Neo4j and the JVM

This talk is mostly about using the Neo4j Java Driver, displays the different types of sessions as well as transaction types that can be used.

It was originally given at the R&D Day at GraphAware: https://graphaware.com

Transcript

1. Neo4j and the JVM Michael Simons, 16th Dezember 2020 The manifold ways of connecting to Neo4j from the JVM
2. Michael Simons @rotnroll666 About me „Michael Simons of Spring Data Neo4j/Reactive/Quarkus/GraalVM fame“ • This ^^ is a quote (and the pressure is high) • Official role: Maintainer of Neo4j-OGM and Spring Data Neo4j • We released Neo4j-OGM 3.2 in 2019, with 20 patch releases following • Not many changes in Spring Data Neo4j 5.x • Changed most everything in Spring Data Neo4j 6
3. Michael Simons @rotnroll666 About me • (:Team {name: "OGM / SDN"}) - [:IS_PART_OF] -> (:Team {name: "Drivers"}) • Very close feedback loop • Happy to test beta code • Nice position between ! users on the one hand, drivers and database on the other "
4. Michael Simons @rotnroll666 About me The rest… • # • $, % and & (no swimming this year… ') • Twitter addict… @rotnroll666 • Enjoys databases with declarative languages in general • Active in both Java and database communities • Likes Spring (wrote a best selling German book about it) • But enjoys other ecosystems too… (Not actually evil, though…)
5. Topics today Where did I use the Neo4j Java Driver?
6. Neo4j-OGM The elephant in the room
7. Michael Simons @rotnroll666 Neo4j-OGM Multiple transport modes, why are they causing issues? • Embedded • Dependency management is a nightmare • Bringing up the database from the mapper is the wrong abstraction layer • Close coupling between OGM and Neo4j versions • Definitely not native ready • HTTP • Not typesafe (we are working on that) • Bolt (Aka the Neo4j Java Driver) • Mostly naming, people mistake bolt-driver with the Java driver itself • Transactional functions (to some extend, later more)
8. Michael Simons @rotnroll666 Neo4j-OGM Spring Data Neo4j 5.x without reactive support. ( • No (sane) way to add a shared abstraction over all modes • Just wrapping lists in reactive datatypes isn’t gonna cut it
9. On the plus side
10. Michael Simons @rotnroll666 Neo4j-OGM On the plus side • More than 20 releases in the last 2 years • Ticket numbers closed to about 4 or 5 • Stable API, ways to mark queries explicitly as r/w and r/o, supports multi database • Native ready since this month • Usage of bolt taught us it’s API and usage • Super flexible mapping, both a benefit and sometimes a PITA • Personal note: I do like maintaining it.
11. Two German Java developers just before Graph Tour 2019 in London „Hmm… Lightweight Spring Data JDBC is cool, should we try this with SDN?“ )
12. Michael Simons @rotnroll666 Goals of SDN⚡RX These days: Spring Data Neo4j 6 • Full support of the reactive story • • • • Neo4j native types • Dedicated ways of decomposing objects into maps to be stored on single nodes / relationships • Clean levels of abstracts (from low to high: Managed driver, Neo4j client, template, repositories, from client onwards with Spring TX Support) • • Compatible with GraalVM native • • Full support of immutable domain objects (Kotlin data classes, Java 16 records, Lombok Data classes) • Full support of all derived finder query building parts • Full support of find by example • • • • Performance •
13. Michael Simons @rotnroll666 (:Topic) - [:RELATED_TO {type: "directly"}] -> (the:Driver) Goals of SDN⚡RX • Full support of the reactive story • Neo4j native types • Dedicated ways of decomposing objects into maps to be stored on single nodes / relationships • Clean levels of abstracts (from low to high: Managed driver, Neo4j client, template, repositories, from client onwards with Spring TX Support) • Compatible with GraalVM native
14. Michael Simons @rotnroll666 Brought in order for today • Full support of the reactive story • Clean levels of abstracts (from low to high: Managed driver, Neo4j client, template, repositories, from client onwards with Spring TX Support) • Neo4j native types • Compatible with GraalVM native
15. Michael Simons @rotnroll666 Session and transaction APIs The big all-in-one-package • org.neo4j.driver.GraphDatabase is the general Driver factory • Instances of org.neo4j.driver.Driver are heavyweight objects and maintain a pool of physical connections • Open as few as possible and keep them around • All sessions are lightweight and basically a lease on connections • Don’t keep them around • Open and close as needed • Concepts are not 1:1 comparable to JDBC wording!
