Apache Hadoop YARN - Enabling Next Generation Data Applications

53,809 views

Published on

This presentation dives into the future of Hadoop: YARN.
-- Why YARN?
-- YARN Architecture and Concepts
-- Building Applications on YARN
-- Next Steps

Published in: Technology
6 Comments
104 Likes
Statistics
Notes
  • Njce! Thanks for sharing.
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • #Hadoop Hadoop Tutorial (Videos and Books) $15 29Gb of data http://www.dbmanagement.info/Tutorials/Hadoop.htm
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • More than 5000 registered IT consultants and Corporates.Search for IT online training Providers at http://www.todaycourses.com
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • Hi All, We are planning to start Hadoop online training batch on this week... If any one interested to attend the demo please register in our website... For this batch we are also provide everyday recorded sessions with Materials. For more information feel free to contact us : siva@keylabstraining.com. For Course Content and Recorded Demo Click Here : http://www.keylabstraining.com/hadoop-online-training-hyderabad-bangalore
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
  • http://dbmanagement.info/Tutorials/MapReduce.htm
       Reply 
    Are you sure you want to  Yes  No
    Your message goes here
No Downloads
Views
Total views
53,809
On SlideShare
0
From Embeds
0
Number of Embeds
30,587
Actions
Shares
0
Downloads
1,629
Comments
6
Likes
104
Embeds 0
No embeds

No notes for slide

Apache Hadoop YARN - Enabling Next Generation Data Applications

  1. 1. © Hortonworks Inc. 2013 Apache Hadoop YARN Enabling next generation data applications Page 1 Arun C. Murthy Founder & Architect @acmurthy (@hortonworks)
  2. 2. © Hortonworks Inc. 2013 Hello! • Founder/Architect at Hortonworks Inc. – Lead - Map-Reduce/YARN/Tez – Formerly, Architect Hadoop MapReduce, Yahoo – Responsible for running Hadoop MapReduce as a service for all of Yahoo (~50k nodes footprint) • Apache Hadoop, ASF – Frmr. VP, Apache Hadoop, ASF (Chair of Apache Hadoop PMC) – Long-term Committer/PMC member (full time >6 years) – Release Manager for hadoop-2.x Page 2
  3. 3. © Hortonworks Inc. 2013 Agenda • Why YARN? • YARN Architecture and Concepts • Building applications on YARN • Next Steps Page 3
  4. 4. © Hortonworks Inc. 2013 Agenda • Why YARN? • YARN Architecture and Concepts • Building applications on YARN • Next Steps Page 4
  5. 5. © Hortonworks Inc. 2013 The 1st Generation of Hadoop: Batch HADOOP 1.0 Built for Web-Scale Batch Apps Single  App   BATCH HDFS Single  App   INTERACTIVE Single  App   BATCH HDFS •  All other usage patterns must leverage that same infrastructure •  Forces the creation of silos for managing mixed workloads Single  App   BATCH HDFS Single  App   ONLINE
  6. 6. © Hortonworks Inc. 2013 Hadoop MapReduce Classic • JobTracker – Manages cluster resources and job scheduling • TaskTracker – Per-node agent – Manage tasks Page 6
  7. 7. © Hortonworks Inc. 2013 MapReduce Classic: Limitations • Scalability – Maximum Cluster size – 4,000 nodes – Maximum concurrent tasks – 40,000 – Coarse synchronization in JobTracker • Availability – Failure kills all queued and running jobs • Hard partition of resources into map and reduce slots – Low resource utilization • Lacks support for alternate paradigms and services – Iterative applications implemented using MapReduce are 10x slower Page 7
  8. 8. © Hortonworks Inc. 2012© Hortonworks Inc. 2013. Confidential and Proprietary. Our Vision: Hadoop as Next-Gen Platform HADOOP 1.0 HDFS   (redundant,  reliable  storage)   MapReduce   (cluster  resource  management    &  data  processing)   HDFS2   (redundant,  reliable  storage)   YARN   (cluster  resource  management)   MapReduce   (data  processing)   Others   (data  processing)   HADOOP 2.0 Single Use System Batch Apps Multi Purpose Platform Batch, Interactive, Online, Streaming, … Page 8
  9. 9. © Hortonworks Inc. 2013 YARN: Taking Hadoop Beyond Batch Page 9 Applica;ons  Run  Na;vely  IN  Hadoop   HDFS2  (Redundant,  Reliable  Storage)   YARN  (Cluster  Resource  Management)       BATCH   (MapReduce)   INTERACTIVE   (Tez)   STREAMING   (Storm,  S4,…)   GRAPH   (Giraph)   IN-­‐MEMORY   (Spark)   HPC  MPI   (OpenMPI)   ONLINE   (HBase)   OTHER   (Search)   (Weave…)   Store ALL DATA in one place… Interact with that data in MULTIPLE WAYS with Predictable Performance and Quality of Service
  10. 10. © Hortonworks Inc. 2013 5 5 Key Benefits of YARN 1.  Scale 2.  New Programming Models & Services 3.  Improved cluster utilization 4.  Agility 5.  Beyond Java Page 10
  11. 11. © Hortonworks Inc. 2013 Agenda • Why YARN • YARN Architecture and Concepts • Building applications on YARN • Next Steps Page 11
  12. 12. © Hortonworks Inc. 2013 A Brief History of YARN •  Originally conceived & architected by the team at Yahoo! – Arun Murthy created the original JIRA in 2008 and led the PMC •  The team at Hortonworks has been working on YARN for 4 years •  YARN based architecture running at scale at Yahoo! – Deployed on 35,000 nodes for 6+ months •  Multitude of YARN applications Page 12
  13. 13. © Hortonworks Inc. 2013 Concepts • Application – Application is a job submitted to the framework – Example – Map Reduce Job • Container – Basic unit of allocation – Fine-grained resource allocation across multiple resource types (memory, cpu, disk, network, gpu etc.) – container_0 = 2GB, 1CPU – container_1 = 1GB, 6 CPU – Replaces the fixed map/reduce slots 13
  14. 14. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   Container  1.1   Container  2.4   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   Container  1.2   Container  1.3   AM  1   Container  2.1   Container  2.2   Container  2.3   AM2   YARN Architecture Scheduler  
  15. 15. © Hortonworks Inc. 2013 Architecture • Resource Manager – Global resource scheduler – Hierarchical queues • Node Manager – Per-machine agent – Manages the life-cycle of container – Container resource monitoring • Application Master – Per-application – Manages application scheduling and task execution – E.g. MapReduce Application Master 15
  16. 16. © Hortonworks Inc. 2013 Design Centre • Split up the two major functions of JobTracker – Cluster resource management – Application life-cycle management • MapReduce becomes user-land library 16
  17. 17. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   Container  1.1   Container  2.4   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   Container  1.2   Container  1.3   AM  1   Container  2.2   Container  2.1   Container  2.3   AM2   YARN Architecture - Walkthrough Scheduler  Client2  
  18. 18. © Hortonworks Inc. 2013 5 Review - Benefits of YARN 1.  Scale 2.  New Programming Models & Services 3.  Improved cluster utilization 4.  Agility 5.  Beyond Java Page 18
  19. 19. © Hortonworks Inc. 2013 Agenda • Why YARN • YARN Architecture and Concepts • Building applications on YARN • Next Steps Page 19
  20. 20. © Hortonworks Inc. 2013 YARN Applications • Data processing applications and services – Online Serving – HOYA (HBase on YARN) – Real-time event processing – Storm, S4, other commercial platforms – Tez – Generic framework to run a complex DAG – MPI: OpenMPI, MPICH2 – Master-Worker – Machine Learning: Spark – Graph processing: Giraph – Enabled by allowing the use of paradigm-specific application master Run all on the same Hadoop cluster! Page 20
  21. 21. © Hortonworks Inc. 2013 YARN – Implementing Applications • What APIs do I need to use? – Only three protocols – Client to ResourceManager – Application submission – ApplicationMaster to ResourceManager – Container allocation – ApplicationMaster to NodeManager – Container launch – Use client libraries for all 3 actions – Module yarn-client – Provides both synchronous and asynchronous libraries – Use 3rd party like Weave –  http://continuuity.github.io/weave/ 21
  22. 22. © Hortonworks Inc. 2013 YARN – Implementing Applications • What do I need to do? – Write a submission Client – Write an ApplicationMaster (well copy-paste) – DistributedShell is the new WordCount! – Get containers, run whatever you want! 22
  23. 23. © Hortonworks Inc. 2013 YARN – Implementing Applications • What else do I need to know? – Resource Allocation & Usage – ResourceRequest ! – Container ! – ContainerLaunchContext! – LocalResource! – ApplicationMaster – ApplicationId – ApplicationAttemptId! – ApplicationSubmissionContext! 23
  24. 24. © Hortonworks Inc. 2013 YARN – Resource Allocation & Usage! • ResourceRequest! – Fine-grained resource ask to the ResourceManager – Ask for a specific amount of resources (memory, cpu etc.) on a specific machine or rack – Use special value of * for resource name for any machine Page 24 ResourceRequest! priority! resourceName! capability! numContainers!
