giraph
giraph copied to clipboard
apache
Mirror of Apache Giraph
Part of the overhead comes from using an atomicinteger that's written to for every entry addition. I don't see any particular value in this keeping this counter, which incurs high...
Three compressed memory efficient OutEdges implementations to hold integer IDs. All three implementations are discussed in: Panagiotis Liakos, Katia Papakonstantinopoulou, Alex Delis: Realizing Memory-Optimized Distributed Graph Processing. IEEE Trans. Knowl....
https://issues.apache.org/jira/browse/GIRAPH-1190 The current implementation takes incoming messages stored as ByteArrayOneMessageToManyIds and prepares them as a map from partition id to a ByteArrayVertexIdMessages, which holds the messages for the corresponding partition....
https://issues.apache.org/jira/browse/GIRAPH-1189 The PageRank implementation uses one-to-many messaging to reduce the amount of traffic sent. The corresponding message request (SendWorkerOneMessageToManyRequest) creates ByteArrayVertexIdMessages instances before adding the messages to the Message Store....
Replaced `com.google.code.findbugs:annotations:2.0.2` with `com.github.stephenc.findbugs:findbugs-annotations:1.3.9-1` and `com.google.code.findbugs:jsr305:2.0.2`. Had to remove a couple annotations that don't exist in the older version of the drop-in replacement. There is a newer version of com.github.stephenc.findbugs:findbugs-annotations,...
…global logic Summary: Some supersteps don't do anything with vertices but just do global worker or master computation or perform aggregation. Not iterating through vertices in these cases can save...
This diff contains the following fixes: 1) Once the OOC is at OFFLAODING state, it may offload data to disk indefinitely. We should call GC manually and get OOC out...