16. Michael Simons @rotnroll666 Session and transaction APIs What you want application wide is as few instances of a driver as possible.
17. Michael Simons @rotnroll666 Session and transaction APIs All major Java platforms do support something like this now
18. Michael Simons @rotnroll666 Session and transaction APIs Reduce driver footprint • i.e. in scenarios where there is more than instance needed (different credentials, different clusters etc.) • Scripting
19. Michael Simons @rotnroll666 Session and transaction APIs Reduce driver footprint • i.e. in scenarios where there is more than instance needed (different credentials, different clusters etc.) • Scripting
20. Michael Simons @rotnroll666 Driver instance hands out sessions Config options apply for all types
21. Michael Simons @rotnroll666 Session purpose Mostly transaction chaining • On the surface • Executes transactional work • Hands out unmanaged transactions • On the inside • Retries your work if necessary and possible • Chains causal bookmarks together
22. Fan out: Transactions • Each session type provides • Auto-commit aka invisible transactions • Managed transactions • Unmanaged transactions • Following examples only showing imperative style 3 x 3 ways of running queries
23. Michael Simons @rotnroll666 Auto-commit transactions The invisible kid.
24. Michael Simons @rotnroll666 Managed transactions The transactions formerly known as transaction functions.
25. Michael Simons @rotnroll666 Unmanaged transactions There might be dragons.
26. Michael Simons @rotnroll666 Unmanaged transactions There might be dragons. • You want to integrate with an external tx manager • You want control over exception handling and retries • Your transactional work is not idempotent
27. Bookmarks • Don’t parse them • Don’t persist them • Sessions handles them for you • Relevant for working against a cluster • Use them when chaining sessions Which I initially though is hard but isn’t
28. Michael Simons @rotnroll666 Bookmarks Usage pattern
29. Michael Simons @rotnroll666 Bookmarks Possible bookmark manager (From SDN 6)
30. The async session
31. The async session Not today.
32. Non blocking, back pressure up into the Bolt server. RxSession
33. Michael Simons @rotnroll666 RxSession: Reactive database access Key facts
34. Michael Simons @rotnroll666 RxSession: Reactive database access Key facts • Fully non blocking, required changes in the Bolt server • Internally using • Netty and existing async behavior • Small parts build with Project Reactor • Build based on PULL_N and TX_STREAMING (See 7687.org, Server v4) • Externally adhering to https://www.reactive-streams.org • Publisher compliant with the TCK (See https://github.com/reactive-streams/ reactive-streams-jvm/tree/master/tck)
35. RxSession drives SDN 6 Well, the reactive parts.
36. Michael Simons @rotnroll666 RxSession: Observations Your mileage may vary. • Better resource usage • More efficient use of threads (non blocking IO is key!) • Slightly less memory usage on the client side • Steep learning curve • All reactive libraries offer a plethora of operators • What is happens on LIMIT / cancel? • Resource handling can be tricky • Reactive context works awesome for user / tenancy information
37. Michael Simons @rotnroll666 RxSession: Observations Your mileage may vary. • For resource usage and some unqualified benchmarking have a look at https://github.com/michael-simons/neo4j-from-the-jvm-ecosystem
38. Michael Simons @rotnroll666 RxSession: Use it with 3rd party libraries Don’t use the publishers alone. • Officially under test are • Project Reactor https://projectreactor.io (Flux and Mono) • RxJava 2 https://github.com/ReactiveX/RxJava (Flowable and Observable) • Works with • SmallRye Mutiny! https://smallrye.io/smallrye-mutiny/ (Multi) • Helidons reactive implementation • Most efficient usage: With Project Reactor (no conversion required)
39. A couple of examples
40. Reading things
41. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments
42. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments Important: This delays the creation of the session until something subscribes!
43. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments Pass the new resource to the actual work
44. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments Clean up the resource. Not needed for tx as this is done via the tx function below
45. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments Create something usable from the publisher API. I usually do this as early as possible.
