Status

Current state: Under Discussion

Discussion thread: here

JIRA: 

Released: TBD

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

Motivation

Generally speaking, applications may consist of one or more jobs, and they may want to share the data with others. In Flink, the jobs in the same application are independent and share nothing among themselves. If a Flink application involves several sequential steps, each step (as an independent job) will have to write its intermediate results to an external sink, so that its results can be used by the following step (job) as sources.

Although functionality wise this works, this programming paradigm has a few shortcomings:

1. In order to share a result, a sink must be provided.

2. Complicated applications become inefficient due to large amount of IO on intermediate result.

3. User experience is weakened for users using programing API (SQL users are not victims here because the temporary tables are created by the framework)

It turns out that interactive programming support is critical to the user experience on Flink in batch processing scenarios. The following code gives an example:

Table a = ….
Table b = a.select(...).filter(...)
int c = b.max()
int d = b.min() // recompute table b from table a
Table c = b.select((‘f1 - min)/(max - min)).filter(...)
c.print() // recompute table b from table a
...
If (b.count() > 10) { // recompute table b from table a
  b.select(UDF1(...)).print()// recompute table b from table a
} else {
  b.select(UDF2(...)).print()
}

In the above code, because b is not cached, it will be computed from scratch multiple times whenever referred later in the program.

To address the above issues, we propose to add support for interactive programming in Flink Table API.

Public Interfaces

1. Add the following two new methods to the Flink Table class.

  /**
    * Cache this table to builtin table service or the specified customized table service.
    *
    * This method provides a hint to Flink that the current table maybe reused later so a
    * cache should be created to avoid regenerating this table.
    *
    * The following code snippet gives an example of how this method could be used.
    *
    * {{{
    *   val t = tEnv.fromCollection(data).as('country, 'color, 'count)
    *
    *   val t1 = t.filter('count < 100).cache()
    *   // t1 is cached after it is computed for the first time.
    *   val x = t1.collect().size
    *
    *   // When t1 is used again to compute t2, it may not be re-computed.
    *   val t2 = t1.groupBy('country).select('country, 'count.sum as 'sum)
    *   val res2 = t2.collect()
    *   res2.foreach(println)
    *
    *   // Similarly when t1 is used again to compute t3, it may not be re-computed.
    *   val t3 = t1.groupBy('color).select('color, 'count.avg as 'avg)
    *   val res3 = t3.collect()
    *   res3.foreach(println)
    *
    * }}}
    *
    * @note Flink optimizer may decide to not use the cache if doing that will accelerate the
    * processing, or if the cache is no longer available for reasons such as the cache has
    * been invalidated.
    * @note The table cache could be create lazily. That means the cache may be created at 
    * the first time when the cached table is computed.
    * @note The table cache will be cleared when the user program exits.
    *
    * @return the current table with a cache hint. The original table reference is not modified
    *               by the execution of this method.
    */
  def cache(): Table


  /**
    * Manually invalidate the cache of this table to release the physical resources. Users are
    * not required to invoke this method to release physical resource unless they want to. The
    * table caches are cleared when user program exits.
    *
    * @note After invalidated, the cache may be re-created if this table is used again.
    */
  def invalidateCache(): Unit

2. Add a close method to the TableEnvironment

/**
  * Close and clean up the table environment.
  */
void close();

Proposed Changes

Cache intermediate results

As mentioned in the motivation section. The key idea of the FLIP is to allow the intermediate process results to be cached, so later references to that result does not result in duplicate computation. To achieve that, we need to introduce Cached Tables.

The cached tables are tables whose contents are saved by Flink as the user application runs. A cached Table can be created in two ways:

Explicit caching

• Users can call cache() method on a table to explicitly tell Flink to cache a Table.
• The cache() method returns a new Table object with a flag set.
• The cache() method does not execute eagerly. Instead, the table will be cached when the DAG that contains the cached table runs.

Auto caching

• As the application runs, if Flink can save an intermediate result with little cost, Flink will do that, even users did not call cache() explicitly. Such case typically occurs at shuffle boundaries.
• Auto caching will be enabled by default but could be turned off by users.

Semantic of cache() method

The semantic of the cache() method is a little different depending on whether auto caching is enabled or not.

• When auto caching is enabled (default behavior)

  Code Block
  language java title Auto caching is enabled (default)
  TableEnvironment tEnv = ... Table t1 = ... Table t2 = t1.cache() ... tEnv.execute() // t1 is cached. Table t3 = t1.select(...) // cache will be used. Table t4 = t2.select(...) // cache will be used. ... // The following two lines of code are equivalent t1.invalidateCache() // cache will be released t2.invalidateCache() // cache will be released ... // The following two lines of code are equivalent. t1.select() // cache will be recreated t2.select() // cache will be recreated

  
  

• When auto caching is disabled

  Code Block
  language java title Auto caching is disabled
  TableEnvironment tEnv = ... Table t1 = ... Table t2 = t1.cache() ... tEnv.execute() // t1 is cached. Table t3 = t1.select(...) // cache will NOT be used. <----- difference Table t4 = t2.select(...) // cache will be used. ... // The following two lines of code are equivalent t1.invalidateCache() // cache will be released t2.invalidateCache() // cache will be released ... t1.print() // cache will not be recreated <-------- different t2.print() // cache will be recreated

  
  

Scope of the cached result

The cached tables are available to the user application using the same TableEnvironment.

Release the cached results

The cached intermediate results will consume some resources and needs to be released eventually. The cached result will be released in two cases.

User application exits

When TableEnvironment is closed, the resources consumed by the cached tables will also be released. This usually happens when user application exits.

Explicit invalidateCache() invocation

Sometimes users may want to release the resource used by a cached table before the application exits. In this case, users can call invalidateCache() on a table. This will immediately release the resources used to cache that table.

Explicitly ignore the cached intermediate result

In some rare cases, users may want to explicitly ignore the cached intermediate result. In this case, user needs to give an explicit hint, such as:

table.hint("ignoreCache")

Right now Flink does not have a hint mechanism yet. So before such a hint mechanism is available. Users are not able to explicitly ignore a cached intermediate result.

Cache a stream table

Theoretically speaking, user can also cache a streaming table. The semantic will be store the result somewhere (potentially with a TTL). However, caching a streaming table is usually not that useful. For simplicity, we would not support stream table caching in the first implementation. When cache() is invoked on a stream table, it will be treated as a No-Op. This leaves us the room to add caching for stream tables in the future without asking users to change their code.

Implementation Details

To let the feature available out of the box, a default file system based cache service will be provided. This section describes the implementation details of the default table service.

Although the implementation details are transparent to the users, there are some changes to Flink internals to make it work.

Default Cache Service

The architecture is illustrated below:

Each cached table consists of two pieces of information:

• Table metadata - name, location, etc.
• Table contents - the actual contents of the table

Default table service stores the metadata in client (e.g. TableEnvironment --> CacheServiceManager) and saves the actual contents in the cache service instances running inside the Task Managers.

The end to end process is the following:

Step 1: Start Cache Service

1. When a DAG of JOB_1 is submitted, the TableEnvironment (planner) checks the DAG to see if any node should be cached. 
2. If there is a node to be cached, the TableEnvironment (CacheServiceManager) will first start a Cache Service job (if the job has not started). To start a cache service job,
   1. The TableServiceManager opens a ServerSocket and listens on that port.
   2. The TableServiceManager submits a Cache Service job.
      1. The Cache Service job contains only source nodes which runs Cache Server code.
      2. The parallelism of the Cache Service job equals to the number of TMs so that each TM has one Cache Service instance running in it.
   3. Each Cache Service instance in the TM will open a server socket and register their HOST:PORT information back to the port opened by TableServiceManager in step 2a.
3. The CacheServiceManager waits until it collects all the HOST:PORT pairs from the Table Service Instances.

Step 2: Execute JOB_1 (write cached tables)

1. The TableEnvironment adds a CacheServiceSink to each cached node (a node with a corresponding Cached Table) in the DAG of JOB_1.
   1. The CacheServiceSink acts as a normal sink, as if user called writeToSink(CacheServiceSink).
   2. Each cached node will be assigned a intermediate result UUID. The UUID is used to tag the when the CacheServiceSink writes the intermediate result to the Cache Service instances.
   3. Since there will be multiple Cache Service instances, the intermediate result will be partitioned and stored in a distributed manner.
2. The TableEnvironment submits the modified DAG to the Flink cluster.
3. After the JOB_1 finishes, the CacheServiceManager stores the [LogicalNode → intermediate result UUID] in memory.

Step 3: Execute JOB_2 (read cached tables)

1. Later on, if JOB_2 is submitted and the DAG contains a node that has already been cached (a mapping to intermediate result UUID is found), TableEnvironment replaces the node and the connected ancestor nodes with a CacheServiceSource.
   1. The CacheServiceSource has the [HOST:PORT] pairs of all the Cache Service instances.
   2. The CacheServiceSource will be reading from the Cache Service instances using the UUID corresponding to the cached intermediate results.
2. When invalidateCache() is invoked, the CacheServiceManager will directly make an RPC call to each Cache Service instance to delete the corresponding intermediate result. It will also remove the LogicalNode to intermedate result UUID mapping.

Step 4: Clean up 

1. When the application exits, the TableEnvironment will be closed.
2. The CacheServiceManager will be closed when TableEnvironment is closed. It will stop the Cache Service job. 
3. All the cached tables will be released when the Cache Service job stops.