  25. 25. © Hortonworks Inc. 2013 YARN – Resource Allocation & Usage! • ResourceRequest! ! Page 25 priority! capability! resourceName! numContainers! ! 0! ! <2gb, 1 core>! host01! 1! rack0! 1! *! 1! 1! <4gb, 1 core>! *! 1!
  26. 26. © Hortonworks Inc. 2013 YARN – Resource Allocation & Usage! • Container! – The basic unit of allocation in YARN – The result of the ResourceRequest provided by ResourceManager to the ApplicationMaster – A specific amount of resources (cpu, memory etc.) on a specific machine Page 26 Container! containerId! resourceName! capability! tokens!
  27. 27. © Hortonworks Inc. 2013 YARN – Resource Allocation & Usage! • ContainerLaunchContext! – The context provided by ApplicationMaster to NodeManager to launch the Container! – Complete specification for a process – LocalResource used to specify container binary and dependencies – NodeManager responsible for downloading from shared namespace (typically HDFS) Page 27 ContainerLaunchContext! container! commands! environment! localResources! LocalResource! uri! type!
  28. 28. © Hortonworks Inc. 2013 YARN - ApplicationMaster • ApplicationMaster – Per-application controller aka container_0! – Parent for all containers of the application – ApplicationMaster negotiates all it’s containers from ResourceManager – ApplicationMaster container is child of ResourceManager – Think init process in Unix – RM restarts the ApplicationMaster attempt if required (unique ApplicationAttemptId) – Code for application is submitted along with Application itself Page 28
  29. 29. © Hortonworks Inc. 2013 YARN - ApplicationMaster • ApplicationMaster – ApplicationSubmissionContext is the complete specification of the ApplicationMaster, provided by Client – ResourceManager responsible for allocating and launching ApplicationMaster container Page 29 ApplicationSubmissionContext! resourceRequest! containerLaunchContext! appName! queue!
