![Poll: what does HashSet<K> use under the covers?
A. K[]
B. Entry<K>[]
C. HashMap<K,V>
D. TreeMap<K,V>](https://image.slidesharecdn.com/javaone-2015-avoiding-bigdata-antipatterns-151028035003-lva1-app6891/85/Avoiding-big-data-antipatterns-71-320.jpg)
Ad
More Related Content
What's hot (20)
Viewers also liked (9)
Ad
Similar to Avoiding big data antipatterns (20)
![[DBA]_HiramFleitas_SQL_PASS_Summit_2017_Summary](https://cdn.slidesharecdn.com/ss_thumbnails/dbahiramfleitassqlpasssummit2017summary-180202201153-thumbnail.jpg?width=560&fit=bounds)
Ad
Recently uploaded (20)
Avoiding big data antipatterns
- 1. Avoiding big data anti-patterns
- 2. whoami • Alex Holmes • Software engineer • @grep_alex • grepalex.com
- 3. Why should I care about anti-patterns? big data ^
- 5. Agenda • Walking through anti-patterns • Looking at why they should be avoided • Consider some mitigations • It’s big data! • A single tool for the job • Polyglot data integration • Full scans FTW! • Tombstones • Counting with Java built-in collections • It’s open I’ll cover: by …
- 6. Meet your protagonists Alex Jade (the amateur) (the pro)
- 8. i want to calculate some statistics on some static user data . . . how big is the data? it’s big data, so huge, 20GB!!! i need to order a hadoop cluster!
- 10. you think you have big data ... but you don’t!
- 13. Poll: how much RAM can a single server support? A. 256 GB B. 512 GB C. 1TB
- 16. keep it simple . . . use MYSQL or POSTGRES or R/python/matlab
- 17. Summary • Simplify your analytics toolchain when working with small data (especially if you don’t already have an investment in big data tooling) • Old(er) tools such as OLTP/OLAP/R/Python still have their place for this type of work
- 18. A single tool for the job
- 19. that looks like a nail!!! nosql PROBLEM
- 21. Big data tools are usually designed to do one thing well (maybe two) ^
- 22. Types of workloads • Low-latency data lookups • Near real-time processing • Interactive analytics • Joins • Full scans • Search • Data movement and integration • ETL
- 23. The old world was simple OLTP/OLAP
- 24. The new world … not so much
- 25. You need to research and find the best-in-class for your function
- 26. Best-in-class big data tools (in my opinion) If you want … Consider … Low-latency lookups Cassandra, memcached Near real-time processing Storm Interactive analytics Vertica, Teradata Full scans, system of record data, ETL, batch processing HDFS, MapReduce, Hive, Pig Data movement and integration Kafka
- 27. Summary • There is no single big data tool that does it all • We live in a polyglot world where new tools are announced every day - don’t believe the hype! • Test the claims on your own hardware and data; start small and fail fast
- 29. i need to move clickstream data from my application to hadoop Application Hadoop Loader Hadoop
- 30. shoot, i need to use that same data in streaming Application Hadoop Loader Hadoop JMS JMS
- 33. that way new consumers can be added without any work! we need a central data repository and pipeline to isolate consumers from the source let’s use kafka! OLTP OLAP / EDW HBase Cassan dra Voldem ort Hadoop SecurityAnalytics Rec. Engine Search Monitoring Social Graph kafka
- 34. Background • Apache project • Originated from LinkedIn • Open-sourced in 2011 • Written in Scala and Java • Borrows concepts in messaging systems and logs • Foundational data movement and integration technology
- 35. What’s the big whoop about Kafka?
- 36. Throughput http://engineering.linkedin.com/kafka/benchmarking-apache-kafka-2-million-writes-second-three-cheap-machines Producer Consumer Consumer Consumer Kafka Server Kafka Server Kafka Server Producer Producer 2,024,032 TPS 2,615,968 TPS 2ms
- 37. O.S. page cache is leveraged OS page cache ...11109876543210 ProducerConsumer A Consumer B Disk writesreadsreads
- 38. Things to look out for • Leverages ZooKeeper, which is tricky to configure • Reads can become slow when the page cache is missed and disk needs to be hit • Lack of security
- 39. Summary • Don’t write your own data integration • Use Kafka for light-weight, fast and scalable data integration
- 40. Full scans FTW!
- 41. i heard that hadoop was designed to work with huge data volumes!
- 42. i’m going to stick my data on hadoop . . .
- 43. and run some joins SELECT * FROM huge_table JOIN ON other_huge_table …
- 45. yes, hadoop is very efficient at batch workloads files are split into large blocks and distributed throughout the cluster “data locality” is a first-class concept, where compute is pushed to storage Scheduler
- 46. but hadoop doesn’t negate all these optimizations we learned when working on relational databases
- 47. so partition your data according to how you will most commonly access it hdfs:/data/tweets/date=20140929/ hdfs:/data/tweets/date=20140930/ hdfs:/data/tweets/date=20140931/ disk io is slow
- 48. and then make sure to include a filter in your queries so that only those partitions are read ... WHERE DATE=20151027
- 49. include projections to reduce data that needs to be read from disk or pushed over the network SELECT id, name FROM...
