferson A
 
SequoiaDB Distributed Relational Database
wangzhonnew
 
Distributed Computing for Everyone
Giovanna Roda
 
Backups
Svet Ivantchev
 
R data-import, data-export
FAO
 
Relational Algebra brief lecture notes for SQL.ppt
SubburamSivakumar1
 
Hailey_Database_Performance_Made_Easy_through_Graphics.pdf
cookie1969
 
AWS SSA Webinar 20 - Getting Started with Data Warehouses on AWS
Cobus Bernard
 
Redo logfile addition in oracle rac 12c
Debasish Nayak
 
As400 load all subfile
aminem_mp
 
Aermap userguide under-revision
Ana Karina Calderón
 
Data Time Travel by Delta Time Machine
Databricks
 
Managing your Black Friday Logs NDC Oslo
David Pilato
 
Beyond php - it's not (just) about the code
Wim Godden
 
spark stream - kafka - the right way
Dori Waldman
 
Adaptive Query Processing on RAW Data
Manos Karpathiotakis
 
Managing your black friday logs - Code Europe
David Pilato
 
Moving 12c database from NON-ASM to ASM
Monowar Mukul
 
Time series with Apache Cassandra - Long version
Patrick McFadin
 
Apache Spark Data intensive processing in practice
Marcin Szymaniuk
 
VirtaThon 2011 - Mining the AWR
Kristofferson A
 

More from DataStax (20)

PPTX
Is Your Enterprise Ready to Shine This Holiday Season?
DataStax
 
PPTX
Designing Fault-Tolerant Applications with DataStax Enterprise and Apache Cas...
DataStax
 
PPTX
Running DataStax Enterprise in VMware Cloud and Hybrid Environments
DataStax
 
PPTX
Best Practices for Getting to Production with DataStax Enterprise Graph
DataStax
 
PPTX
Webinar | Data Management for Hybrid and Multi-Cloud: A Four-Step Journey
DataStax
 
PPTX
Webinar | How to Understand Apache Cassandra™ Performance Through Read/Writ...
DataStax
 
PDF
Webinar | Better Together: Apache Cassandra and Apache Kafka
DataStax
 
PDF
Top 10 Best Practices for Apache Cassandra and DataStax Enterprise
DataStax
 
PDF
Introduction to Apache Cassandra™ + What’s New in 4.0
DataStax
 
PPTX
Webinar: How Active Everywhere Database Architecture Accelerates Hybrid Cloud...
DataStax
 
PPTX
Webinar | Aligning GDPR Requirements with Today's Hybrid Cloud Realities
DataStax
 
PDF
Designing a Distributed Cloud Database for Dummies
DataStax
 
PDF
How to Power Innovation with Geo-Distributed Data Management in Hybrid Cloud
DataStax
 
PDF
How to Evaluate Cloud Databases for eCommerce
DataStax
 
PPTX
Webinar: DataStax Enterprise 6: 10 Ways to Multiply the Power of Apache Cassa...
DataStax
 
PPTX
Webinar: DataStax and Microsoft Azure: Empowering the Right-Now Enterprise wi...
DataStax
 
PPTX
Webinar - Real-Time Customer Experience for the Right-Now Enterprise featurin...
DataStax
 
PPTX
Datastax - The Architect's guide to customer experience (CX)
DataStax
 
PPTX
An Operational Data Layer is Critical for Transformative Banking Applications
DataStax
 
PPTX
Becoming a Customer-Centric Enterprise Via Real-Time Data and Design Thinking
DataStax
 
Is Your Enterprise Ready to Shine This Holiday Season?
DataStax
 
Designing Fault-Tolerant Applications with DataStax Enterprise and Apache Cas...
DataStax
 
Running DataStax Enterprise in VMware Cloud and Hybrid Environments
DataStax
 
Best Practices for Getting to Production with DataStax Enterprise Graph
DataStax
 
Webinar | Data Management for Hybrid and Multi-Cloud: A Four-Step Journey
DataStax
 
