...

...

Discussion thread: http://apache-flink-mailing-list-archive.1008284.n3.nabble.com/DISCUSS-FLIP-49-Unified-Memory-Configuration-for-TaskExecutors-td31436.html

JIRA:

...

Please keep the discussion on the mailing list rather than commenting on the wiki (wiki discussions get unwieldy fast).

...

Task executor memory configuration options. (See Memory Pools and Configuration Keys)A summary of backwards compatibility  

Proposed Changes

Unifying Managed Memory for Batch and Streaming

The basic idea is to consider memory used by RocksDB state backends as part of managed memory, and extend memory manager so that state backends memory consumers can simply reserve certain amount of memory from it but not necessarily allocate the memory from it. In this way, users should be able to switch between streaming and batch jobs, without having to modify the cluster configurations.

Memory Use Cases

...

• Streaming jobs with RocksDBStateBackend
• Off-heap memory
• Implicitly allocated by the state backend
• Cannot exceed total memory size, which is configured during initialization
• Batch jobs
• OffEither on-heap or off-heap memory
• Explicitly allocated from the memory manager
• Cannot exceed total memory allocated from memory manager

To make managed memory work with both cases, we should always allocate managed memory off-heap.

Unifying Explicit and Implicit Memory Allocation

• Memory consumers can acquire memory in two ways
• Explicitly acquire from MemoryManager, in the form of MemorySegment.
• Reserve from MemoryManager, in which case should return “use up to X bytes”, and implicitly allocate the memory by the consumer itself.
• MemoryManager never pre-allocate any memory pages, so that we keep the managed memory budget available for both allocation from MemoryManager and allocation directly from memory consumers.
• For off-heap memory explicitly acquired from MemoryManager, Flink always allocate with Unsafe.allocateMemory(), which is not limited by the JVM -XX:MaxDirectMemorySize parameter.
• This eliminates the uncertainty about how many off-heap managed memory should be accounted for JVM max direct memory. 
• The fallback drawback is that Unsafe is no longer supported in Java 12.

Separate On-Heap and Off-Heap Memory Pools for Managed Memory

Currently (Flink 1.9), all managed memory are allocated with the same type, either on-heap or off-heap. This is good with the current use cases, where we do not necessary need both on-heap and off-heap managed memory in the same task executor.

With the design in this proposal, memory usage of state backends is also considered as managed memory, which means we may have scenarios where jobs in the same cluster need different types of managed memory. E.g., a streaming job with MemoryStateBackend and another streaming job with RocksDBStateBackend.

Therefore, we separate the managed memory pool into the on-heap pool and the off-heap pool. We use an off-heap fraction to decide what fraction of managed memory should go into the off-heap pool, and leave the rest to the on-heap pool. Users can still configure the cluster to use all on-heap / off-heap managed memory by setting the off-heap fraction to 0 / 1.

Memory Pools and Configuration Keys

Image Removed

Framework Heap Memory

Memory Pools and Configuration Keys

Image Added

Framework Heap Memory

On-heap memory for the Flink task manager framework. It is not accounted for slot resource profiles.

(taskmanager.memory.framework.heap.size)

(default 128mb)

Framework Off-Heap Memory

OffOn-heap memory for the Flink task manager framework. It is not accounted for slot resource profiles.

(taskmanager.memory.framework.off-heap.size)

(default 128mb)

Task Heap Memory

...

(taskmanager.memory.task.heap.size)

Task Off-Heap Memory

Off-heap memory for user code.

(taskmanager.memory.task.offheapoff-heap.size) 

(default 0b)

Network Memory

Off-heap memory for shuffle service, e.g., network buffers.

(taskmanager.memory.network.[min/max/fraction]) or (taskmanager.network.memory.[min/max/fraction])

(default min=64mb, max=1gb, fraction=0.1)

Managed Memory

OnOff-heap and off-heap Flink managed memory.

(taskmanager.memory.managed.[size|fraction])(taskmanager.memory.managed.offheap-fraction)

(default fraction=0.5, offheap-fraction=0.0)

On-Heap Managed Memory = Managed Memory * (1 - offheap-fraction)

Off-Heap Managed Memory = Managed Memory * offheap-fraction

4)

JVM Metaspace

Off-heap memory for JVM metaspace.

