Soumettre la recherche
Mettre en ligne
Java Flight Recorder Behind the Scenes
•
7 j'aime
•
3,412 vues
S
Staffan Larsen
Suivre
Presentation from JavaOne 2013
Lire moins
Lire la suite
Technologie
Actualités & Politique
Signaler
Partager
Signaler
Partager
1 sur 42
Télécharger maintenant
Télécharger pour lire hors ligne
Recommandé
Spark Summit EU talk by Shaun Klopfenstein and Neelesh Shastry
Spark Summit EU talk by Shaun Klopfenstein and Neelesh Shastry
Spark Summit
Java mission control and java flight recorder
Java mission control and java flight recorder
Wolfgang Weigend
Rundeck's History and Future
Rundeck's History and Future
dev2ops
Get Rid Of OutOfMemoryError messages
Get Rid Of OutOfMemoryError messages
Poonam Bajaj Parhar
Introducing BinarySortedMultiMap - A new Flink state primitive to boost your ...
Introducing BinarySortedMultiMap - A new Flink state primitive to boost your ...
Flink Forward
Seven Habits of Highly Effective Jenkins Users (2014 edition!)
Seven Habits of Highly Effective Jenkins Users (2014 edition!)
Andrew Bayer
G1 Garbage Collector: Details and Tuning
G1 Garbage Collector: Details and Tuning
Simone Bordet
Using the New Apache Flink Kubernetes Operator in a Production Deployment
Using the New Apache Flink Kubernetes Operator in a Production Deployment
Flink Forward
Recommandé
Spark Summit EU talk by Shaun Klopfenstein and Neelesh Shastry
Spark Summit EU talk by Shaun Klopfenstein and Neelesh Shastry
Spark Summit
Java mission control and java flight recorder
Java mission control and java flight recorder
Wolfgang Weigend
Rundeck's History and Future
Rundeck's History and Future
dev2ops
Get Rid Of OutOfMemoryError messages
Get Rid Of OutOfMemoryError messages
Poonam Bajaj Parhar
Introducing BinarySortedMultiMap - A new Flink state primitive to boost your ...
Introducing BinarySortedMultiMap - A new Flink state primitive to boost your ...
Flink Forward
Seven Habits of Highly Effective Jenkins Users (2014 edition!)
Seven Habits of Highly Effective Jenkins Users (2014 edition!)
Andrew Bayer
G1 Garbage Collector: Details and Tuning
G1 Garbage Collector: Details and Tuning
Simone Bordet
Using the New Apache Flink Kubernetes Operator in a Production Deployment
Using the New Apache Flink Kubernetes Operator in a Production Deployment
Flink Forward
Real-Time Stream Processing with KSQL and Apache Kafka
Real-Time Stream Processing with KSQL and Apache Kafka
confluent
Flink powered stream processing platform at Pinterest
Flink powered stream processing platform at Pinterest
Flink Forward
Premier Inside-Out: Apache Druid
Premier Inside-Out: Apache Druid
Hortonworks
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
HostedbyConfluent
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
Flink Forward
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
Chen Cheng-Wei
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
Monica Beckwith
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
Amazon Web Services
Apache Kafka Best Practices
Apache Kafka Best Practices
DataWorks Summit/Hadoop Summit
Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28
Amazon Web Services
Amazon Aurora
Amazon Aurora
Amazon Web Services
Kafka tiered-storage-meetup-2022-final-presented
Kafka tiered-storage-meetup-2022-final-presented
Sumant Tambe
Storing 16 Bytes at Scale
Storing 16 Bytes at Scale
Fabian Reinartz
GraalVm and Quarkus
GraalVm and Quarkus
Sascha Rodekamp
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
DataStax
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
DataStax
Apache Arrow: High Performance Columnar Data Framework
Apache Arrow: High Performance Columnar Data Framework
Wes McKinney
SpringOne Tour: Spring Boot 3 and Beyond
SpringOne Tour: Spring Boot 3 and Beyond
VMware Tanzu
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Amazon Web Services
Optimizing Apache Spark UDFs
Optimizing Apache Spark UDFs
Databricks
Diagnosing Your Application on the JVM
Diagnosing Your Application on the JVM
Staffan Larsen
JSR107 State of the Union JavaOne 2013
JSR107 State of the Union JavaOne 2013