Webinar | How to Understand Apache Cassandra™ Performance Through Read/Writ...
DataStax
 
Webinar | Better Together: Apache Cassandra and Apache Kafka
DataStax
 
Top 10 Best Practices for Apache Cassandra and DataStax Enterprise
DataStax
 
Introduction to Apache Cassandra™ + What’s New in 4.0
DataStax
 
Webinar: How Active Everywhere Database Architecture Accelerates Hybrid Cloud...
DataStax
 
Webinar | Aligning GDPR Requirements with Today's Hybrid Cloud Realities
DataStax
 
Designing a Distributed Cloud Database for Dummies
DataStax
 
How to Power Innovation with Geo-Distributed Data Management in Hybrid Cloud
DataStax
 
How to Evaluate Cloud Databases for eCommerce
DataStax
 
Webinar: DataStax Enterprise 6: 10 Ways to Multiply the Power of Apache Cassa...
DataStax
 
Webinar: DataStax and Microsoft Azure: Empowering the Right-Now Enterprise wi...
DataStax
 
Webinar - Real-Time Customer Experience for the Right-Now Enterprise featurin...
DataStax
 
Datastax - The Architect's guide to customer experience (CX)
DataStax
 
An Operational Data Layer is Critical for Transformative Banking Applications
DataStax
 
Becoming a Customer-Centric Enterprise Via Real-Time Data and Design Thinking
DataStax
 

Recently uploaded (20)

PDF
2025 Textile ERP Trends: SAP, Odoo & Oracle
Aetos Technologies
 
PDF
Adobe Premiere Pro 2025 (v24.5.0.057) Crack free
ghrom2211g
 
PDF
How to Migrate SBCGlobal Email to Yahoo Easily
raymondjones7273
 
PDF
Why TechBuilder is the Future of Pickup and Delivery App Development (1).pdf
TechBuilder
 
PDF
Softaken Excel to vCard Converter Software.pdf
markwillsonmw004
 
PDF
Internet Downloader Manager (IDM) Crack 6.42 Build 42 Updates Latest 2025
itskinga12
 
PDF
Claude Code: Everyone is a 10x Developer - A Comprehensive AI-Powered CLI Tool
William Chong
 
PPTX
Agentic AI : A Practical Guide. Undersating, Implementing and Scaling Autono...
IT IDOL Technologies
 
PPTX
Operating system designcfffgfgggggggvggggggggg
rmskumara09
 
PPTX
Lecture 3: Operating Systems Introduction to Computer Hardware Systems
hci7se
 
PDF
Nekopoi APK 2025 free lastest update
umarali35kp
 
PDF
Adobe Illustrator 28.6 Crack My Vision of Vector Design
ghrom2211g
 
PDF
Internet Downloader Manager (IDM) Crack 6.42 Build 41
ghrom2211g
 
PPTX
Introduction to Artificial Intelligence
lohedi9363
 
PPTX
ai tools demonstartion for schools and inter college
harshavardhanreddyka11
 
PPTX
L1 - Introduction to python Backend.pptx
MoizAhmed480282
 
PDF
How Creative Agencies Leverage Project Management Software.pdf
Orangescrum
 
PPTX
Transform Your Business with a Software ERP System
Canada Software Company
 
PDF
AI in Product Development-omnex systems
omnex systems
 
PDF
Addressing The Cult of Project Management Tools-Why Disconnected Work is Hold...
OnePlan Solutions
 
2025 Textile ERP Trends: SAP, Odoo & Oracle
Aetos Technologies
 
Adobe Premiere Pro 2025 (v24.5.0.057) Crack free
ghrom2211g
 
How to Migrate SBCGlobal Email to Yahoo Easily
raymondjones7273
 
Why TechBuilder is the Future of Pickup and Delivery App Development (1).pdf
TechBuilder
 
Softaken Excel to vCard Converter Software.pdf
markwillsonmw004
 
Internet Downloader Manager (IDM) Crack 6.42 Build 42 Updates Latest 2025
itskinga12
 
