Diese Präsentation wurde erfolgreich gemeldet.
Wir verwenden Ihre LinkedIn Profilangaben und Informationen zu Ihren Aktivitäten, um Anzeigen zu personalisieren und Ihnen relevantere Inhalte anzuzeigen. Sie können Ihre Anzeigeneinstellungen jederzeit ändern.
NoSQL Matters
@adriancolyer
1. when choosing a data store / processing
platform
2. when it comes to getting the most out of that
platform
3. when we t...
The 13 horsemen of the apocalypse...
Your application(s)
Anomaly (Prevented By) Tolerable? Mitigation (M,G,A…)
Dirty Write...
Your application(s)
Anomaly (Prevented By) Tolerable? Mitigation
Read Skew MAV Isolation +
item-cut
Lost Update Repeatable...
Your application(s)
Anomaly (Prevented By) Tolerable? Mitigation
Non-monotonic
read
Monotonic reads
Non-monotonic
write
Mo...
Your Developers
“we believe there is considerable work to be
done to improve the programmability of highly-
available syst...
Your Developers
“...an unacceptable burden to place on
developers” - Google 2012 (F1)
Consistency and all that...
If you accept a weaker consistency model
make sure it’s a genuine trade-off and you’re
getting...
PACELC (pass-elk)
Operations & all the other use cases
…it is important to consider the data accesses that don’t
use the API. These include ...
it tears you apart
with suspense!
“
”
Why is it so hard?
“We have found that the standard verification techniques in
industry are necessary but not sufficient. ...
In the ALPS...
… or a walk in the park?
(Web)Scale
The USL
Source : McSherry et al. 2015
Credit: Neil Gunther
(Web)Scale
Source : McSherry et al. 2015
Big?!
How Big?
“Working sets are Zipf-distributed. We can therefore store
in memory all but the very largest datasets, which we ...
Musketeer
Performance
40-80% of all MR jobs would perform
better on a single machine!
(and cost less, and be easier to
operate, and ...
COST
The Configuration that Outperforms a Single
Thread
“You can have a second computer
once you’ve shown you know how
to ...
vs a single thread...
FlashGraph vs Pregel
● Pregel: 1B vertices, 127B edges, 300
machines
● FlashGraph: 3.4B vertices, 129B edges, 1
machine
ApproxHadoop
BlinkDB
Sometimes it pays to wait (a little bit)
What’s the bottleneck?
● Network I/O?
● Disk I/O?
● CPU?
Measure before optimising… and avoid
excessive serialization and ...
X (multi-core)
Distributed X
In-memory X
Flash Optimised X
NVMM X
NVMM & RDMA X
X (establish baseline COST)
ALPS, ACID 2.0,
CRDTs, CAC,
COPS, CRON,
CALM, CAP, &
CRAP!
Coordination Avoidance
Invariant-Confluence for application level
constraints
● NOT NULL
● PRIMARY KEY (read & delete, but...
Life Beyond...
“In recent years, many ‘NoSQL’ designs have avoided
cross-partition transactions entirely, effectively prov...
Your application(s)
From anomalies to invariants...
Invariant Type Affected Txns I-Confluent?
Some closing thoughts
● Do you need eventual?
● Have you planned for anomalies?
● Does it actually work?
● Are you distrib...
http://blog.acolyer.org @adriancolyer
References
● Highly Available Transactions, Virtues & Limitations - Bailis et al. 2014 http:
//blog.acolyer.org/2014/11/07...
References
● Don’t Settle for Eventual: Scalable Causal Consistency for Wide-Area
Storage with COPS - LLoyd et al. 2011 ht...
References
● Musketeer: all for one, one for all in data processing systems - Gog et al.
2015 http://blog.acolyer.org/2015...
References
● ApproxHadoop: Bringing Approximations to Hadoop Frameworks - Goiri
2015 http://blog.acolyer.org/2015/04/16/ap...
References
● The Declarative Imperative: Experiences and Conjectures in Distributed
Logic - Hellerstein 2010 http://blog.a...
References
● Consistency Analysis in Bloom: A Calm and Collected Approach - Alvaro et
al. 2011 http://blog.acolyer.org/201...
References
● Coordination Avoidance in Database Systems - Bailis et al. 2014 http:
//blog.acolyer.org/2015/03/19/coordinat...
References
● Life Beyond Distributed Transactions - Helland 2007 http://blog.acolyer.
org/2014/11/20/life-beyond-distribut...
Image Credits
● ALPS + Dublin Park: Wikimedia Commons
● Movies: IMDB
● Monotone Commuters: http://www.yenko.net/ubbthreads...
Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015
Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015
Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015
Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015
Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015
Nächste SlideShare
Wird geladen in …5
×

Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015

736 Aufrufe

Veröffentlicht am

NoSQL matters, on that much I'm sure we can all agree. But if we take a closer look, what really matters when it comes to choosing a data store and/or a data processing platform? What really matters when it comes to getting the most out of that platform? And what is really going to matter as we take things to the next level?

Veröffentlicht in: Daten & Analysen
  • Als Erste(r) kommentieren

Adrian Colyer - Keynote: NoSQL matters - NoSQL matters Dublin 2015

  1. 1. NoSQL Matters @adriancolyer
  2. 2. 1. when choosing a data store / processing platform 2. when it comes to getting the most out of that platform 3. when we take things to the next level What really matters...
  3. 3. The 13 horsemen of the apocalypse... Your application(s) Anomaly (Prevented By) Tolerable? Mitigation (M,G,A…) Dirty Writes Read Uncommitted Dirty Reads Read Committed Fuzzy Reads (non-repeat- able) Item-Cut Isolation Phantoms Predicate-Cut Isolation ...
  4. 4. Your application(s) Anomaly (Prevented By) Tolerable? Mitigation Read Skew MAV Isolation + item-cut Lost Update Repeatable Read Cursor Lost Update Cursor Stability Write Skew Repeatable Read Stale Reads Partition-intolerance
  5. 5. Your application(s) Anomaly (Prevented By) Tolerable? Mitigation Non-monotonic read Monotonic reads Non-monotonic write Monotonic writes Invisible cause Writes-follow-reads Disappearing writes Read-your-writes (for sessions)
  6. 6. Your Developers “we believe there is considerable work to be done to improve the programmability of highly- available systems” - Bailis et al. 2014 (HAT)
  7. 7. Your Developers “...an unacceptable burden to place on developers” - Google 2012 (F1)
  8. 8. Consistency and all that... If you accept a weaker consistency model make sure it’s a genuine trade-off and you’re getting something (you need) in return. You can have causal consistency with (C)AC
  9. 9. PACELC (pass-elk)
  10. 10. Operations & all the other use cases …it is important to consider the data accesses that don’t use the API. These include back-ups, bulk import and deletion of data, bulk migrations from one data format to another, replica creation, asynchronous replication, consistency monitoring tools, and operational debugging. An alternate store would also have to provide atomic write transactions, efficient granular writes, and few latency outliers. - Facebook 2013 (TAO) “ ”
  11. 11. it tears you apart with suspense! “ ”
  12. 12. Why is it so hard? “We have found that the standard verification techniques in industry are necessary but not sufficient. We use deep design reviews, code reviews, static code analysis, stress testing, fault-injection testing, and many other techniques, but we still find that subtle bugs can hide in complex concurrent fault-tolerant systems.” - Amazon 2014
  13. 13. In the ALPS...
  14. 14. … or a walk in the park?
  15. 15. (Web)Scale The USL Source : McSherry et al. 2015 Credit: Neil Gunther
  16. 16. (Web)Scale Source : McSherry et al. 2015
  17. 17. Big?!
  18. 18. How Big? “Working sets are Zipf-distributed. We can therefore store in memory all but the very largest datasets, which we avoid storing in memory altogether. For example, the distribution of input sizes of MapReduce jobs at Facebook is heavy- tailed. Furthermore, 96% of active jobs can have their entire data simultaneously fit in the corresponding clusters’ memory” - Tachyon, Lie et al. 2014