Hazelcast
Contenu connexe
Tendances
Real-Time Stream Processing with KSQL and Apache Kafka
Real-Time Stream Processing with KSQL and Apache Kafka
confluent
Flink powered stream processing platform at Pinterest
Flink powered stream processing platform at Pinterest
Flink Forward
Premier Inside-Out: Apache Druid
Premier Inside-Out: Apache Druid
Hortonworks
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
HostedbyConfluent
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
Flink Forward
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
Chen Cheng-Wei
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
Monica Beckwith
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
Amazon Web Services
Apache Kafka Best Practices
Apache Kafka Best Practices
DataWorks Summit/Hadoop Summit
Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28
Amazon Web Services
Amazon Aurora
Amazon Aurora
Amazon Web Services
Kafka tiered-storage-meetup-2022-final-presented
Kafka tiered-storage-meetup-2022-final-presented
Sumant Tambe
Storing 16 Bytes at Scale
Storing 16 Bytes at Scale
Fabian Reinartz
GraalVm and Quarkus
GraalVm and Quarkus
Sascha Rodekamp
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
DataStax
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
DataStax
Apache Arrow: High Performance Columnar Data Framework
Apache Arrow: High Performance Columnar Data Framework
Wes McKinney
SpringOne Tour: Spring Boot 3 and Beyond
SpringOne Tour: Spring Boot 3 and Beyond
VMware Tanzu
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Amazon Web Services
Optimizing Apache Spark UDFs
Optimizing Apache Spark UDFs
Databricks
Tendances
(20)
Real-Time Stream Processing with KSQL and Apache Kafka
Real-Time Stream Processing with KSQL and Apache Kafka
Flink powered stream processing platform at Pinterest
Flink powered stream processing platform at Pinterest
Premier Inside-Out: Apache Druid
Premier Inside-Out: Apache Druid
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
Redis + Kafka = Performance at Scale | Julien Ruaux, Redis Labs
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
Demystifying flink memory allocation and tuning - Roshan Naik, Uber
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
GitLab Auto DevOps 大解析—CI/CD 原來可以這樣做
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
Garbage First Garbage Collector (G1 GC) - Migration to, Expectations and Adva...
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
ABD301-Analyzing Streaming Data in Real Time with Amazon Kinesis
Apache Kafka Best Practices
Apache Kafka Best Practices
Aurora Deep Dive | AWS Floor28
Aurora Deep Dive | AWS Floor28
Amazon Aurora
Amazon Aurora
Kafka tiered-storage-meetup-2022-final-presented
Kafka tiered-storage-meetup-2022-final-presented
Storing 16 Bytes at Scale
Storing 16 Bytes at Scale
GraalVm and Quarkus
GraalVm and Quarkus
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
A Detailed Look At cassandra.yaml (Edward Capriolo, The Last Pickle) | Cassan...
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
Optimizing Your Cluster with Coordinator Nodes (Eric Lubow, SimpleReach) | Ca...
Apache Arrow: High Performance Columnar Data Framework
Apache Arrow: High Performance Columnar Data Framework
SpringOne Tour: Spring Boot 3 and Beyond
SpringOne Tour: Spring Boot 3 and Beyond
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Best Practices for Database Migration to the Cloud: Improve Application Perfo...
Optimizing Apache Spark UDFs
Optimizing Apache Spark UDFs
Similaire à Java Flight Recorder Behind the Scenes
Diagnosing Your Application on the JVM
Diagnosing Your Application on the JVM
Staffan Larsen
JSR107 State of the Union JavaOne 2013
JSR107 State of the Union JavaOne 2013
Hazelcast
MySQL Webinar 2/4 Performance tuning, hardware, optimisation
MySQL Webinar 2/4 Performance tuning, hardware, optimisation
Mark Swarbrick
JVMs in Containers - Best Practices
JVMs in Containers - Best Practices
David Delabassee
Java code coverage with JCov. Implementation details and use cases.
