CODEWITHSUNDEEP
hadoop-yarnschedulinghadoop2
Sarriman
Sarriman

Reputation: 422

Explanation for Scheduling Categories in Hadoop

Trying to categorize Hadoop's schedulers (not only the official FIFO-Capacity-Fair) i seem to find that there is no clear understanding as to what each of the 4 categories that a Scheduler can belong to, are defining as a behavior control. The 4 Categories i seem to conclude are:

• Static Scheduling

• Dynamic Scheduling

• Resource Availability

• Time Awareness

Can someone point me into a direction or a source i am not aware of, in order to find out more, or clarify on it's own the above?

Upvotes: 0

Views: 59

Answers (1)

Sarriman
Sarriman

Reputation: 422

While researching into this more i came to the following conclusion, which i am posting in order to help.

A Scheduler can belong in 2 Categories. Static Scheduling(FIFO,Fair,Delay,Capacity) and Dynamic Scheduling(Deadline Constraint, Resource Aware).

After choosing the desired scheduler, the tuning of various properties inside YARN's code, can define if that Scheduler will work based on Resource Availability(Delay Scheduler, Matchmaking Scheduler, Resource Aware) or based in the Time(Deadline Constraint, Delay) constraint.

Again this is an understanding of mine through some research that i have done.

Upvotes: 1

Related Questions