  19. 19. Musketeer
  20. 20. Performance 40-80% of all MR jobs would perform better on a single machine! (and cost less, and be easier to operate, and have many fewer failures…)
  21. 21. COST The Configuration that Outperforms a Single Thread “You can have a second computer once you’ve shown you know how to use the first one.” - Paul Barham
  22. 22. vs a single thread...
  23. 23. FlashGraph vs Pregel ● Pregel: 1B vertices, 127B edges, 300 machines ● FlashGraph: 3.4B vertices, 129B edges, 1 machine
  24. 24. ApproxHadoop
  25. 25. BlinkDB
  26. 26. Sometimes it pays to wait (a little bit)
  27. 27. What’s the bottleneck? ● Network I/O? ● Disk I/O? ● CPU? Measure before optimising… and avoid excessive serialization and deserialization!
  28. 28. X (multi-core) Distributed X In-memory X Flash Optimised X NVMM X NVMM & RDMA X X (establish baseline COST)
  29. 29. ALPS, ACID 2.0, CRDTs, CAC, COPS, CRON, CALM, CAP, & CRAP!
  30. 30. Coordination Avoidance Invariant-Confluence for application level constraints ● NOT NULL ● PRIMARY KEY (read & delete, but not insert) ● UNIQUE (read & delete, insert?) ● FOREIGN KEY (insert, cascade delete, but delete)
  31. 31. Life Beyond... “In recent years, many ‘NoSQL’ designs have avoided cross-partition transactions entirely, effectively providing Read Uncommitted isolation…” - Bailis et al. 2014 From: “Life Beyond Distributed Transactions”, To: “Read-Atomic Multiple Partition” Transactions (RAMP)
  32. 32. Your application(s) From anomalies to invariants... Invariant Type Affected Txns I-Confluent?
  33. 33. Some closing thoughts ● Do you need eventual? ● Have you planned for anomalies? ● Does it actually work? ● Are you distributing for the right reasons? (AL…) ● Do you need exact? ● Do you need it ASAP? ● Can you keep CALM? ● Do you understand your application’s invariants?
  34. 34. http://blog.acolyer.org @adriancolyer
  35. 35. References ● Highly Available Transactions, Virtues & Limitations - Bailis et al. 2014 http: //blog.acolyer.org/2014/11/07/highly-available-transactions-virtues-and- limitations/ ● Building on Quicksand - Helland 2009 http://blog.acolyer. org/2015/03/23/building-on-quicksand/ ● F1: A Distributed SQL Database that Scales - Google 2012 http://blog. acolyer.org/2015/01/06/f1-a-distributed-sql-database-that-scales/ ● Scalability! But at what COST? - McSherry et al. 2015 http://blog.acolyer. org/?p=941 (to appear, June 5th 2015) ● Applying the Universal Scalability Law to Organisations - Colyer 2015 http: //blog.acolyer.org/2015/04/29/applying-the-universal-scalability-law-to- organisations/
  36. 36. References ● Don’t Settle for Eventual: Scalable Causal Consistency for Wide-Area Storage with COPS - LLoyd et al. 2011 http://blog.acolyer. org/2015/03/17/consistency-availability-and-convergence-cops/ ● Consistency, Availability, and Convergence - Mahajan et al. 2014 http: //blog.acolyer.org/2015/03/17/consistency-availability-and-convergence- cops/ ● Tachyon: Reliable, Memory-Speed Storage for Cluster Computing - Lie et al. 2014 http://blog.acolyer.org/2014/12/04/tachyon-reliable-memory- speed-storage-for-cluster-computing/