Java code coverage with JCov. Implementation details and use cases.
Alexandre (Shura) Iline
JVMs in Containers
JVMs in Containers
David Delabassee
ASML_FlightRecorderMeetsJava.pdf
ASML_FlightRecorderMeetsJava.pdf
Miro Wengner
Coherence 12.1.2 Hidden Gems
Coherence 12.1.2 Hidden Gems
harvraja
DevDays: Profiling With Java Flight Recorder
DevDays: Profiling With Java Flight Recorder
Miro Wengner
Keynote (Nandini Ramani) - The Role of Java in Heterogeneous Computing & How ...
Keynote (Nandini Ramani) - The Role of Java in Heterogeneous Computing & How ...
AMD Developer Central
Java Mission Control in Java SE 7U40
Java Mission Control in Java SE 7U40
Roger Brinkley
GlassFish in Production Environments
GlassFish in Production Environments
Bruno Borges
Effective Spark on Multi-Tenant Clusters
Effective Spark on Multi-Tenant Clusters
DataWorks Summit/Hadoop Summit
Running your Java EE 6 Applications in the Cloud
Running your Java EE 6 Applications in the Cloud
Arun Gupta
SFBay Area Solr Meetup - June 18th: Benchmarking Solr Performance
SFBay Area Solr Meetup - June 18th: Benchmarking Solr Performance
Lucidworks (Archived)
Spark to Production @Windward
Spark to Production @Windward
Demi Ben-Ari
Java API for WebSocket 1.0: Java EE 7 and GlassFish
Java API for WebSocket 1.0: Java EE 7 and GlassFish
Arun Gupta
C* Summit 2013: Real-Time Big Data with Storm, Cassandra, and In-Memory Compu...
C* Summit 2013: Real-Time Big Data with Storm, Cassandra, and In-Memory Compu...
DataStax Academy
Tuning Java for Big Data
Tuning Java for Big Data
Scott Seighman
JFokus 2011 - Running your Java EE 6 apps in the Cloud
JFokus 2011 - Running your Java EE 6 apps in the Cloud
Arun Gupta
Similaire à Java Flight Recorder Behind the Scenes
(20)
Diagnosing Your Application on the JVM
Diagnosing Your Application on the JVM
JSR107 State of the Union JavaOne 2013
JSR107 State of the Union JavaOne 2013
MySQL Webinar 2/4 Performance tuning, hardware, optimisation
MySQL Webinar 2/4 Performance tuning, hardware, optimisation
JVMs in Containers - Best Practices
JVMs in Containers - Best Practices
Java code coverage with JCov. Implementation details and use cases.
Java code coverage with JCov. Implementation details and use cases.
JVMs in Containers
JVMs in Containers
ASML_FlightRecorderMeetsJava.pdf
ASML_FlightRecorderMeetsJava.pdf
Coherence 12.1.2 Hidden Gems
Coherence 12.1.2 Hidden Gems
DevDays: Profiling With Java Flight Recorder
DevDays: Profiling With Java Flight Recorder
Keynote (Nandini Ramani) - The Role of Java in Heterogeneous Computing & How ...
Keynote (Nandini Ramani) - The Role of Java in Heterogeneous Computing & How ...
Java Mission Control in Java SE 7U40
Java Mission Control in Java SE 7U40
GlassFish in Production Environments
GlassFish in Production Environments
Effective Spark on Multi-Tenant Clusters
Effective Spark on Multi-Tenant Clusters
Running your Java EE 6 Applications in the Cloud
Running your Java EE 6 Applications in the Cloud
SFBay Area Solr Meetup - June 18th: Benchmarking Solr Performance
SFBay Area Solr Meetup - June 18th: Benchmarking Solr Performance
Spark to Production @Windward
Spark to Production @Windward
Java API for WebSocket 1.0: Java EE 7 and GlassFish
Java API for WebSocket 1.0: Java EE 7 and GlassFish
C* Summit 2013: Real-Time Big Data with Storm, Cassandra, and In-Memory Compu...
C* Summit 2013: Real-Time Big Data with Storm, Cassandra, and In-Memory Compu...