Claude Code: Everyone is a 10x Developer - A Comprehensive AI-Powered CLI Tool
William Chong
 
Agentic AI : A Practical Guide. Undersating, Implementing and Scaling Autono...
IT IDOL Technologies
 
Operating system designcfffgfgggggggvggggggggg
rmskumara09
 
Lecture 3: Operating Systems Introduction to Computer Hardware Systems
hci7se
 
Nekopoi APK 2025 free lastest update
umarali35kp
 
Adobe Illustrator 28.6 Crack My Vision of Vector Design
ghrom2211g
 
Internet Downloader Manager (IDM) Crack 6.42 Build 41
ghrom2211g
 
Introduction to Artificial Intelligence
lohedi9363
 
ai tools demonstartion for schools and inter college
harshavardhanreddyka11
 
L1 - Introduction to python Backend.pptx
MoizAhmed480282
 
How Creative Agencies Leverage Project Management Software.pdf
Orangescrum
 
Transform Your Business with a Software ERP System
Canada Software Company
 
AI in Product Development-omnex systems
omnex systems
 
Addressing The Cult of Project Management Tools-Why Disconnected Work is Hold...
OnePlan Solutions
 

What is in All of Those SSTable Files Not Just the Data One but All the Rest Too! (John Schulz, The Pythian Group) | Cassandra Summit 2016

