Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

[YUNIKORN-2919] Document spark-operator: Account spark.memory.offHeap.size #508

Merged
merged 2 commits into from
Dec 24, 2024
Merged
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
4 changes: 2 additions & 2 deletions docs/user_guide/workloads/run_spark.md
Original file line number Diff line number Diff line change
@@ -30,7 +30,7 @@ under the License.
:::note
Pre-requisites:
- This tutorial assumes YuniKorn is [installed](../../get_started/get_started.md) under the namespace `yunikorn`
- Use spark-operator version >= 2.0 to enable support for YuniKorn gang scheduling
- Use spark-operator version >= 2.1.0 to enable support for YuniKorn gang scheduling
:::

### Install YuniKorn
@@ -48,7 +48,7 @@ helm install yunikorn yunikorn/yunikorn --create-namespace --namespace yunikorn
We should install `spark-operator` with `controller.batchScheduler.enable=true` and set `controller.batchScheduler.default=yunikorn` to enable Gang Scheduling. It's optional to set the default scheduler to YuniKorn since you can specify it later on, but it's recommended to do so.
Also, note that the total requested memory for the Spark job is the sum of memory requested for the driver and that for all executors, where each is computed as below:
* Driver requested memory = `spark.driver.memory` + `spark.driver.memoryOverhead`
* Executor requested memory = `spark.executor.memory` + `spark.executor.memoryOverhead` + `spark.executor.pyspark.memory`
* Executor requested memory = `spark.executor.memory` + `spark.executor.memoryOverhead` + `spark.executor.pyspark.memory` + `spark.memory.offHeap.size`

```shell script
helm repo add spark-operator https://kubeflow.github.io/spark-operator