Tuning Java for Big Data
Tuning Java for Big Data
JFokus 2011 - Running your Java EE 6 apps in the Cloud
JFokus 2011 - Running your Java EE 6 apps in the Cloud
Dernier
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
Slibray Presentation
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
Manik S Magar
Scanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL Certs
Rizwan Syed
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
Fwdays
CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):
comworks
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Mark Simos
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easy
Alfredo García Lavilla
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
hariprasad279825
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Zilliz
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
Kalema Edgar
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brand
gvaughan
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptx
NavinnSomaal
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
Sergiu Bodiu
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
Florian Wilhelm
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
LoriGlavin3
Hyperautomation and AI/ML: A Strategy for Digital Transformation Success.pdf
Hyperautomation and AI/ML: A Strategy for Digital Transformation Success.pdf
Precisely
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?
Mattias Andersson
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
Fwdays
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
Addepto
Dernier
(20)
Connect Wave/ connectwave Pitch Deck Presentation
Connect Wave/ connectwave Pitch Deck Presentation
Anypoint Exchange: It’s Not Just a Repo!
Anypoint Exchange: It’s Not Just a Repo!
Scanning the Internet for External Cloud Exposures via SSL Certs
Scanning the Internet for External Cloud Exposures via SSL Certs
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
"Subclassing and Composition – A Pythonic Tour of Trade-Offs", Hynek Schlawack
CloudStudio User manual (basic edition):
CloudStudio User manual (basic edition):
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
TrustArc Webinar - How to Build Consumer Trust Through Data Privacy
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Tampa BSides - Chef's Tour of Microsoft Security Adoption Framework (SAF)
Commit 2024 - Secret Management made easy
Commit 2024 - Secret Management made easy
Artificial intelligence in cctv survelliance.pptx
Artificial intelligence in cctv survelliance.pptx
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Leverage Zilliz Serverless - Up to 50X Saving for Your Vector Storage Cost
Unleash Your Potential - Namagunga Girls Coding Club
Unleash Your Potential - Namagunga Girls Coding Club
WordPress Websites for Engineers: Elevate Your Brand
WordPress Websites for Engineers: Elevate Your Brand
SAP Build Work Zone - Overview L2-L3.pptx
SAP Build Work Zone - Overview L2-L3.pptx
DevEX - reference for building teams, processes, and platforms
DevEX - reference for building teams, processes, and platforms
Streamlining Python Development: A Guide to a Modern Project Setup
Streamlining Python Development: A Guide to a Modern Project Setup
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Merck Moving Beyond Passwords: FIDO Paris Seminar.pptx
Hyperautomation and AI/ML: A Strategy for Digital Transformation Success.pdf
Hyperautomation and AI/ML: A Strategy for Digital Transformation Success.pdf
Are Multi-Cloud and Serverless Good or Bad?
Are Multi-Cloud and Serverless Good or Bad?
"Debugging python applications inside k8s environment", Andrii Soldatenko
"Debugging python applications inside k8s environment", Andrii Soldatenko
Gen AI in Business - Global Trends Report 2024.pdf
Gen AI in Business - Global Trends Report 2024.pdf
Java Flight Recorder Behind the Scenes
1.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.1
2.
Java Flight Recorder Behind
the Scenes Staffan Larsen Java SE Serviceability Architect staffan.larsen@oracle.com @stalar
3.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.3 The following is intended to outline our general product direction. It is intended for information purposes only, and may not be incorporated into any contract. It is not a commitment to deliver any material, code, or functionality, and should not be relied upon in making purchasing decisions. The development, release, and timing of any features or functionality described for Oracle s products remains at the sole discretion of Oracle.