  • 1. John H Schulz What is in all of those sstable files? Not just the data one, but all the rest too!
  • 2. 1 Bio and My Employer 2 Some background 3 The Test Table and data 4 The actual file contents 5 Conclusions and Other Stuff 2© DataStax, All Rights Reserved.
  • 4. BIO Likes databases more than cheese sandwiches with peppered bacon and onions Open source databases since 2003 Relational databases since 1984 Owned one of the first Apple IIs in the neighborhood Lives in Michigan near the Detroit Zoo
  • 5. About Pythian 11,400 Pythian currently manages more than 11,400 systems. 400+ Pythian currently employs more than 400 people in 200 cities in 35 countries 1997 Pythian was founded in 1997 Global Leader In IT Transformation And Operational Excellence Unparalleled Expertise • Top 5% in databases, applications, infrastructure, Big Data, Cloud, Data Science, and DevOps Unmatched Certifications • 9 Oracle ACEs, 4 Oracle ACE Directors, 1 Oracle ACE Associate • 6 Microsoft MVPs, 1 Microsoft Certified Master • 5 Google Platform Qualified Developers • 1 Cloudera Champion of Big Data • 1 Mongo DB Certified DBA Associate Level • DataStax Certified Partner, 1 MVP, 3 Architect, 2 Certified administrators Broad Technical Experience • Oracle, Microsoft, MySQL, Oracle EBS, Hadoop, Cassandra, MongoDB, virtualization, configuration management, monitoring, trending, and more.
  • 8. Some Terminology Partition Key The column(s) which are hashed to create a token mapped by the partitioner Partition The rows and columns which belong to a specific Partition Key Clustering Key The column(s) which, with the Partition Key define a row Column A specific entity with a type (can be user defined) a value, version, time to live and a name
  • 9. More on columns A single column can stand on its own in an SSTable file What is the minimum information needed to find and use a column Column Name Data type Cluster column values Partition column values Version (usually a timestamp) Time to Live Value
  • 10. The Cassandra data directory structure data key-spaces Tables backups snapshots SSTable Files
  • 11. Example partial directory tree from Cassandra 2.1 [root@Snoopy-3 node1]# ls -lR data data: total 12 drwxr-xr-x. 3 root root 4096 Aug 19 13:59 stuff drwxr-xr-x. 19 root root 4096 Aug 19 13:58 system drwxr-xr-x. 4 root root 4096 Aug 19 13:58 system_traces data/stuff: total 4 drwxr-xr-x. 2 root root 4096 Aug 19 14:00 simplefields-bdd61590663611e69c3e1d84c92693ab data/stuff/simplefields-bdd61590663611e69c3e1d84c92693ab: total 36
  • 12. SSTable files in 2.1.9 8 Aug 19 14:00 stuff-simplefields-ka-1-CRC.db 1044 Aug 19 14:00 stuff-simplefields-ka-1-Data.db 9 Aug 19 14:00 stuff-simplefields-ka-1-Digest.sha1 24 Aug 19 14:00 stuff-simplefields-ka-1-Filter.db 130 Aug 19 14:00 stuff-simplefields-ka-1-Index.db 4460 Aug 19 14:00 stuff-simplefields-ka-1-Statistics.db 116 Aug 19 14:00 stuff-simplefields-ka-1-Summary.db 79 Aug 19 14:00 stuff-simplefields-ka-1-TOC.txt
  • 13. The SSTable files Cassandra 3.0 8 Aug 19 10:42 mb-1-big-CRC.db 481 Aug 19 10:42 mb-1-big-Data.db 10 Aug 19 10:42 mb-1-big-Digest.crc32 24 Aug 19 10:42 mb-1-big-Filter.db 84 Aug 19 10:42 mb-1-big-Index.db 4799 Aug 19 10:42 mb-1-big-Statistics.db 80 Aug 19 10:42 mb-1-big-Summary.db 80 Aug 19 10:42 mb-1-big-TOC.txt
