1. Adds the option of compressing outputs.
2. Adds batching to the serialization to prevent OOM on the read side.
3. Slight renaming of config options.
4. Use Spark's buffer size for reads in addition to writes.
External Sorting for Aggregator and CoGroupedRDDs (Revisited)
(This pull request is re-opened from https://github.com/apache/incubator-spark/pull/303, which was closed because Jenkins / github was misbehaving)
The target issue for this patch is the out-of-memory exceptions triggered by aggregate operations such as reduce, groupBy, join, and cogroup. The existing AppendOnlyMap used by these operations resides purely in memory, and grows with the size of the input data until the amount of allocated memory is exceeded. Under large workloads, this problem is aggravated by the fact that OOM frequently occurs only after a very long (> 1 hour) map phase, in which case the entire job must be restarted.
The solution is to spill the contents of this map to disk once a certain memory threshold is exceeded. This functionality is provided by ExternalAppendOnlyMap, which additionally sorts this buffer before writing it out to disk, and later merges these buffers back in sorted order.
Under normal circumstances in which OOM is not triggered, ExternalAppendOnlyMap is simply a wrapper around AppendOnlyMap and incurs little overhead. Only when the memory usage is expected to exceed the given threshold does ExternalAppendOnlyMap spill to disk.
Aside from trivial formatting changes, use nulls instead of Options for
DiskMapIterator, and add documentation for spark.shuffle.externalSorting
and spark.shuffle.memoryFraction.
Also, set spark.shuffle.memoryFraction to 0.3, and spark.storage.memoryFraction = 0.6.
- When a resourceOffers() call has multiple offers, force the TaskSets
to consider them in increasing order of locality levels so that they
get a chance to launch stuff locally across all offers
- Simplify ClusterScheduler.prioritizeContainers
- Add docs on the new configuration options
throughout the docs: SPARK_VERSION, SCALA_VERSION, and MESOS_VERSION.
To use them, e.g. use {{site.SPARK_VERSION}}.
Also removes uses of {{HOME_PATH}} which were being resolved to ""
by the templating system anyway.
instead of the maximum number of outstanding fetches. This should make
it faster when there are many small map output files, as well as more
robust to overallocating memory on large map outputs.
- Rework/expand the nav bar with more of the docs site
- Removing parts of docs about EC2 and Mesos that differentiate between
running 0.5 and before
- Merged subheadings from running-on-amazon-ec2.html that are still relevant
(i.e., "Using a newer version of Spark" and "Accessing Data in S3") into
ec2-scripts.html and deleted running-on-amazon-ec2.html
- Added some TODO comments to a few docs
- Updated the blurb about AMP Camp
- Renamed programming-guide to spark-programming-guide
- Fixing typos/etc. in Standalone Spark doc
which can be compiled via jekyll, using the command `jekyll`. To compile
and run a local webserver to serve the doc as a website, run
`jekyll --server`.