  30. 30. © Hortonworks Inc. 2013 YARN Application API - Overview • hadoop-yarn-client module • YarnClient is submission client api • Both synchronous & asynchronous APIs for resource allocation and container start/stop • Synchronous API – AMRMClient! – AMNMClient! • Asynchronous API – AMRMClientAsync! – AMNMClientAsync! Page 30
  31. 31. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   Container  1.1   Container  2.4   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   Container  1.2   Container  1.3   AM  1   Container  2.2   Container  2.1   Container  2.3   AM2   Client2     New Application Request: YarnClient.createApplication! Submit Application: YarnClient.submitApplication! 1 2 YARN Application API – The Client Scheduler  
  32. 32. © Hortonworks Inc. 2013 YARN Application API – The Client • YarnClient! – createApplication to create application – submitApplication to start application –  Application developer needs to provide ApplicationSubmissionContext! – APIs to get other information from ResourceManager –  getAllQueues! –  getApplications! –  getNodeReports! – APIs to manipulate submitted application e.g. killApplication! Page 32
  33. 33. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   YARN Application API – Resource Allocation ResourceManager   NodeManager   NodeManager   NodeManager   AM   registerApplicationMaster! 1 4 AMRMClient.allocate! Container! 2 3 NodeManager   NodeManager   NodeManager   NodeManager   unregisterApplicationMaster! Scheduler  
  34. 34. © Hortonworks Inc. 2013 YARN Application API – Resource Allocation • AMRMClient - Synchronous API for ApplicationMaster to interact with ResourceManager – Prologue / epilogue – registerApplicationMaster / unregisterApplicationMaster – Resource negotiation with ResourceManager –  Internal book-keeping - addContainerRequest / removeContainerRequest / releaseAssignedContainer! –  Main API – allocate! – Helper APIs for cluster information –  getAvailableResources! –  getClusterNodeCount! Page 34
  35. 35. © Hortonworks Inc. 2013 YARN Application API – Resource Allocation • AMRMClientAsync - Asynchronous API for ApplicationMaster – Extension of AMRMClient to provide asynchronous CallbackHandler! – Callbacks make it easier to build mental model of interaction with ResourceManager for the application developer –  onContainersAllocated! –  onContainersCompleted! –  onNodesUpdated! –  onError! –  onShutdownRequest! Page 35
  36. 36. © Hortonworks Inc. 2012 NodeManager   NodeManager     NodeManager     NodeManager     YARN Application API – Using Resources Container  1.1   NodeManager   NodeManager   NodeManager   NodeManager     NodeManager   NodeManager     NodeManager     NodeManager     AM  1   AMNMClient.startContainer! AMNMClient.getContainerStatus! ResourceManager   Scheduler  
  37. 37. © Hortonworks Inc. 2013 YARN Application API – Using Resources • AMNMClient - Synchronous API for ApplicationMaster to launch / stop containers at NodeManager – Simple (trivial) APIs –  startContainer! –  stopContainer! –  getContainerStatus! Page 37
  38. 38. © Hortonworks Inc. 2013 YARN Application API – Using Resources • AMNMClient - Asynchronous API for ApplicationMaster to launch / stop containers at NodeManager – Simple (trivial) APIs –  startContainerAsync! –  stopContainerAsync! –  getContainerStatusAsync! – CallbackHandler to make it easier to build mental model of interaction with NodeManager for the application developer –  onContainerStarted! –  onContainerStopped! –  onStartContainerError! –  onContainerStatusReceived! Page 38
  39. 39. © Hortonworks Inc. 2013 YARN Application API - Development • Un-Managed Mode for ApplicationMaster – Run ApplicationMaster on development machine rather than in- cluster – No submission client – hadoop-yarn-applications-unmanaged-am-launcher! – Easier to step through debugger, browse logs etc.! Page 39 ! $ bin/hadoop jar hadoop-yarn-applications-unmanaged-am-launcher.jar ! Client ! –jar my-application-master.jar ! –cmd ‘java MyApplicationMaster <args>’!
  40. 40. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • Overview – YARN application to run n copies for a Shell command – Simplest example of a YARN application – get n containers and run a specific Unix command Page 40 ! $ bin/hadoop jar hadoop-yarn-applications-distributedshell.jar ! org.apache.hadoop.yarn.applications.distributedshell.Client ! –shell_command ‘/bin/date’ ! –num_containers <n>! Code: https://github.com/hortonworks/simple-yarn-app
  41. 41. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • Code Overview – User submits application to ResourceManager via org.apache.hadoop.yarn.applications.distributedsh ell.Client! – Client provides ApplicationSubmissionContext to the ResourceManager – It is responsibility of org.apache.hadoop.yarn.applications.distributedsh ell.ApplicationMaster to negotiate n containers – ApplicationMaster launches containers with the user-specified command as ContainerLaunchContext.commands! Page 41
  42. 42. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • Client – Code Walkthrough – hadoop-yarn-client module – Steps: – YarnClient.createApplication! – Specify ApplicationSubmissionContext, in particular, ContainerLaunchContext with commands, and other key pieces such as resource capability for ApplicationMaster container and queue, appName, appType etc. – YarnClient.submitApplication! Page 42 1 . // Create yarnClient! 2 .   YarnClient yarnClient = YarnClient.createYarnClient();! 3 .   yarnClient.init(new Configuration());! 4 .   yarnClient.start();! 5 . ! 6 .    // Create application via yarnClient! 7 .  YarnClientApplication app = yarnClient.createApplication();! 8 . !
