I am new to EMR and I am running an EMR cluster, with 1 master (32gb) and 5 core nodes (16gb). I launch 11 apps. The apps have to be separated in case one of them fail (all of them are streaming apps). I must mention that I also got ElasticSearch running on the cluster.
After some time the master node is running out of memory and stops responding and some apps starting to fail. In the process overview I found many smaller hadoop processes with that occupy 1-1.3GB of RAM. I guess these are the driver processes from each app. I tried to reduce the the driver memory under "spark.driver.memory" to 512MB, but it's still at 1.3GB after relaunching the apps. Is this because of yarn?
ES just allocates ca. 6.5 GB of RAM of the master node
I had to specify the driver memory in spark-submit command like this:
spark-submit --driver-memory 500M
because to specify it inside the python file is too late, when you run the driver in client mode, because it allocates the memory before
I'm trying to run multiple spark clients on Airflow(ETL scheduler).
I'm running in cluster mode on YARN, therefore ApplicationMaster Executor and Driver are all running on executor in Yarn context.
However, my spark client which sample the process and monitor the state is running in airflow worker.
The problem is that the Spark client take lot's of memory ~500 MB per job. It may sound as not much in terms of executors or drivers but for the role of spark client it sounds crazy.
My question is, how can I configure/manipulate spark client memory/cpu requirements can I limit it's intervals ? can I limit it's memory with flags?
So in spark code it make a distinction if it's running in standalone mode or cluster mode. For standalone it set a default of -Xmx 1G and in cluster mode it doesn't have default but it trying to read java options from environment variable called SPARK_SUBMIT_OPTS.
So if you wanna set any java opts for the client java process only use SPARK_SUBMIT_OPTS
For me the only difference would be that in the first case the Driver would be inside the App Master JVM whereas it would be next to it the above client mode configuration. What would be different in those two cases ?
Spark deployment mode - client vs cluster
Client mode
It's interactive. If you want to get a job result (dynamic analysis) at your machine (client - the Driver program), client deployment mode is quite useful.
Easier for developing/debugging.
End user have the control where the Driver Program is running
Always up application: expose your Spark job launcher as REST service or a Web UI.
Cluster mode
Easier for resource allocation (let the master decide): Fire and forget
Due to this Fire-N-Forget behavior, It's good to submit long, resource hungry job in cluster mode.
Monitoring Driver Program is done from Master Web UI like other workers. (not from client machine)
Stop at the end: once job is finished, allocated resources are freed automatically.
In cluster mode, if driver program has some issue, YARN restarts the driver without killing the executors.
In client mode, YARN automatically kills all executors if your driver is killed.
The spark docs have the following paragraph that describes the difference between yarn client and yarn cluster:
There are two deploy modes that can be used to launch Spark applications on YARN. In cluster mode, the Spark driver runs inside an application master process which is managed by YARN on the cluster, and the client can go away after initiating the application. In client mode, the driver runs in the client process, and the application master is only used for requesting resources from YARN.
I'm assuming there are two choices for a reason. If so, how do you choose which one to use?
Please use facts to justify your response so that this question and answer(s) meet stackoverflow's requirements.
There are a few similar questions on stackoverflow, however those questions focus on the difference between the two approaches, but don't focus on when one approach is more suitable than the other.
A common deployment strategy is to submit your application from a gateway machine that is physically co-located with your worker machines (e.g. Master node in a standalone EC2 cluster). In this setup, client mode is appropriate. In client mode, the driver is launched directly within the spark-submit process which acts as a client to the cluster. The input and output of the application is attached to the console. Thus, this mode is especially suitable for applications that involve the REPL (e.g. Spark shell).
Alternatively, if your application is submitted from a machine far from the worker machines (e.g. locally on your laptop), it is common to use cluster mode to minimize network latency between the drivers and the executors. Note that cluster mode is currently not supported for Mesos clusters. Currently only YARN supports cluster mode for Python applications." -- Submitting Applications
What I understand from this is that both strategies use the cluster to distribute tasks; the difference is where the "driver program" runs: locally with spark-submit, or, also in the cluster.
When you should use either of them is detailed in the quote above, but I also did another thing: for big jars, I used rsync to copy them to the cluster (or even to master node) with 100 times the network speed, and then submitted from the cluster. This can be better than "cluster mode" for big jars. Note that client mode does not probably transfer the jar to the master. At that point the difference between the 2 is minimal. Probably client mode is better when the driver program is idle most of the time, to make full use of cores on the local machine and perhaps avoid transferring the jar to the master (even on loopback interface a big jar takes quite a bit of seconds). And with client mode you can transfer (rsync) the jar on any cluster node.
On the other hand, if the driver is very intensive, in cpu or I/O, cluster mode may be more appropriate, to better balance the cluster (in client mode, the local machine would run both the driver and as many workers as possible, making it over loaded and making it that local tasks will be slower, making it such that the whole job may end up waiting for a couple of tasks from the local machine).
Conclusion :
To sum up, if I am in the same local network with the cluster, I would
use the client mode and submit it from my laptop. If the cluster is
far away, I would either submit locally with cluster mode, or rsync
the jar to the remote cluster and submit it there, in client or
cluster mode, depending on how heavy the driver program is on
resources.*
AFAIK With the driver program running in the cluster, it is less vulnerable to remote disconnects crashing the driver and the entire spark job.This is especially useful for long running jobs such as stream processing type workloads.
Spark Jobs Running on YARN
When running Spark on YARN, each Spark executor runs as a YARN container. Where MapReduce schedules a container and fires up a JVM for each task, Spark hosts multiple tasks within the same container. This approach enables several orders of magnitude faster task startup time.