- 50. hash joins require network io which is slow
- 51. 65.23VRSN 39.54MSFT 526.62GOOGL PriceSymbol RestonVRSN RedmondMSFT Mtn ViewGOOGL HeadquartersSymbol merge joins are way more efficient Records in all datasets sorted by join key Headquarters Redmond Mtn View Reston65.23VRSN 39.54MSFT 526.62GOOGL PriceSymbol The merge algorithm streams and performs an inline merge of the datasets
- 52. you’ll have to bucket and sort your data and tell your query engine to use a sort- merge-bucket (SMB) join — Hive properties to enable a SMB join set hive.auto.convert.sortmerge.join=true; set hive.optimize.bucketmapjoin = true; set hive.optimize.bucketmapjoin.sortedmerge = true;
- 53. and look at using a columnar data format like parquet Column Strorage Column 1 (Symbol) 526.62 MSFT 05-10-2014 05-10-2014 39.54 GOOGL Column 2 (Date) Column 3 (Price)
- 54. Summary • Partition, filter and project your data (same as you used to do with relational databases) • Look at bucketing and sorting your data to support advanced join techniques such as sort-merge- bucket • Consider storing your data in columnar form
- 55. Tombstones
- 56. i need to store data in a Highly Available persistent queue . . . and we already have Cassandra deployed . . . bingo!!!
- 57. What is Cassandra? • Low-latency distributed database • Apache project modeled after Dynamo and BigTable • Data replication for fault tolerance and scale • Multi-datacenter support • CAP
- 60. V VVV V VVVVV V KKKKKKKKKKK V VVV V VVVVV V KKKKKKKKKKK tombstone markers indicate that the column has been deleted deletes in Cassandra are soft; deleted columns are marked with tombstones these tombstoned columns slow-down reads
- 61. by default tombstones stay around for 10 days IF you want to know why, read up on gc_grace_secs, and reappearing deletes
- 62. don’t use Cassandra, use kafka
- 63. design your schema and read patterns to avoid tombstones getting in the way of your reads
- 64. keep track of consumer offsets, and add a time or bucket semantic to rows. only delete rows after some time has elapsed, or once all consumers have consumed them. ID bucket 1 2 2 1 msg 81583 723804 offset 723803 81582 msg 81582 723803 msg 81581 723802 bucket bucket consumer consumer . . . . . . ID 1 2
- 65. Summary • Try to avoid use cases that require high volume deletes and slice queries that scan over tombstone columns • Design your schema and delete/read patterns with tombstone avoidance in mind
- 66. Counting with Java’s built-in collections
- 67. I’m going to COUNT THE DISTINCT NUMBER OF USERS THAT viewed a tweet
- 71. Poll: what does HashSet<K> use under the covers? A. K[] B. Entry<K>[] C. HashMap<K,V> D. TreeMap<K,V>
- 73. Memory consumption HashSet = 32 * SIZE + 4 * CAPACITY String = 8 * (int) ((((no chars) * 2) + 45) / 8) Average user is 6 characters long = 64 bytes number of elements in set set capacity (array length) For 10,000,000 users this is at least 1GiB
- 74. USE HYPERLOGLOG TO WORK WITH approximate DISTINCt COUNTS @SCALE
- 75. HyperLogLog • Cardinality estimation algorithm • Uses (a lot) less space than sets • Doesn’t provide exact distinct counts (being “close” is probably good enough) • Cardinality Estimation for Big Data: http://druid.io/ blog/2012/05/04/fast-cheap-and-98-right- cardinality-estimation-for-big-data.html
- 76. 1 billion distinct elements = 1.5kb memory standard error = 2%
- 78. Hashes 10110100100101010101100111001011 Good hash functions should result in each bit having a 50% probability of occurring h(entity):
- 79. Bit pattern observations 1xxxxxxxxx..x 01xxxxxxxx..x 001xxxxxxx..x 0001xxxxxx..x 50% of hashed values will look like: 25% of hashed values will look like: 12.5% of hashed values will look like: 6.25% of hashed values will look like:
- 80. 0 0 0 0 0 0 0 0 0 00 0 0 0 00 000 0 0 00 0 register
- 81. 0 0 0 0 0 0 0 0 0 00 0 0 0 00 000 0 0 00 0 h(entity): register index: 4 register value: 1 register harmonic_mean ( HLL estimated cardinality = (= 1 01010 01 0 0 01 1
- 82. HLL Java library • https://github.com/aggregateknowledge/java-hll • Neat implementation - it automatically promotes internal data structure to HLL once it grows beyond a certain size
- 83. Approximate count algorithms • HyperLogLog (distinct counts) • CountMinSketch (frequencies of members) • Bloom Filter (set membership)
- 84. Summary • Data skew is a reality when working at Internet scale • Java’s builtin collections have a large memory footprint don’t scale • For high-cardinality data use approximate estimation algorithms
- 85. stepping away . . . MATH
- 86. It’s open
- 87. prototyping/VIABILITY - DONE coding - done testing - done performance & scalability testing - done monitoring - done i’m ready to ship!
- 92. How the old world worked OLTP/OLAP Authentication Authorization DBA
- 93. security’s not my job!!! we disagree infosec
- 94. Important questions to ask • Is my data encrypted when it’s in motion? • Is my data encrypted on disk? • Are there ACL’s defining who has access to what? • Are these checks enabled by default?
- 95. How do tools stack up? ACL’s At-rest encryption In-motion encryption Enabled by default Ease of use Oracle Hadoop Cassandra ZooKeeper Kafka
- 96. Summary • Enable security for your tools! • Include security as part of evaluating a tool • Ask vendors and project owners to step up to the plate
- 97. We’re done!
- 98. Conclusions • Don’t assume that a particular big data technology will work for your use case - verify it for yourself on your own hardware and data early on in the evaluation of a tool • Be wary of the “new hotness” and vendor claims - they may burn you • Make sure that load/scale testing is a required part of your go-to-production plan
- 99. Thanks for your time!