  43. 43. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell Page 43 9.  // Set up the container launch context for the application master! 1 0.    ContainerLaunchContext amContainer = ! 1 1.        Records.newRecord(ContainerLaunchContext.class);! 1 2.    List<String> command = new List<String>();! 1 3.    commands.add("$JAVA_HOME/bin/java");! 1 4.    commands.add("-Xmx256M");! 1 5.    commands.add(! 1 6.        "org.apache.hadoop.yarn.applications.distributedshell.ApplicationMaster");! 1 7.    commands.add("--container_memory 1024” );! 1 8.    commands.add("--container_cores 1”);! 1 9.    commands.add("--num_containers 3");;! 2 0.    amContainer.setCommands(commands);! 2 1. ! 2 2.    // Set up resource type requirements for ApplicationMaster! 2 3.    Resource capability = Records.newRecord(Resource.class);! 2 4.    capability.setMemory(256);! 2 5.    capability.setVirtualCores(2);! 2 6. ! • Client – Code Walkthrough Command to launch ApplicationMaster process Resources required for ApplicationMaster container!
  44. 44. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell Page 44 2 7.    // Finally, set-up ApplicationSubmissionContext for the application! 2 8.    ApplicationSubmissionContext appContext = ! 2 9.        app.getApplicationSubmissionContext();! 3 0.    appContext.setQueue("my-queue");            // queue ! 3 1.    appContext.setAMContainerSpec(amContainer);! 3 2.    appContext.setResource(capability);! 3 3.    appContext.setApplicationName("my-app");     // application name! 3 4.    appContext.setApplicationType(”DISTRIBUTED_SHELL");    // application type! 3 5. ! 3 6.    // Submit application! 3 7.    yarnClient.submitApplication(appContext); • Client – Code Walkthrough ApplicationSubmissionContext ! for ApplicationMaster Submit application to ResourceManager
  45. 45. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • ApplicationMaster – Code Walkthrough – Again, hadoop-yarn-client module – Steps: – AMRMClient.registerApplication! – Negotiate containers from ResourceManager by providing ContainerRequest to AMRMClient.addContainerRequest – Take the resultant Container returned via subsequent call to AMRMClient.allocate, build ContainerLaunchContext with Container and commands, then launch them using AMNMClient.launchContainer! – Use LocalResources to specify software/configuration dependencies for each worker container – Wait till done… AllocateResponse.getCompletedContainersStatuses from subsequent calls to AMRMClient.allocate – AMRMClient.unregisterApplication! Page 45
  46. 46. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • ApplicationMaster – Code Walkthrough Page 46 1.  // Initialize clients to ResourceManager and NodeManagers! 2 .        Configuration conf = new Configuration();! 3 . ! 4 .        AMRMClient rmClient = AMRMClientAsync.createAMRMClient();! 5 .        rmClient.init(conf);! 6 .        rmClient.start();! 7 . ! 8 .        NMClient nmClient = NMClient.createNMClient();! 9 .        nmClientAsync.init(conf);! 1 0.     nmClientAsync.start();! 1 1.! 1 2.     // Register with ResourceManager! 1 3.     rmClient.registerApplicationMaster("", 0, "");! Initialize clients to ResourceManager and NodeManagers Register with ResourceManager
  47. 47. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • ApplicationMaster – Code Walkthrough Page 47 1 5.    // Priority for worker containers - priorities are intra-application! 1 6.    Priority priority = Records.newRecord(Priority.class);! 1 7.    priority.setPriority(0);      ! 1 8. ! 1 9.    // Resource requirements for worker containers! 2 0.    Resource capability = Records.newRecord(Resource.class);! 2 1.    capability.setMemory(128);! 2 2.    capability.setVirtualCores(1);! 2 3. ! 2 4.    // Make container requests to ResourceManager! 2 5.    for (int i = 0; i < n; ++i) {! 2 6.      ContainerRequest containerAsk = new ContainerRequest(capability, null, null, priority);! 2 7.      rmClient.addContainerRequest(containerAsk);! 2 8.    } ! Setup requirements for worker containers Make resource requests to ResourceManager