  • 14. Parts of an SSTable file name Pre 2.2 stuff-simplefields-ia-1-Data.db 2.2 going forward mb-1-big-Data.db Version Counter What it is Keyspace table name Version Counter What it is
  • 15. Data File Size Changed!!! Cassandra 2.1 1044 Aug 19 14:00 stuff-simplefields-ka-1-Data.db Cassandra 3.0 481 Aug 19 10:42 mb-1-big-Data.db
  • 16. A relational view of how the files are connected © DataStax, All Rights Reserved. 16 DATA Summary Index Filter Yes/No Statistics Digest Compression TOC CRC METADATA
  • 17. Metadata • Partitioner • Commit log location • Field names and types • Bloom filter chance • Counter information • Repair time • Cardinality estimate • Checksums • Compression Algorithm • Filenames © DataStax, All Rights Reserved. 17
  • 18. The Test Table and Data
  • 19. My test table CREATE TABLE stuff.simplefields ( field1 text, field2 text, field3 text, field4 timestamp, field5 decimal, PRIMARY KEY ((field1, field2), field3) ) WITH CLUSTERING ORDER BY (field3 ASC) AND compression = {};
  • 20. Some test data to put in it insert into simpleFields (field1, field2, field3, field4, field5) values('FV1', 'FV2' , 'FV3', '2016-08-11 11:37:25.976', '127.00'); insert into simpleFields (field1, field2, field3, field4, field5) values('GV1', 'GV2' , 'GV3', '2016-08-11 11:37:25.976', '127.10'); insert into simpleFields (field1, field2, field3, field4, field5) values('AV1', 'AV2' , 'AV3', '2016-08-11 11:37:25.976', '127.20'); insert into simpleFields (field1, field2, field3, field4, field5) values('BV1', 'BV2' , 'BV3', '2016-08-11 11:37:25.976', '127.30'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV3', '2016-08-11 11:37:25.976', '127.40'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV4', '2016-08-11 11:37:25.976', '127.50'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV5', '2016-08-11 11:37:25.976', '127.60'); insert into simpleFields (field1, field2, field3, field4, field5) values('CV1', 'CV2' , 'CV6', '2016-08-11 11:37:25.976', '127.70');
  • 21. My Data in the Table field1 | field2 | field3 | field4 | field5 --------+--------+--------+-------------------------+-------- CV1 | CV2 | CV3 | 2016-08-11 11:37:25.976 | 127.40 CV1 | CV2 | CV4 | 2016-08-11 11:37:25.976 | 127.50 CV1 | CV2 | CV5 | 2016-08-11 11:37:25.976 | 127.60 CV1 | CV2 | CV6 | 2016-08-11 11:37:25.976 | 127.70 FV1 | FV2 | FV3 | 2016-08-11 11:37:25.976 | 127.00 GV1 | GV2 | GV3 | 2016-08-11 11:37:25.976 | 127.10 BV1 | BV2 | BV3 | 2016-08-11 11:37:25.976 | 127.30 AV1 | AV2 | AV3 | 2016-08-11 11:37:25.976 | 127.20
  • 22. How I got my clusters going Using CCM – Cassandra Cluster Manager https://github.com/pcmanus/ccm Example of creating a cluster: ccm create --vnodes --nodes=1 –v 2.1.9--start ver-219 ccm node1 nodetool status Datacenter: datacenter1 ======================= Status=Up/Down |/ State=Normal/Leaving/Joining/Moving -- Address Load Tokens Owns Host ID Rack UN 127.0.0.1 196.38 KB 256 ? 6ecec1cd-424d-4173-a83f-aefe485f4e2c rack1 © DataStax, All Rights Reserved. 22
  • 23. Clusters currently on my Machine ver-12a ver-219 ver227 ver_33 ver-12a-multi ver-219-single ver-227-single ver_20_11a ver-221 ver-301 test ver-2011-single ver_225 ver-308 ver-120-single ver-213-multi ver_225_6 ver-308-single © DataStax, All Rights Reserved. 23
  • 24. A peek inside the files
  • 25. Quick overview of each File CRC – A single CRC of the Data file Compression – The name of the compression engine used Data – The SSTable data stored in sorted order by Partition/Cluster Key Digest – One or more CRCs one per 2GB segment of the Data file Filter – The bloom filter – folded hashes of all key values in the Data file Index – A list of all partition/Cluster Keys Statistics – Hodgepodge of metadata about the SSTable Summary – A list of some of the Partition/Cluster keys – Think second level index TOC – Text list of the files
  • 26. A peek inside the less interesting files hexdump mb-1-big-CRC.db 0000000 0100 0000 c7ff 8b91 0000008 cat mb-1-big-Digest.crc32 4291269003 cat mb-1-big-TOC.txt Filter.db CRC.db Data.db Digest.crc32 Summary.db TOC.txt Statistics.db Index.db
  • 27. Cassandra 2.1 Index 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 0 0 0000010 0 0 0 0 0 0 0 0 0 0 0 f 0 003 F V 0000020 1 0 0 003 F V 2 0 0 0 0 0 0 0 001 340 0000030 0 0 0 0 0 f 0 003 G V 1 0 0 003 G V 0000040 2 0 0 0 0 0 0 0 002 m 0 0 0 0 0 f 0000050 0 003 B V 1 0 0 003 B V 2 0 0 0 0 0 0000060 0 0 002 372 0 0 0 0 0 f 0 003 A V 1 0 0000070 0 003 A V 2 0 0 0 0 0 0 0 003 207 0 0 0000080 0 0
  • 28. Cassandra 3.0 Index hexdump -c mb-1-big-Index.db 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 0 0 0000010 0 f 0 003 F V 1 0 0 003 F V 2 0 200 312 0000020 0 0 f 0 003 G V 1 0 0 003 G V 2 0 201 0000030 017 0 0 f 0 003 B V 1 0 0 003 B V 2 0 0000040 201 U 0 0 f 0 003 A V 1 0 0 003 A V 2 0000050 0 201 233 0
  • 29. Cassandra 2.1 Summary 0000000 0 0 0 200 0 0 0 001 0 0 0 0 0 0 0 030 0000010 0 0 0 200 0 0 0 001 004 0 0 0 0 003 C V 0000020 1 0 0 003 C V 2 0 0 0 0 0 0 0 0 0 0000030 0 0 0 f 0 003 C V 1 0 0 003 C V 2 0 0000040 0 0 0 f 0 003 A V 1 0 0 003 A V 2 0 0000050 0 004 m m a p 0 0 0 001 0 0 0 0 0 0 0000060 0 0 0 004 m m a p 0 0 0 001 0 0 0 0 0000070 0 0 0 0
  • 30. Cassandra 3.0 Summary hexdump -c mb-1-big-Summary.db 0000000 0 0 0 200 0 0 0 001 0 0 0 0 0 0 0 030 0000010 0 0 0 200 0 0 0 001 004 0 0 0 0 003 C V 0000020 1 0 0 003 C V 2 0 0 0 0 0 0 0 0 0 0000030 0 0 0 f 0 003 C V 1 0 0 003 C V 2 0 0000040 0 0 0 f 0 003 A V 1 0 0 003 A V 2 0
  • 31. Cassandra 3.0 Statistics Part Beginning hexdump -c mb-1-big-Statistics.db 0000000 0 0 0 004 0 0 0 0 0 0 0 $ 0 0 0 001 0000010 0 0 0 Y 0 0 0 002 0 0 0 y 0 0 0 003 0000020 0 0 021 250 0 + o r g . a p a c h e 0000030 . c a s s a n d r a . d h t . M 0000040 u r m u r 3 P a r t i t i o n e 0000050 r ? 204 z 341 G 256 024 { 0 0 0 034 377 377 377 0000060 376 r 031 001 005 256 346 275 001 302 263 326 005 334 361 364 0000070 t 354 241 224 n 324 234 337 001 0 0 0 227 0 0 0 0000080 0 0 0 0 001 0 0 0 0 0 0 0 0 0 0 0
