Skip to content

Conversation

@rxin
Copy link
Contributor

@rxin rxin commented Nov 23, 2016

What changes were proposed in this pull request?

TaskMemoryManager has a memory leak detector that gets called at task completion callback and checks whether any memory has not been released. If they are not released by the time the callback is invoked, TaskMemoryManager releases them.

The current error message says something like the following:

WARN  [Executor task launch worker-0]
org.apache.spark.memory.TaskMemoryManager - leak 16.3 MB memory from
org.apache.spark.unsafe.map.BytesToBytesMap@33fb6a15
In practice, there are multiple reasons why these can be triggered in the normal code path (e.g. limit, or task failures), and the fact that these messages are log means the "leak" is fixed by TaskMemoryManager.

To not confuse users, this patch downgrade the message from warning to debug level, and avoids using the word "leak" since it is not actually a leak.

How was this patch tested?

N/A - this is a simple logging improvement.

@rxin
Copy link
Contributor Author

rxin commented Nov 23, 2016

cc @JoshRosen

@SparkQA
Copy link

SparkQA commented Nov 23, 2016

Test build #69058 has finished for PR 15989 at commit 46b5af9.

  • This patch passes all tests.
  • This patch merges cleanly.
  • This patch adds no public classes.

@hvanhovell
Copy link
Contributor

LGTM. Merging to master/branch-2.1. Thanks!

asfgit pushed a commit that referenced this pull request Nov 23, 2016
## What changes were proposed in this pull request?
TaskMemoryManager has a memory leak detector that gets called at task completion callback and checks whether any memory has not been released. If they are not released by the time the callback is invoked, TaskMemoryManager releases them.

The current error message says something like the following:
```
WARN  [Executor task launch worker-0]
org.apache.spark.memory.TaskMemoryManager - leak 16.3 MB memory from
org.apache.spark.unsafe.map.BytesToBytesMap33fb6a15
In practice, there are multiple reasons why these can be triggered in the normal code path (e.g. limit, or task failures), and the fact that these messages are log means the "leak" is fixed by TaskMemoryManager.
```

To not confuse users, this patch downgrade the message from warning to debug level, and avoids using the word "leak" since it is not actually a leak.

## How was this patch tested?
N/A - this is a simple logging improvement.

Author: Reynold Xin <[email protected]>

Closes #15989 from rxin/SPARK-18557.

(cherry picked from commit 9785ed4)
Signed-off-by: Herman van Hovell <[email protected]>
@asfgit asfgit closed this in 9785ed4 Nov 23, 2016
robert3005 pushed a commit to palantir/spark that referenced this pull request Dec 2, 2016
## What changes were proposed in this pull request?
TaskMemoryManager has a memory leak detector that gets called at task completion callback and checks whether any memory has not been released. If they are not released by the time the callback is invoked, TaskMemoryManager releases them.

The current error message says something like the following:
```
WARN  [Executor task launch worker-0]
org.apache.spark.memory.TaskMemoryManager - leak 16.3 MB memory from
org.apache.spark.unsafe.map.BytesToBytesMap33fb6a15
In practice, there are multiple reasons why these can be triggered in the normal code path (e.g. limit, or task failures), and the fact that these messages are log means the "leak" is fixed by TaskMemoryManager.
```

To not confuse users, this patch downgrade the message from warning to debug level, and avoids using the word "leak" since it is not actually a leak.

## How was this patch tested?
N/A - this is a simple logging improvement.

Author: Reynold Xin <[email protected]>

Closes apache#15989 from rxin/SPARK-18557.
uzadude pushed a commit to uzadude/spark that referenced this pull request Jan 27, 2017
## What changes were proposed in this pull request?
TaskMemoryManager has a memory leak detector that gets called at task completion callback and checks whether any memory has not been released. If they are not released by the time the callback is invoked, TaskMemoryManager releases them.

The current error message says something like the following:
```
WARN  [Executor task launch worker-0]
org.apache.spark.memory.TaskMemoryManager - leak 16.3 MB memory from
org.apache.spark.unsafe.map.BytesToBytesMap33fb6a15
In practice, there are multiple reasons why these can be triggered in the normal code path (e.g. limit, or task failures), and the fact that these messages are log means the "leak" is fixed by TaskMemoryManager.
```

To not confuse users, this patch downgrade the message from warning to debug level, and avoids using the word "leak" since it is not actually a leak.

## How was this patch tested?
N/A - this is a simple logging improvement.

Author: Reynold Xin <[email protected]>

Closes apache#15989 from rxin/SPARK-18557.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants