RE: Resources or advice on minimising memory usage in Giraph/Hadoop code ?

2012-06-07 Thread David Garcia
ever reason, it may not be necessary to instantiate them all. From: Claudio Martella [claudio.marte...@gmail.com] Sent: Thursday, June 07, 2012 1:35 AM To: user@giraph.apache.org Subject: Re: Resources or advice on minimising memory usage in Giraph/Hadoop cod

Re: Resources or advice on minimising memory usage in Giraph/Hadoop code ?

2012-06-07 Thread André Kelpe
Hi! One interesting jvm option I learned about lately is -XX:+UseCompressedStrings, which will use a byte [] for all strings, that are fully defined in ASCII. Given that you are working with URIs, I assume that this is true for most of your strings, so I would give it a shot. For more info on JVM

Re: Resources or advice on minimising memory usage in Giraph/Hadoop code ?

2012-06-06 Thread Claudio Martella
Won't this just postpone the pain? On Thursday, June 7, 2012, David Garcia wrote: > Based upon what you have mentioned, o think you are getting heap errors > because every vertex in your graph will be loaded into memory prior to > super step one. So if you have a large graph, with lots of state

Re: Resources or advice on minimising memory usage in Giraph/Hadoop code ?

2012-06-06 Thread David Garcia
Based upon what you have mentioned, o think you are getting heap errors because every vertex in your graph will be loaded into memory prior to super step one. So if you have a large graph, with lots of state, you probably have memory issues from the very beginning. A simple way to mitigate the

Re: Resources or advice on minimising memory usage in Giraph/Hadoop code ?

2012-06-06 Thread Avery Ching
No article or book, but here's a few tips. 1) Use aggregators! This can drastically can reduce the amount of memory use by combining messages on the server side. 2) -Dmapred.child.java.opts="-Xss128k" or some other value (should affect the RPC threads or netty threads) 3) You'll want to minimi