  • 32. Cassandra 3.0 Statistics Tail end 00001230 70 65 29 01 28 6f 72 67 2e 61 70 61 63 68 65 2e |pe).(org.apache.| 00001240 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d 61 72 |cassandra.db.mar| 00001250 73 68 61 6c 2e 55 54 46 38 54 79 70 65 00 02 06 |shal.UTF8Type...| 00001260 66 69 65 6c 64 34 28 6f 72 67 2e 61 70 61 63 68 |field4(org.apach| 00001270 65 2e 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d |e.cassandra.db.m| 00001280 61 72 73 68 61 6c 2e 55 54 46 38 54 79 70 65 06 |arshal.UTF8Type.| 00001290 66 69 65 6c 64 35 28 6f 72 67 2e 61 70 61 63 68 |field5(org.apach| 000012a0 65 2e 63 61 73 73 61 6e 64 72 61 2e 64 62 2e 6d |e.cassandra.db.m| 000012b0 61 72 73 68 61 6c 2e 55 54 46 38 54 79 70 65 |arshal.UTF8Type|
  • 33. Cassandra 2.1 First Part © DataStax, All Rights Reserved. 33 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 177 377 0000010 377 377 200 0 0 0 0 0 0 0 0 t 0 003 C V 0000020 3 0 0 0 0 0 0 005 : p w w [ 341 0 0 0000030 0 0 0 017 0 003 C V 3 0 0 006 f i e l 0000040 d 4 0 0 0 005 : p w w [ 341 0 0 0 027 0000050 2 0 1 6 - 0 8 - 1 1 1 1 : 3 7 0000060 : 2 5 . 9 7 6 0 017 0 003 C V 3 0 0 0000070 006 f i e l d 5 0 0 0 005 : p w w [ 0000080 341 0 0 0 006 1 2 7 . 4 0 0 t 0 003 C 0000090 V 4 0 0 0 0 0 0 005 : p w w n / 0 00000a0 0 0 0 0 017 0 003 C V 4 0 0 006 f i e 00000b0 l d 4 0 0 0 005 : p w w n / 0 0 0 00000c0 027 2 0 1 6 - 0 8 - 1 1 1 1 : 3 00000d0 7 : 2 5 . 9 7 6 0 017 0 003 C V 4 0 00000e0 0 006 f i e l d 5 0 0 0 005 : p w w 00000f0 n / 0 0 0 006 1 2 7 . 5 0 0 t 0 003 0000100 C V 5 0 0 0 0 0 0 005 : p w w 200 261 0000110 0 0 0 0 0 017 0 003 C V 5 0 0 006 f i 0000120 e l d 4 0 0 0 005 : p w w 200 261 0 0 0000130 0 027 2 0 1 6 - 0 8 - 1 1 1 1 :
  • 34. Cassandra 2.1 Second Part © DataStax, All Rights Reserved. 34 0000140 3 7 : 2 5 . 9 7 6 0 017 0 003 C V 5 0000150 0 0 006 f i e l d 5 0 0 0 005 : p w 0000160 w 200 261 0 0 0 006 1 2 7 . 6 0 0 t 0 0000170 003 C V 6 0 0 0 0 0 0 005 : p w w 213 0000180 a 0 0 0 0 0 017 0 003 C V 6 0 0 006 f 0000190 i e l d 4 0 0 0 005 : p w w 213 a 0 00001a0 0 0 027 2 0 1 6 - 0 8 - 1 1 1 1 00001b0 : 3 7 : 2 5 . 9 7 6 0 017 0 003 C V 00001c0 6 0 0 006 f i e l d 5 0 0 0 005 : p 00001d0 w w 213 a 0 0 0 006 1 2 7 . 7 0 0 0 00001e0 0 f 0 003 F V 1 0 0 003 F V 2 0 177 377 00001f0 377 377 200 0 0 0 0 0 0 0 0 t 0 003 F V 0000200 3 0 0 0 0 0 0 005 : p w v 271 d 0 0 0000210 0 0 0 017 0 003 F V 3 0 0 006 f i e l 0000220 d 4 0 0 0 005 : p w v 271 d 0 0 0 027 0000230 2 0 1 6 - 0 8 - 1 1 1 1 : 3 7 0000240 : 2 5 . 9 7 6 0 017 0 003 F V 3 0 0 0000250 006 f i e l d 5 0 0 0 005 : p w v 271 0000260 d 0 0 0 006 1 2 7 . 0 0 0 0 0 f 0
  • 35. Cassandra 2.1 Third Part © DataStax, All Rights Reserved. 35 0000270 003 G V 1 0 0 003 G V 2 0 177 377 377 377 200 0000280 0 0 0 0 0 0 0 0 t 0 003 G V 3 0 0 0000290 0 0 0 0 005 : p w w : 255 0 0 0 0 0 00002a0 017 0 003 G V 3 0 0 006 f i e l d 4 0 00002b0 0 0 005 : p w w : 255 0 0 0 027 2 0 1 00002c0 6 - 0 8 - 1 1 1 1 : 3 7 : 2 5 00002d0 . 9 7 6 0 017 0 003 G V 3 0 0 006 f i 00002e0 e l d 5 0 0 0 005 : p w w : 255 0 0 00002f0 0 006 1 2 7 . 1 0 0 0 0 f 0 003 B V 0000300 1 0 0 003 B V 2 0 177 377 377 377 200 0 0 0 0000310 0 0 0 0 0 t 0 003 B V 3 0 0 0 0 0 0000320 0 005 : p w w Q 234 0 0 0 0 0 017 0 003 0000330 B V 3 0 0 006 f i e l d 4 0 0 0 005 0000340 : p w w Q 234 0 0 0 027 2 0 1 6 - 0 0000350 8 - 1 1 1 1 : 3 7 : 2 5 . 9 7 0000360 6 0 017 0 003 B V 3 0 0 006 f i e l d 0000370 5 0 0 0 005 : p w w Q 234 0 0 0 006 1 0000380 2 7 . 3 0 0 0 0 f 0 003 A V 1 0 0 0000390 003 A V 2 0 177 377 377 377 200 0 0 0 0 0 0