  48. 48. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • ApplicationMaster – Code Walkthrough Page 48 ! 3 0.    // Obtain allocated containers and launch ! 3 1.    int allocatedContainers = 0;! 3 2.    while (allocatedContainers < n) {! 3 3.      AllocateResponse response = rmClient.allocate(0);! 3 4.      for (Container container : response.getAllocatedContainers()) {! 3 5.        ++allocatedContainers;! 3 6. ! 3 7.        // Launch container by create ContainerLaunchContext! 3 8.        ContainerLaunchContext ctx = Records.newRecord(ContainerLaunchContext.class);! 3 9.        ctx.setCommands(Collections.singletonList(”/bin/date”));! 4 0.        nmClient.startContainer(container, ctx);! 4 1.      }! 4 2.      Thread.sleep(100);! 4 3.    } ! Setup requirements for worker containers Make resource requests to ResourceManager
  49. 49. © Hortonworks Inc. 2013 Sample YARN Application – DistributedShell • ApplicationMaster – Code Walkthrough Page 49 4     // Now wait for containers to complete! 4     int completedContainers = 0;! 4     while (completedContainers < n) {! 4       AllocateResponse response = rmClient.allocate(completedContainers/n);! 4       for (ContainerStatus status : response.getCompletedContainersStatuses()) {! 5         if (status.getExitStatus() == 0) {! 5           ++completedContainers;! 5         }  ! 5       }! 4       Thread.sleep(100);! 5     }! 5  ! 5     // Un-register with ResourceManager! 5     rmClient.unregisterApplicationMaster(SUCCEEDED, "", "");! Wait for containers to complete successfully Un-register with ResourceManager
  50. 50. © Hortonworks Inc. 2013 Apache Hadoop MapReduce on YARN • Original use-case • Most complex application to build – Data-locality – Fault tolerance – ApplicationMaster recovery: Check point to HDFS – Intra-application Priorities: Maps v/s Reduces – Needed complex feedback mechanism from ResourceManager – Security – Isolation • Binary compatible with Apache Hadoop 1.x Page 50
  51. 51. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   map  1.1   reduce2.1   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   map1.2   reduce1.1   MR  AM  1   map2.1   map2.2   reduce2.2   MR  AM2   Apache Hadoop MapReduce on YARN Scheduler  
  52. 52. © Hortonworks Inc. 2013 Apache Tez on YARN • Replaces MapReduce as primitive for Pig, Hive, Cascading etc. – Smaller latency for interactive queries – Higher throughput for batch queries – http://incubator.apache.org/projects/tez.html Page 52 Hive - Tez SELECT a.x, AVERAGE(b.y) AS avg FROM a JOIN b ON (a.id = b.id) GROUP BY a UNION SELECT x, AVERAGE(y) AS AVG FROM c GROUP BY x ORDER BY AVG; Hive - MR
  53. 53. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   map  1.1   vertex1.2.2   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   map1.2   reduce1.1   MR  AM  1   vertex1.1.1   vertex1.1.2   vertex1.2.1   Tez  AM1   Apache Tez on YARN Scheduler  
  54. 54. © Hortonworks Inc. 2013 HOYA - Apache HBase on YARN • Hoya – Apache HBase becomes user-level application • Use cases – Small HBase cluster in large YARN cluster – Dynamic HBase clusters – Transient/intermittent clusters for workflows • APIs to create, start, stop & delete HBase clusters • Flex cluster size: increase/decrease size with load • Recover from Region Server loss with new container. Page 54 Code: https://github.com/hortonworks/hoya