46. Michael Simons @rotnroll666 Project Reactor Standard for Spring environments Apply back pressure here, i.e. - take(n) (Can replace a LIMIT inside the query*) - skip(n) (Can replace a SKIP inside the query*) - takeUntil() - elementAt(i) - bufferUntil
47. Michael Simons @rotnroll666 RxJava2 For example with Micronaut
48. Michael Simons @rotnroll666 RxJava2 For example with Micronaut Yikes… No async clean up here… At least, session creation is correctly deferred
49. Michael Simons @rotnroll666 SmallRye Mutiny! Used in Quarkus
50. Michael Simons @rotnroll666 SmallRye Mutiny! Used in Quarkus Better. Async clean up, too.
51. Writing things
52. Michael Simons @rotnroll666 Very much the same with tx functions
53. Michael Simons @rotnroll666 Very much the same with tx functions Better don’t do a take(n), skip here in case of multiple results. Also not the best idea to expose such a flow directly as an endpoint. take(n) is effectively a „cancel“ and so is an interrupted browser request. in such case a transaction would be rolled back!
54. Michael Simons @rotnroll666 Take control on cancel You will need unmanaged transactions.
55. Michael Simons @rotnroll666 Take control on cancel You will need unmanaged transactions.
56. Michael Simons @rotnroll666 Take control on cancel You will need unmanaged transactions. Create a nested resource: RxSession provides transactions, transactions provides supplier for committing or rolling them back.
57. Michael Simons @rotnroll666 Take control on cancel You will need unmanaged transactions. The usingWhen form has distinct hooks for async complete, error and cancel. Cancel can be „good“ (taken enough) or „bad“ (somethings gone…)
58. Michael Simons @rotnroll666 Just a couple of pitfalls • Inner project semantics, i.e. here Project Reactor, on a Flux • flatMap: Takes every item and maps it concurrently into a new Publisher • concatMap: Same as above, but one at a time • switchMap: Takes the next item, cancels the inner publisher and maps that item into a new Publisher • Different behaviors between libraries Flux/Mono are both Publishers (never null items, strict failures) Multi/Uni feel like Publishers, but only Multi is. Uni drops everything after 1st
59. Michael Simons @rotnroll666 Just a couple of pitfalls • Cancel: Good or bad? Intended or not accident? • Be careful with transaction boundaries
60. SDN 6 For a better living. Imperative and reactive repositories Imperative and reactive templates Imperative and reactive client Integrated with Spring Transactions and Context Driver
61. Michael Simons @rotnroll666 SDN 6 reactive examples
62. Michael Simons @rotnroll666 SDN 6 reactive examples
63. https://github.com/neo4j-examples/sdn6-reactive-multidatabase More examples https://github.com/michael-simons/neo4j-from-the-jvm-ecosystem Spring Data Neo4j 6 centric, with focus on multi db and fabric JVM in general: Plain driver and various OGMs in various frameworks
64. The holy Graal? • High-performance runtime • GraalVM compiler: JIT written in Java, integrated with Hotspot VM through JVM Compiler interface • Language independent representation of guest languages: A graph • The Truffle framework for creating new interpreters for additional languages • The polyglot runtime • Native Image Speaking about GraalVM
65. Just an overview, too many things to cover.
66. Michael Simons @rotnroll666 GraalVM Native image Ahead-of-Time-Compilation • Create native executables or shared libraries • Restrictions apply • Classes / methods used via reflections must be enumerated in most cases • Same applies for resources • There’s an agent helping you todo this • Faster startup, lower memory overhead • Maybe less throughput than a fully warmed up JVM
67. Michael Simons @rotnroll666 GraalVM Native image Does it work with the Neo4j Java driver? YES! • Netty needed some „support“ (a couple of substitutions, mainly inspired by our work with the Quarkus people) • SSL is kinda hard to get right, requires JNI and others enabled on native image generation • Tested to work with • Quarkus • Spring Boot (with SDN 6, currently in Beta) • Helidon (from Oracle, currently in Alpha, PR-Stage)
68. Michael Simons @rotnroll666 GraalVM Native image Use cases • Workloads that needs to be super elastic • „Scale to 0“ • CLI
69. Michael Simons @rotnroll666 GraalVM Polyglot programming Lots of fun.
70. Thank you for your time.
71. Michael Simons @rotnroll666 Image sources • Ventilator by Ronan Furuta: https://unsplash.com/photos/7p8cFwM0Y1s • Bookmarks by Kasturi Roy: https://unsplash.com/photos/bYy9hndOx0k • Goblet by Shannon Douglas: https://unsplash.com/photos/8yeyzVfMv0M • Graal and Neo4j polyglot by Oleg Šelajev for my Medium article: https://medium.com/graalvm/3-ways-to-polyglot-with-graalvm- fb28c1542b45