  • 36. Cassandra 3.0 Part A © DataStax, All Rights Reserved. 36 0000000 0 f 0 003 C V 1 0 0 003 C V 2 0 177 377 0000010 377 377 200 0 0 0 0 0 0 0 $ 0 003 C V 3 0000020 $ 032 270 221 b 027 2 0 1 6 - 0 8 - 1 1 0000030 1 1 : 3 7 : 2 5 . 9 7 6 b 006 1 0000040 2 7 . 4 0 $ 0 003 C V 4 % + 300 D a 0000050 b 027 2 0 1 6 - 0 8 - 1 1 1 1 : 0000060 3 7 : 2 5 . 9 7 6 b 006 1 2 7 . 5 0000070 0 $ 0 003 C V 5 % , 300 N 234 b 027 2 0 0000080 1 6 - 0 8 - 1 1 1 1 : 3 7 : 2 0000090 5 . 9 7 6 b 006 1 2 7 . 6 0 $ 0 003 00000a0 C V 6 % , 300 [ ` b 027 2 0 1 6 - 0 00000b0 8 - 1 1 1 1 : 3 7 : 2 5 . 9 7 00000c0 6 b 006 1 2 7 . 7 0 001 0 f 0 003 F V 00000d0 1 0 0 003 F V 2 0 177 377 377 377 200 0 0 0 00000e0 0 0 0 0 $ 0 003 F V 3 # 032 0 b 027 2 00000f0 0 1 6 - 0 8 - 1 1 1 1 : 3 7 : 0000100 2 5 . 9 7 6 b 006 1 2 7 . 0 0 001 0 0000110 f 0 003 G V 1 0 0 003 G V 2 0 177 377 377
  • 37. Cassandra 3.0 Part B © DataStax, All Rights Reserved. 37 0000120 377 200 0 0 0 0 0 0 0 $ 0 003 G V 3 $ 0000130 032 221 346 b 027 2 0 1 6 - 0 8 - 1 1 0000140 1 1 : 3 7 : 2 5 . 9 7 6 b 006 1 2 0000150 7 . 1 0 001 0 f 0 003 B V 1 0 0 003 B 0000160 V 2 0 177 377 377 377 200 0 0 0 0 0 0 0 $ 0000170 0 003 B V 3 $ 032 255 036 b 027 2 0 1 6 - 0000180 0 8 - 1 1 1 1 : 3 7 : 2 5 . 9 0000190 7 6 b 006 1 2 7 . 3 0 001 0 f 0 003 A 00001a0 V 1 0 0 003 A V 2 0 177 377 377 377 200 0 0 00001b0 0 0 0 0 0 $ 0 003 A V 3 $ 032 240 > b 00001c0 027 2 0 1 6 - 0 8 - 1 1 1 1 : 3 00001d0 7 : 2 5 . 9 7 6 b 006 1 2 7 . 2 0 00001e0 001
  • 38. Bloom Filter contents 2.1 and 3.0 hexdump stuff-simplefields-ka-1-Filter.db 0000000 0000 0500 0000 0200 0032 2090 3002 0009 0000010 1803 0110 0022 40c0 hexdump mb-1-big-Filter.db 0000000 0000 0500 0000 0200 1808 5068 8822 0019 0000010 1012 1100 0002 4180 © DataStax, All Rights Reserved. 38
  • 40. Conclusions • Cassandra 3.0 SSTable format is considerably more compact • In Cassandra 3.0 field names are no longer stored in the data file field Ids are used instead • TTL and Timestamps are grouped • TTL and Timestamp Differences are stored as deltas • Sorted String Tables are sorted…. By Partition Key Token Value • The Statistics file is full of a bunch of different stuff • Data, Summary, Filter and Statistics have important useful stuff © DataStax, All Rights Reserved. 40
  • 42. QUESTIONS? © DataStax, All Rights Reserved. 42
  • 43. In version "ka", contents of Statistics component file is split in to 3 types of metadata called validation, compaction and stats. Validation metadata is used to validate SSTable which includes partitioner and bloom filter fp chance fields. Compaction metadata includes ancestors information which is also available in older formats and a new field called cardinality estimator. Cardinality estimator is used to efficiently pre-allocate bloom filter space in a merged compaction file by estimating how much the input SStables overlap. Stats metadata contains rest of the information available in old formats and two additional fields. First one is a flag to track the presence of local/remote counter shards and the other one is for storing the repair time. © DataStax, All Rights Reserved. 43