  37. 37. References ● Musketeer: all for one, one for all in data processing systems - Gog et al. 2015 http://blog.acolyer.org/2015/04/27/musketeer-part-i-whats-the-best- data-processing-system/ and http://blog.acolyer. org/2015/04/28/musketeer-part-ii-one-for-all-and-all-for-one/ ● Pregel: A System for Large-Scale Graph Processing - Google 2010 http: //blog.acolyer.org/2015/05/26/pregel-a-system-for-large-scale-graph- processing/ ● FlashGraph: Processing Billion Node Graphs on an array of commodity SSDs - Zheng et al. 2015 http://blog.acolyer.org/?p=935
  38. 38. References ● ApproxHadoop: Bringing Approximations to Hadoop Frameworks - Goiri 2015 http://blog.acolyer.org/2015/04/16/approxhadoop-bringing- approximations-to-mapreduce-frameworks/ ● BlinkDB: http://blinkdb.org/ ● Making Sense of Performance in Data Analytics Frameworks - Ousterhout et al 2015 http://blog.acolyer.org/2015/04/20/making-sense-of- performance-in-data-analytics-frameworks/ ● A Comprehensive Study of Convergent and Commutative Replicated Data Types - Shapiro et al. 2011 http://blog.acolyer.org/2015/03/18/a- comprehensive-study-of-convergent-and-commutative-replicated-data- types/
  39. 39. References ● The Declarative Imperative: Experiences and Conjectures in Distributed Logic - Hellerstein 2010 http://blog.acolyer.org/2014/11/13/the-declarative- imperative-experiences-and-conjectures-in-distributed-logic/ ● Fast Remote Memory - Dragojevic et al. 2014 http://blog.acolyer. org/2015/05/20/farm-fast-remote-memory/ ● Mojim: A Reliable and Highly-Available Non-Volatile Memory System - Zhang et al. 2015 http://blog.acolyer.org/2015/04/14/mojim-a-reliable-and- highly-available-non-volatile-memory-system/
  40. 40. References ● Consistency Analysis in Bloom: A Calm and Collected Approach - Alvaro et al. 2011 http://blog.acolyer.org/2015/03/16/consistency-analysis-in-bloom- a-calm-and-collected-approach/ ● Edelweiss: Automatic Storage Reclamation for Distributed Programming - Conway et al. 2014 http://blog.acolyer.org/2015/02/20/edelweiss- automatic-storage-reclamation-for-distributed-programming/ ● Scalable Atomic Visibility with RAMP Transactions - Bailis et al. 2014 http: //blog.acolyer.org/2015/03/27/scalable-atomic-visibility-with-ramp- transactions/
  41. 41. References ● Coordination Avoidance in Database Systems - Bailis et al. 2014 http: //blog.acolyer.org/2015/03/19/coordination-avoidance-in-database- systems/ ● Putting Consistency Back into Eventual Consistency - Balegas et al. 2015 http://blog.acolyer.org/2015/05/04/putting-consistency-back-into-eventual- consistency/ ● Use of Formal Methods at Amazon Web Services - Newcombe et al. 2014 http://blog.acolyer.org/2014/11/24/use-of-formal-methods-at-amazon-web- services/ ● Consistency Trade-offs in Modern Distributed Database Systems Design - Abadi 2012 http://cs-www.cs.yale.edu/homes/dna/papers/abadi-pacelc.pdf
  42. 42. References ● Life Beyond Distributed Transactions - Helland 2007 http://blog.acolyer. org/2014/11/20/life-beyond-distributed-transactions/
  43. 43. Image Credits ● ALPS + Dublin Park: Wikimedia Commons ● Movies: IMDB ● Monotone Commuters: http://www.yenko.net/ubbthreads/ubbthreads. php/topics/312207/re-old-street-scenes ● Elk picture by Jim Richmond: http://commons.wikimedia.org/wiki/File:Rm- elk-locking-antlers.jpg

×