(taskmanager.memory.jvm-metaspace)

(default 192mb96mb)

JVM Overhead

Off-heap memory for thread stack space, I/O direct memory, compile cache, etc.

(taskmanager.memory.jvm-overhead.[min/max/fraction])

(default min=128mb192mb, max=1gb, fraction=0.1)

Total Flink Memory

Coarser config option for total flink memory, to make it easily configurable for users.

This includes Framework Heap Memory, Framework Off-Heap Memory, Task Heap Memory, Task Off-Heap Memory, Network Memory Network Memory, and Managed Memory.

This excludes JVM Metaspace and JVM Overhead.

(taskmanager.memory.flink.size)

Total Process Memory

...

• JVM heap memory
• Includes Framework Heap Memory, Task Heap Memory, and On- Heap Managed Memory
• Explicitly set both  -Xmx and -Xms to this value
• JVM direct memory
  • Includes Framework Off-Heap Memory, Task Off-heap Memory , and Network Memory and JVM Overhead
  • Explicitly set -XX:MaxDirectMemorySize to this value
  • For Off-heap Managed Memory, we always allocate memory with Unsafe.allocateMemory(), which will not be limited by this parameter.
• JVM metaspace
• Set -XX:MaxMetaspaceSize to configured JVM Metaspace

...

• If both Task Heap Memory and Managed Memory are configured, we use these to derive Total Flink Memory
• If Network Memory Memory is configured explicitly, we use that value
• Otherwise, we compute it such that it makes up the configured fraction of the final Total Flink Memory (see getAbsoluteOrInverseFraction())
• If Total Flink Memory is configured, but not Task Heap Memory and Managed Memory, then we deriveNetwork Memory Network Memory and Managed Memory, and leave the rest (excluding Framework Heap Memory, Framework Off-Heap Memory and  and Task Off-Heap Memory) as Task Heap Memory.
• IfNetwork  Network Memory is configured explicitly, we use that value
• Otherwise we compute it such that it makes up the configured fraction of the Total Flink Memory (see getAbsoluteOrFraction())
• If Managed Memory is configured explicitly, we use that value
• Otherwise we compute it such that it makes up the configured fraction of the Total Flink Memory (see getAbsoluteOrFraction())
• If only the Total Process Memory is configured, we derive the Total Flink Memory in the following way
• We get (or compute relative) and subtract the JVM Overhead from Total Process Memory (see getAbsoluteOrFraction())
• We subtract JVM Metaspace from the remaining
• We leave the rest as Total Flink Memory

...

• Introduce new configuration options
• Introduce TaskExecutorSpecificsdata structures and utilities.
  • Data structure to store memory / pool sizes of task executor
  • Utility for calculating memory / pool sizes from configuration
  • Utility for generating dynamic configurations
  • Utility for generating JVM parameters

...

Step 3. Launch task executor with new memory calculation logics

• Invoke TaskExecutorSpecifics data structures and utilities introduced in Step 2 to generate JVM parameters and dynamic configurations for launching new task executors.
  • In startup scripts
  • In resource managers
• Task executor uses TaskExecutorSpecifics to uses data structures and utilities introduced in Step 2 to set memory pool sizes and slot resource profiles.
  • MemoryManager
  • ShuffleEnvironment
  • TaskSlotTable

...

Compatibility, Deprecation, and Migration Plan

This FLIP changes how users configure cluster resources, which in some cases may require re-configuring of cluster if migrated from prior versions.

Deprecated configuration keys are as follows:

Test Plan

• We need to update existing and add new integration tests dedicated to validate the new memory configuration behaviors.
• It is also expected that other regular integration and end-to-end tests should fail if this is broken.

Limitations

• The proposed design uses Unsafe.allocateMemory() for allocating managed memory, which is no longer supported Java 12. We need to look for alternative solutions in the future.

Follow Ups

• This FLIP requires very good documentation to help users understand how to properly configure Flink processes and which knobs to turn in which cases.
• It would be good to expose configured memory pool sizes in the web UI, so that users see immediately what amount of memory TMs assume to use for what purpose.

Rejected Alternatives

Regarding JVM direct memory, we have the following alternative.

...