Details
-
Type:
New Feature
-
Status: Resolved
-
Priority:
Major
-
Resolution: Later
-
Affects Version/s: None
-
Fix Version/s: None
-
Component/s: benchmarks
-
Labels:None
-
Target Version/s:
-
Release Note:Introducing PUMA benchmark suite
-
Tags:benchmarks
Description
A benchmark suite which represents a broad range of "real-world" MapReduce applications exhibiting application characteristics with high/low computation and high/low shuffle volumes. These benchmarks have been published as part of MaRCO (http://dx.doi.org/10.1016/j.jpdc.2012.12.012) project in JPDC '12.
Activity
- All
- Comments
- Work Log
- History
- Activity
- Transitions
Thanks Karthik, I have uploaded the patch for the first example. Will it be better if I create the patch for all benchmarks and upload them or wait for this one to pass through review and commit and then proceed with the other ones?
Hey Faraz, thanks for sharing the links. Looks like there are quite a few new examples you implemented. It might be easier to create one subtask per example for easier review and commit.
Here is a link to the document for free download (http://web.ics.purdue.edu/~fahmad/papers/marco.pdf).
The benchmarks are available to view separately as http://web.ics.purdue.edu/~fahmad/benchmarks.htm
Here is the link to download the code and input data sets (http://web.ics.purdue.edu/~fahmad/benchmarks/datasets.htm).
Once committed, they should be available through the release.
Reading the link provided in the description requires a payment of US$39.95.
Removing the fix versions; usually, those are set by the committer at commit time.
There hasn't been any traffic on this jira in some time. [~algol], are you still interested in getting this benchmark suite into Hadoop? Please comment on you interest in finishing up this work and I'll be happy to keep this ticket open. Otherwise, I'll close this ticket at the end of the week.