# Copyright 2008 The Apache Software Foundation Licensed under the # Apache License, Version 2.0 (the "License"); you may not use this # file except in compliance with the License. You may obtain a copy # of the License at http://www.apache.org/licenses/LICENSE-2.0 Unless # required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or # implied. See the License for the specific language governing # permissions and limitations under the License. This package implements fair scheduling for MapReduce jobs. Fair scheduling is a method of assigning resources to jobs such that all jobs get, on average, an equal share of resources over time. When there is a single job running, that job uses the entire cluster. When other jobs are submitted, tasks slots that free up are assigned to the new jobs, so that each job gets roughly the same amount of CPU time. Unlike the default Hadoop scheduler, which forms a queue of jobs, this lets short jobs finish in reasonable time while not starving long jobs. It is also a reasonable way to share a cluster between a number of users. Finally, fair sharing can also work with job priorities - the priorities are used as weights to determine the fraction of total compute time that each job should get. In addition to providing fair sharing, the Fair Scheduler allows assigning jobs to "pools" with guaranteed minimum shares. When a pool contains jobs, it gets at least its minimum share, but when a pool does not need its full capacity, the excess is shared between other running jobs. Thus pools are a way to guarantee capacity for particular user groups while utilizing the cluster efficiently when these users are not submitting any jobs. Within each pool, fair sharing is used to share capacity between the running jobs. By default the pool is set based on the queue.name property in the jobconf which will be introduced with the Hadoop Resource Manager (JIRA 3445), but it's possible to also have a pool per user or per Unix user group. The fair scheduler lets all jobs run by default, but it is also possible to limit the number of running jobs per user and per pool through the config file. This can be useful when a user must submit hundreds of jobs at once, or in general to improve performance if running too many jobs at once would cause too much intermediate data to be created or too much context-switching. Limiting the jobs does not cause any subsequently submitted jobs to fail, only to wait in the sheduler's queue until some of the user's earlier jobs finish. Jobs to run from each user/pool are chosen in order of priority and then submit time, as in the default FIFO scheduler in Hadoop. Finally, the fair scheduler provides several extension points where the basic functionality can be extended. For example, the weight calculation can be modified to give a priority boost to new jobs, implementing a "shortest job first" like policy which will reduce response times for interactive jobs even further. -------------------------------------------------------------------------------- BUILDING: In HADOOP_HOME, run ant package to build Hadoop and its contrib packages. -------------------------------------------------------------------------------- INSTALLING: To run the fair scheduler in your Hadoop installation, you need to put it on the CLASSPATH. The easiest way is to copy the hadoop-*-fairscheduler.jar from HADOOP_HOME/build/contrib/fairscheduler to HADOOP_HOME/lib. Alternatively you can modify HADOOP_CLASSPATH to include this jar, in conf/hadoop-env.sh. You will also need to set the following property in the Hadoop config file (conf/hadoop-site.xml) to have Hadoop use the fair scheduler: mapred.jobtracker.taskScheduler org.apache.hadoop.mapred.FairScheduler Once you restart the cluster, you can check that the fair scheduler is running by going to http:///scheduler on the JobTracker's web UI. A "job scheduler administration" page should be visible there. This page is described in the Administration section. -------------------------------------------------------------------------------- CONFIGURING: The following properties can be set in hadoop-site.xml to configure the scheduler: mapred.fairscheduler.allocation.file: Specifies an absolute path to an XML file which contains the allocations for each pool, as well as the per-pool and per-user limits on number of running jobs. If this property is not provided, allocations are not used. This file must be in XML format, and can contain three types of elements: - pool elements, which may contain elements for minMaps, minReduces and maxRunningJobs (limit the number of jobs from the pool to run at once). - user elements, which may contain a maxRunningJobs to limit jobs. - A userMaxJobsDefault element, which sets the running job limit for any users that do not have their own elements. The following example file shows how to create each type of element: 5 5 6 3 This example creates a pool sample_pool with a guarantee of 5 map slots and 5 reduce slots. It also limits the number of running jobs per user to 3, except for sample_user, who can run 6 jobs concurrently. Any pool not defined in the allocations file will have no guaranteed capacity. Also, any pool or user with no max running jobs set in the file will be allowed to run an unlimited number of jobs. mapred.fairscheduler.assignmultiple: Allows the scheduler to assign both a map task and a reduce task on each heartbeat, which improves cluster throughput when there are many small tasks to run. Boolean value, default: false. mapred.fairscheduler.sizebasedweight: Take into account job sizes in calculating their weights for fair sharing. By default, weights are only based on job priorities. Setting this flag to true will make them based on the size of the job (number of tasks needed) as well, though not linearly (the weight will be proportional to the log of the number of tasks needed). This lets larger jobs get larger fair shares while still providing enough of a share to small jobs to let them finish fast. Boolean value, default: false. mapred.fairscheduler.poolnameproperty: Specify which jobconf property is used to determine the pool that a job belongs in. String, default: queue.name (the same property as the queue name in the Hadoop Resource Manager, JIRA 3445). You can use user.name or group.name to base it on the Unix user or Unix group for example. mapred.fairscheduler.weightadjuster: An extensibility point that lets you specify a class to adjust the weights of running jobs. This class should implement the WeightAdjuster interface. There is currently one example implementation - NewJobWeightBooster, which increases the weight of jobs for the first 5 minutes of their lifetime to let short jobs finish faster. To use it, set the weightadjuster property to the full class name, org.apache.hadoop.mapred.NewJobWeightBooster. NewJobWeightBooster itself provides two parameters for setting the duration and boost factor - mapred.newjobweightbooster.factor (default 3) and mapred.newjobweightbooster.duration (in milliseconds, default 300000 for 5 minutes). mapred.fairscheduler.loadmanager: An extensibility point that lets you specify a class that determines how many maps and reduces can run on a given TaskTracker. This class should implement the LoadManager interface. By default the task caps in the Hadoop config file are used, but this option could be used to make the load based on available memory and CPU utilization for example. mapred.fairscheduler.taskselector: An extensibility point that lets you specify a class that determines which task from within a job to launch on a given tracker. This can be used to change either the locality policy (e.g. keep some jobs within a particular rack) or the speculative execution algorithm (select when to launch speculative tasks). The default implementation uses Hadoop's default algorithms from JobInProgress. -------------------------------------------------------------------------------- ADMINISTRATION: The fair scheduler provides support for administration at runtime through two mechanisms. First, it is possible to modify pools' allocations and user and pool running job limits at runtime by editing the allocation config file. The scheduler will reload this file 10-15 seconds after it sees that it was modified. Second, current jobs, pools, and fair shares can be examined through the JobTracker's web interface, at http:///scheduler. On this interface, it is also possible to modify jobs' priorities or move jobs from one pool to another and see the effects on the fair shares (this requires JavaScript). The following fields can be seen for each job on the web interface: Submitted - Date and time job was submitted. JobID, User, Name - Job identifiers as on the standard web UI. Pool - Current pool of job. Select another value to move job to another pool. Priority - Current priority. Select another value to change the job's priority. Maps/Reduces Finished: Number of tasks finished / total tasks. Maps/Reduces Running: Tasks currently running. Map/Reduce Fair Share: The average number of task slots that this job should have at any given time according to fair sharing. The actual number of tasks will go up and down depending on how much compute time the job has had, but on average it will get its fair share amount. In addition, it is possible to turn on an "advanced" view for the web UI, by going to http:///scheduler?advanced. This view shows four more columns used for calculations internally: Maps/Reduce Weight: Weight of the job in the fair sharing calculations. This depends on priority and potentially also on job size and job age if the sizebasedweight and NewJobWeightBooster are enabled. Map/Reduce Deficit: The job's scheduling deficit in macine-seconds - the amount of resources it should have gotten according to its fair share, minus how many it actually got. Positive deficit means the job will be scheduled again in the near future because it needs to catch up to its fair share. The scheduler schedules jobs with higher deficit ahead of others. Please see the Implementation section of this document for details. Finally, the web interface provides a button for switching to FIFO scheduling, at runtime, at the bottom of the page, in case this becomes necessary and it is inconvenient to restart the MapReduce cluster. -------------------------------------------------------------------------------- IMPLEMENTATION: There are two aspects to implementing fair scheduling: Calculating each job's fair share, and choosing which job to run when a task slot becomes available. To select jobs to run, the scheduler then keeps track of a "deficit" for each job - the difference between the amount of compute time it should have gotten on an ideal scheduler, and the amount of compute time it actually got. This is a measure of how "unfair" we've been to the job. Every few hundred milliseconds, the scheduler updates the deficit of each job by looking at how many tasks each job had running during this interval vs. its fair share. Whenever a task slot becomes available, it is assigned to the job with the highest deficit. There is one exception - if there were one or more jobs who were not meeting their pool capacity guarantees, we only choose among these "needy" jobs (based again on their deficit), to ensure that the scheduler meets pool guarantees as soon as possible. The fair shares are calculated by dividing the capacity of the cluster among runnable jobs according to a "weight" for each job. By default the weight is based on priority, with each level of priority having 2x higher weight than the next (for example, VERY_HIGH has 4x the weight of NORMAL). However, weights can also be based on job sizes and ages, as described in the Configuring section. For jobs that are in a pool, fair shares also take into account the minimum guarantee for that pool. This capacity is divided among the jobs in that pool according again to their weights. Finally, when limits on a user's running jobs or a pool's running jobs are in place, we choose which jobs get to run by sorting all jobs in order of priority and then submit time, as in the standard Hadoop scheduler. Any jobs that fall after the user/pool's limit in this ordering are queued up and wait idle until they can be run. During this time, they are ignored from the fair sharing calculations and do not gain or lose deficit (their fair share is set to zero).