4.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.4 Java Flight Recorder § Tracer and Profiler § Non-intrusive § Built into the JVM itself § On-demand profiling § After-the-fact capture and analysis § First released in 7u40 Photograph by Jeffrey Milstein http://butdoesitfloat.com/In-the-presence-of-massive-gravitational-fields-this-perfect
5.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.5 Tracer and Profiler § Captures both JVM and application data – Garbage Collections – Synchronization – Compiler – CPU Usage – Exceptions – I/O § Sampling-based profiler – Very low overhead – Accurate data
6.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.6 Non-Intrusive § Typical overhead in benchmarks: 2-3% (!) § Often not noticeable in typical production environments § Turn on and off in runtime § Information already available in the JVM – Zero extra cost Photograph: Andrew Rivett http://www.flickr.com/photos/veggiefrog/3435380297/
7.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.7 Built into the JVM itself § Core of JFR is inside the JVM § Can easily interact with other JVM subsystems § Optimized C++ code § Supporting functionality written in Java
8.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.8 On-Demand Profiling § Start from Java Mission Control – Or from the command line § Easily configure the amount of information to capture § For a profile, a higher overhead can be acceptable § When done, no overhead § Powerful GUI for analysis
9.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.9 After-the-Fact Analysis § In its default mode, very low overhead § Designed to be always-on § Uses circular buffers to store data – In-memory or on-disk § When an SLA breach is detected, dump the current buffers § Dump will have information leading up to the problem n n+1 n+2n+3 n+4
10.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.10
11.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.11 Agenda § Overview of JFR § Demo! § Configuration topics § Implementation details
12.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.12 Configuration § Enable -‐XX:+UnlockCommercialFeatures -‐XX:+FlightRecorder § Start -‐XX:StartFlightRecording=filename=<path>,duration=<time> § Or jcmd <pid> JFR.start filename=<path> duration=<time>
13.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.13 Advanced Configuration Per Recording Session Max age of data maxage=<time> Max size to keep maxsize=<size> Global Settings (-XX:FlightRecorderOptions) Max stack trace depth stackdepth=<n> (default 64) Save recording on exit dumponexit=true Logging loglevel=[ERROR|WARN|INFO| DEBUG|TRACE] Repository path repository=<path>
14.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.14 Recording Sessions § Recordings can specify exactly which information to capture – ~80 events with 3 settings each § But: two preconfigured settings – “default”: provides as much information as possible while keeping overhead to a minimum – “profile”: has more information, but also higher overhead § You can configure your own favorites in Mission Control
15.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.15 Many Simultaneous Recording Sessions § This works great § Each session can have its own settings § Caveat: If there are multiple sessions all of them get the union of the enabled events – Ex: If event A is enabled in on recording, all recordings will see event A – Ex: If event B has two different thresholds, the lower value will apply
16.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.16 JVM How Is It Built? § Information gathering – Instrumentation calls all over the JVM – Application information via Java API § Collected in Thread Local buffers ⇢ Global Buffers ⇢Disk § Binary, proprietary file format § Managed via JMX JFR Thread Buffers Disk JMX Java API Global Buffers GC RuntimeCompiler
17.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.17 “Everything Is an Event” § Header § Payload – Event specific data Event ID End Time Start Time Thread ID StackTrace ID Payload Header Optional Size
18.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.18 Event Types § Instant – Single point in time – Ex: Thread starts § Duration – Timing for something – Ex: GC § Requestable – Happens with a specified frequency – Ex: CPU Usage every second
19.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.19 Event Meta Data § For every event – Name, Path, Description § For every payload item – Name, Type, Description, Content Type
20.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.20 “Content Type” § Describes the semantics of a value § Used to correctly display the value in the UI Content Type Displayed as Bytes 4 MB Percentage 34 % Address 0x23CDA540 Millis 17 ms Nanos 4711 ns
21.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.21 Event Definition in Hotspot <event id="ThreadSleep" path="java/thread_sleep" label="Java Thread Sleep" ...> <value field="time" type="MILLIS" label="Sleep Time"/> </event> § XML definitions are processed into C++ classes
22.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.22 Event Emission in Hotspot JVM_Sleep(int millis) { EventThreadSleep event; ... // actual sleep happens here event.set_time(millis); event.commit(); } § Done! Data is now available in JFR
