How to setup mapreduce not to use yarn

[#1]

When I execute a mapreduce job (MR2), it is using YARN and I can see the job/operation details in Cloudera Navigator for sourceType:YARN. But instead I want to execute the mapreduce job(MR2) without using YARN so that the operation  details in Cloudera Navigator will be with sourceType:MAPREDUCE.

Could you please provide the configuration changes to be done in Cloudera for this scenario?

Also, please provide the configuration details to configure MR1. I am using Cloudera version of CDH 5.13.

[#2]

I am not an expert, but I thought that Mapreduce2 is an a.k.a for YARN, so the framework uses YARN resource manager and node managers. So the identification of the application is correct in Navigator.

[#3]

Yes, but I still want to have some data to be created with sourceType:Mapreduce. So, I have to execute by mapreduce job not to use Yarn. Please let me know the configuration to achive this.

[#4 · Accepted solution]

Hello Nukala,

YARN (MapReduce 2 Included) is the recommended for usage. MapReduce is included in YARN for backward compatibility.

The only configuration to access MapReduce is to change the "Alternatives Priority" property under YARN's configuration from 92 to 91.

To change the Alternatives Priority property, do the following:

1. Go to the MapReduce or YARN service.

2. Click the Configuration tab.

3. Select Scope > Gateway Default Group.

4. Select Category > All.

5. Type Alternatives in Search box.

6. In the Alternatives Priority property, set the priority value.

7. Click Save Changes to commit the changes.

8. Redeploy the client configuration.

The link to the document covering changing of priorities is located here:

Configuring Alternatives Priority for Services Dependent on MapReduce

[#5]

Thanks.