| Issue 112: | MapReduce support | |
| 395 people starred this issue and may be notified of changes. | Back to list |
It would be really great if there was some way of using MapReduce principles to do background task execution. I know this was mentioned by me in the context of background task execution, but I just wanted to give it its own thread for better follow- up. Google might well provide background task execution that has nothing to do with MapReduce. Perhaps the number of stars on this issue will also indicate whether others are interested in using MapReduce as well. |
|
,
Apr 10, 2008
(No comment was entered for this change.)
Labels: -Type-Defect Type-Feature
|
|
,
Apr 11, 2008
Web 2.0 is all about having smart apps. Having MapReduce would make it easy for us developers to make our apps smart. |
|
,
May 02, 2008
A long discussion touching on the subject and showing that mapreduce may address or may make things easier: http://groups.google.com/group/google- appengine/browse_thread/thread/74abbe3d70a50d20/7e9cb422f7b5ea7e ./alex -- .w( the_mindstorm )p. Alexandru Popescu |
|
,
May 29, 2008
Python MapReduce tasks - it seems the most "Googley" solution for scalable background tasks that manipulate the datastore. By "Googley" I mean -- it would be scalable, meterable, aligned with the current APIs, and would obsolete most of the current tickets for: background tasks, batch processing, schema management. ( Issue #6 plus several others) |
|
,
Jun 02, 2008
eventually google needs to give us some kind of tool for data mining, right now it's very easy to load google with data but not to bulk-retrieve, or get some valuable information from our info |
|
,
Jul 08, 2008
please? |
|
,
Jul 08, 2008
For an example of a database with MapReduce, check out the latest trunk of CouchDB. Maybe having MapReduce in the views/indicies of GAE would be good. |
|
,
Jul 23, 2008
Without group by, aggregate functions in bigtable, map-reduce is the only way to do reports and analytics. |
|
,
Sep 25, 2008
(No comment was entered for this change.)
Status: Acknowledged
|
|
,
Nov 30, 2008
It was hinted at qCon that GAE is to eventually allow long running background jobs. Mapreduce seems a perfect fit. |
|
,
Apr 15, 2009
I think it's a necessity for folks with lot's of data, or lot's of traffic. |
|
,
Apr 15, 2009
It would be an interesting feature if it could be integrated like Amazon's Simple MapReduce Service. It seems right now that AppEngine is purely towards the serving side of Web Apps. |
|
,
May 06, 2009
This is no doubt a big step on providing such services on the cloud. Reporting and data mining are reality in any systems, and this could be the solution as it is already applied to Google's current structure. Maybe this can even be implemented in a lower layer that would work more transparent to the end-developers, with simple integrated methods to adhere to it. |
|
,
May 20, 2009
+1 I have to build some kind of reports and the lack of aggregation operators in GQL will force me to compute the aggregates manually by 1000-entities chunks, with a cron to build the aggregate incrementally. Map-Reduce would be perfect for this. |
|
 Sign in to add a comment
|