23.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.23 Thread Park <event id="ThreadPark" path="java/thread_park" label="Java Thread Park" has_thread="true" has_stacktrace="true" is_instant="false"> <value type="CLASS" field="klass" label="Class Parked On"/> <value type="MILLIS" field="timeout" label="Park Timeout"/> <value type="ADDRESS" field="address" label="Address of Object Parked"/> </event>
24.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.24 Thread Park UnsafePark(jboolean isAbsolute, jlong time) { EventThreadPark event; JavaThreadParkedState jtps(thread, time != 0); thread-‐>parker()-‐>park(isAbsolute != 0, time); if (event.should_commit()) { oop obj = thread-‐>current_park_blocker(); event.set_klass(obj ? obj-‐>klass() : NULL); event.set_timeout(time); event.set_address(obj ? (TYPE_ADDRESS)obj : 0); event.commit(); } }
25.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.25 Event Graph
26.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.26 Event Details
27.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.27 Buffers § “Circular” § Designed for low contention Thread Local Buffers Global Buffers Disk Chunk Repository JVM Java
28.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.28 Filtering Early § Enable/disable event § Thresholds – Only if duration is longer than X § Enable/disable stack trace § Frequency – Sample every X
29.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.29 File Format § Self-contained – Everything needed to parse an event is included in the file – New events instantly viewable in the UI § Binary, proprietary § Designed for fast writing § Single file, no dependencies Event Records Event Definitions …Header
30.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.30 Dynamic Runtime and Long-Running Recordings § Can’t leak memory – Can’t aggregate information eternally – Can’t keep references that prohibits class unloading § Dynamic Runtime – Classes can come and go – Threads can come and go § Solutions: Constant Pools, Checkpoints
31.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.31 Problem: Many Events Reference Classes § If every event contained the class name as a string, we would waste lots of space § Solution: Class IDs class Klass { … u8 _trace_id; … } 17event id class count java.lang.Integer 2 … 17event id class count 4711 2 …
32.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.32 Problem: When Do We Write the Class IDs? § IDs need to be part of the file § Classes can be unloaded at any time – Class may not be around until end of recording § Solution: write Class ID when classes are unloaded
33.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.33 Problem: Size of the Class List § Many classes are loaded, not all are referenced in events, we want to save space § Solution: when a class ID is referenced, the class is also “tagged” – Write only tagged classes in the JFR file #define CLASS_USED 1 void use_class_id(Klass* const klass) { klass-‐>_trace_id |= CLASS_USED; }
34.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.34 Problem: Leaking Memory § Over time many classes will be tagged, the size of the class list will increase § Solution: reset the tags each time a class list is written to disk § We call this a “Checkpoint” § A recording file may contain many class lists, each one is only valid for the data immediately preceding it
35.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.35 Constant Pools § The Class List is a special case of a Constant Pool § Classes § Methods § Threads § Thread Groups § Stack Traces § Strings class_pool.lookup(4711) ➞ java.lang.Integer method_pool.lookup(1729) ➞ java.lang.Math:pow()
36.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.36 Checkpoints § At regular intervals, a “checkpoint” is created in the recording § Has everything needed to parse the recording since the last checkpoint checkpoint = events + constant pools + event meta-data events constant pools meta-data checkpoint JFR File
37.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.37 Optimizations § Fast Timestamps – Fast, high resolution CPU time where available – Invariant TSC instructions § Stack Traces – Each event stores the thread’s stack trace – Pool of stack traces
38.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.38 Differences vs. JRockit § I/O: File path, Socket address § Exceptions § Reverse call trace view in Mission Control § Easier configuration in Mission Control § Deeper (configurable) stack traces § Internal JVM differences: GC
39.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.39 More Information § Whitepaper http://www.oracle.com/missioncontrol § User Guide http://docs.oracle.com/javase/7/docs/technotes/guides/jfr/index.html § Forum https://forums.oracle.com/community/developer/english/java/ java_hotspot_virtual_machine/java_mission_control
40.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.40 -‐XX:+UnlockCommercialFeatures -‐XX:+FlightRecorder Remember
41.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.41 @stalar staffan.larsen@oracle.com serviceability-dev@openjdk.java.net
42.
Copyright © 2013,
Oracle and/or its affiliates. All rights reserved.42
Télécharger maintenant