  55. 55. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   map  1.1   Hbase  Master1.1   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   map1.2   reduce1.1   MR  AM  1   RegionServer  1.1   RegionServer  1.2   RegionServer  1.3   HOYA  AM1   HOYA - Apache HBase on YARN Scheduler  
  56. 56. © Hortonworks Inc. 2013 HOYA - Highlights • Cluster specification stored as JSON in HDFS • Config directory cached - dynamically patched before pushing up as local resources for Master & RegionServers • HBase tar file stored in HDFS -clusters can use the same/different HBase versions • Handling of cluster flexing is the same code as unplanned container loss. • No Hoya code on RegionServers: client and AM only Page 56
  57. 57. © Hortonworks Inc. 2013 Storm on YARN • Ability to deploy multiple Storm clusters on YARN for real-time event processing • Yahoo – Primary contributor – 200+ nodes in production • Ability to recover from faulty nodes – Get new containers • Auto-scale for load balancing – Get new containers as load increases – Release containers as load decreases Page 57 Code: https://github.com/yahoo/storm-yarn
  58. 58. © Hortonworks Inc. 2012 NodeManager   NodeManager   NodeManager   NodeManager   map  1.1   Nimbus1.4   ResourceManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   NodeManager   map1.2   reduce1.1   MR  AM  1   Nimbus1.1   Nimbus1.2   Nimbus1.3   Storm  AM1   Storm on YARN Scheduler  
  59. 59. © Hortonworks Inc. 2013 General Architectural Considerations • Fault Tolerance – Checkpoint • Security • Always-On services • Scheduler features – Whitelist resources – Blacklist resources – Labels for machines – License management Page 59
  60. 60. © Hortonworks Inc. 2013 Agenda • Why YARN • YARN Architecture and Concepts • Building applications on YARN • Next Steps Page 60
  61. 61. © Hortonworks Inc. 2013 HDP 2.0 Community Preview & YARN Certification Program Goal: Accelerate # of certified YARN-based solutions •  HDP 2.0 Community Preview – Contains latest community Beta of Apache Hadoop 2.0 & YARN – Delivered as easy to use Sandbox VM, as well as RPMs and Tarballs – Enables YARN Cert Program Community & commercial ecosystem to test and certify new and existing YARN-based apps •  YARN Certification Program – More than 14 partners in program at launch –  Splunk* –  Elastic Search* –  Altiscale* –  Concurrent* –  Microsoft –  Platfora –  Tableau –  (IBM) DataStage –  Informatica –  Karmasphere –  and others * Already certified
  62. 62. © Hortonworks Inc. 2013 Forum & Office Hours • YARN Forum – Community of Hadoop YARN developers – Focused on collaboration and Q&A –  http://hortonworks.com/community/forums/forum/yarn • Office Hours – http://www.meetup.com/HSquared-Hadoop-Hortonworks-User- Group/ – Bi-weekly office hours with Hortonworks engineers & architects – Every other Thursday, starting on Aug 15th from 4:30 – 5:30 PM (PST) – At Hortonworks Palo Alto HQ – West Bayshore Rd. Palo Alto, CA 94303 USA Page 62
  63. 63. © Hortonworks Inc. 2013 Technical Resources • ASF hadoop-2.1.0-beta – Coming soon! –  http://hadoop.apache.org/common/releases.html – Release Documentation: http://hadoop.apache.org/common/docs/r2.1.0-beta • Blogs: –  http://hortonworks.com/hadoop/yarn –  http://hortonworks.com/blog/category/apache-hadoop/yarn/ –  http://hortonworks.com/blog/introducing-apache-hadoop-yarn/ –  http://hortonworks.com/blog/apache-hadoop-yarn-background- and-an-overview/ Page 63
  64. 64. © Hortonworks Inc. 2013 What Next? • Download the Book • Download the Community preview • Join the Beta Program • Participate via the YARN forum • Come see us during the YARN Office Hours • Follow us… @hortonworks Thank You! Page 64 http://hortonworks.com/hadoop/yarn

×