Failover

If a Cache Service instance fails, since Cache Service is a Flink job. The failed subtask will be brought up again. However, all the intermediate result which has a partition on the failed subtask will become unavailable.

In this case, the CacheServiceSource / CacheServiceSink will throw an exception and the job will fail. As a result, CacheServiceManager invalidates the caches that are impacted. The TableEnvironment will resubmit the original DAG without using the cache. Note that because there is no cache available, the TableEnvironment (planner) will add a CacheServiceSink to the cached nodes, therefore the cache will be recreated after the execution of the original DAG.

The above process is transparent to the users.

Impact to optimization

When users explicitly caches a table, the DAG is implicitly changed. This is because the optimizer may decide to change the cached node if it were not explicitly cached. As of now, when a node has more than one downstream nodes in the DAG, that node will be computed multiple times. This could be solved by introducing partial optimization for subgraphs, which is a feature available in Blink but not in Flink yet. This is an existing problem in Flink and this FLIP does not attempt to address that.

How to run the cache services?

We propose to run the cache service instances using a Flink Job. Doing this has a few benefits:

1. Leaving the cache service to user space decouples it from the Flink runtime.
2. Flink job is a powerful primitive in Flink. This means the cache service can reuse many existing mechanism
   
   1. resource specifications
   2. monitoring framework
   3. lifecycle management
   4. state management
   5. fault tolerance 

Integration with external shuffle service (FLIP-31)

To achieve auto caching, the cache service needs to be integrated with shuffle service. This is sort of a natural move. Inherently, external shuffle service and cache service share a lot of similarities:

With a few extensions, the default cache service could also serve as the default external shuffle service. The architecture would look like something below:

The cache service will serve as a storage for both cache and shuffle data. This means both the storage tier and the RPC protocol will support SubPartitions.

Compatibility, Deprecation, and Migration Plan

This FLIP proposes a new feature in Flink. It is fully backwards compatible.

Test Plan

Unit tests and Integration Tests will be added to test the proposed functionalities.

Rejected Alternatives

The semantic of the cache() / invalidateCache() API has gone through extended discussions. The rejected alternative semantics are documented below:

Rejected API Option 1

API

// Do not return anything
void cache();
// Alternatively, return the original table for chaining purpose.
Table cache();
 
// Physically uncache the table to release resource
void uncache();
 
// To explicitly ignore cache, not part of this proposal but in the future
Table hint(“ingoreCache”)

Semantic

TableEnvironment tEnv = ...
Table t = …
...
// cache the table
t.cache();
…
// Assume some execution physically created the cache.
tEnv.execute()
...
// The optimizer decide whether the cache will be used or not.
t.foo();
...
// Physically release the cache.
t.uncache();

Pros

• Simple and intuitive, users only need to deal one variable of Table class

Cons

• Side effect: a table may be cached / uncached in a method invocation, while the caller does not know about this.

  Code Block
  language java title Side Effect Option 1
  { Table t = … // cache the table t.cache(); ... // The cache may be released in foo() foo(t); ... // cache is no longer available t.bar() } void foo(Table t) { // cache the table t.cache(); …. // Physically drop the table t.uncache(); }

  
  

Rejected API Option 2

API

// Return CachedTable
CachedTable cache();
 
public interface CachedTable extends Table {
  // Physically uncache the table to release resource
  void uncache();
}

Semantic

TableEnvironment tEnv = ...
Table t = …
...
// cache the table
CachedTable cachedT = t.cache();
…
// Assume some execution physically created the cache.
tEnv.execute()
...
// Use the original DAG.
t.foo();
...
// Use cache
cachedT.bar()
...
// Physically release the cache.
t.uncache();

Pros

• No side effect

Cons

• Optimizer has no chance to kick in.

• Users have to distinguish between original table / cached table.

• Adding auto cache becomes a backward incompatible change.

Rejected API Option 3

API

// Return a CacheHandle
CacheHandle cache();
 
public interface CacheHandle {
  // Physically release cache resource
  void release();
}

Semantic

TableEnvironment tEnv = ...
Table t = …
...
// cache the table
CacheHandle handle1 = t.cache();
CacheHandle handle2 = t.cache;
…
// Assume some execution physically created the cache.
tEnv.execute()
...
// Optimizer decide whether cache will be used.
t.foo();
...
// Release the first handle, cache will not be released because handle 2 has not been released
handle1.release();
// Release the second handle, the cache will be released because all the cache handle have been released.
handle2.release();

Pros

• No side effect

• Users only deal with the variable.

• Easy to add auto caching.

Cons

• The behavior of t.foo() changes after t.cache(), the concern is that this is considered as “modifies” table t, which is against the immutable principle.