Spark supports two modes for running on YARN, “yarn-cluster” mode and “yarn-client” mode. Broadly, yarn-cluster mode makes sense for production jobs, while yarn-client mode makes sense for interactive and debugging uses where you want to see your application’s output immediately.
Understanding the difference requires an understanding of YARN’s Application Master concept. In YARN, each application instance has an Application Master process, which is the first container started for that application. The application is responsible for requesting resources from the ResourceManager, and, when allocated them, telling NodeManagers to start containers on its behalf. Application Masters obviate the need for an active client — the process starting the application can go away and coordination continues from a process managed by YARN running on the cluster.
In yarn-cluster mode, the driver runs in the Application Master. This means that the same process is responsible for both driving the application and requesting resources from YARN, and this process runs inside a YARN container. The client that starts the app doesn’t need to stick around for its entire lifetime.
yarn-cluster mode
The yarn-cluster mode is not well suited to using Spark interactively, but the yarn-client mode is. Spark applications that require user input, like spark-shell and PySpark, need the Spark driver to run inside the client process that initiates the Spark application. In yarn-client mode, the Application Master is merely present to request executor containers from YARN. The client communicates with those containers to schedule work after they start:
yarn-client mode
This table offers a concise list of differences between these modes:
Reference: https://blog.cloudera.com/blog/2014/05/apache-spark-resource-management-and-yarn-app-models/ - Apache Spark Resource Management and YARN App Models (web.archive.com mirror)
In yarn-cluster mode, the driver program will run on the node where application master is running where as in yarn-client mode the driver program will run on the node on which job is submitted on centralized gateway node.
Both answers by Ram and Chavati/Wai Lee are useful and helpful, but here I just want to added a couple of other points:
Much has been written about the proximity of driver to the executors, and that is a big factor. The other factors are:
Will the driver process be around until execution of job is finished?
How's returned data being handled?
#1. In spark client mode, the driver process must be up and running the whole time when the spark job is in execution. So if you have a truly long job that say take many hours to run, you need to make sure the driver process is still up and running, and that the driver session is not auto-logout.
On the other hand, after submitting a job to run in cluster mode, the process can go away. The cluster mode will keep running. So this is typically how a production job will run: the job can be triggered by a timer, or by an external event and then the job will run to its completion without worrying about the lifetime of the process submitting the spark job.
#2. In client mode, you can call sc.collect() to gather all the data back from all executors, and then write/save the returned data to a local Linux file on local disk. Now this may not work for cluster mode, as the 'driver' typically run in a different remote host. The data written up therefore need to be persisted in a common mounted volume (such as GPFS, NFS) or in distributed file system like HDFS. If the job is running under Hadoop/YARN, the more common way for cluster mode is simply ask each executor to persist the data to HDFS, and not to run collect( ) at all. Collect() actually have scalability issue when there are a large number of executors returning large amount of data - it can overwhelm the driver process.
The doc https://spark.apache.org/docs/1.1.0/submitting-applications.html
describes deploy-mode as :
--deploy-mode: Whether to deploy your driver on the worker nodes (cluster) or locally as an external client (client) (default: client)
Using this diagram fig1 as a guide (taken from http://spark.apache.org/docs/1.2.0/cluster-overview.html) :
If I kick off a Spark job :
./bin/spark-submit \
--class com.driver \
--master spark://MY_MASTER:7077 \
--executor-memory 845M \
--deploy-mode client \
./bin/Driver.jar
Then the Driver Program will be MY_MASTER as specified in fig1 MY_MASTER
If instead I use --deploy-mode cluster then the Driver Program will be shared among the Worker Nodes ? If this is true then does this mean that the Driver Program box in fig1 can be dropped (as it is no longer utilized) as the SparkContext will also be shared among the worker nodes ?
What conditions should cluster be used instead of client ?
No, when deploy-mode is client, the Driver Program is not necessarily the master node. You could run spark-submit on your laptop, and the Driver Program would run on your laptop.
On the contrary, when deploy-mode is cluster, then cluster manager (master node) is used to find a slave having enough available resources to execute the Driver Program. As a result, the Driver Program would run on one of the slave nodes. As its execution is delegated, you can not get the result from Driver Program, it must store its results in a file, database, etc.
Client mode
Want to get a job result (dynamic analysis)
Easier for developing/debugging
Control where your Driver Program is running
Always up application: expose your Spark job launcher as REST service or a Web UI
Cluster mode
Easier for resource allocation (let the master decide): Fire and forget
Monitor your Driver Program from Master Web UI like other workers
Stop at the end: one job is finished, allocated resources are freed
I think this may help you understand.In the document https://spark.apache.org/docs/latest/submitting-applications.html
It says " A common deployment strategy is to submit your application from a gateway machine that is physically co-located with your worker machines (e.g. Master node in a standalone EC2 cluster). In this setup, client mode is appropriate. In client mode, the driver is launched directly within the spark-submit process which acts as a client to the cluster. The input and output of the application is attached to the console. Thus, this mode is especially suitable for applications that involve the REPL (e.g. Spark shell).
Alternatively, if your application is submitted from a machine far from the worker machines (e.g. locally on your laptop), it is common to use cluster mode to minimize network latency between the drivers and the executors. Note that cluster mode is currently not supported for Mesos clusters or Python applications."
What about HADR?
In cluster mode, YARN restarts the driver without killing the executors.
In client mode, YARN automatically kills all executors if your driver is killed.