Re: running job with giraph dependency anomaly
Yeah. I haven't changed anything with the standard Giraph stuff. I just made my own vertex and and VertexInputFormat. We are in a 64bit environment. . .is it possible that building a jar with 32bit tools would be a problem? I wouldn't think so, since that addressing native-dependency issues was sort of the *point* of java. . .but, this seems really odd to me. Are there some dependency restrictions that I should know about? We have to use Jackson 1.6 (because we use cloudera distribution of hadoop), and there are other libraries we use. Thx again for the feedback. -David On 2/7/12 8:08 PM, "Avery Ching" wrote: >If you're using GiraphJob, the mapper class should be set for you. >That's weird. > >Avery > >On 2/7/12 5:58 PM, David Garcia wrote: >> That's interesting. Yes, I don't need native libraries. The problem >>I'm >> having is that after I run job.waitForCompletion(..), >> The job runs a mapper that is something other than GraphMapper. It >> doesn't complain that a Mapper isn't defined or anything. It runs >> something else. As I mentioned below, the map-class doesn't appear to >>be >> defined. >> >> >> On 2/7/12 7:50 PM, "Jakob Homan" wrote: >> >>> That's not necessarily a bad thing. Hadoop (not Giraph) has native >>> code library it can use for improved performance. You'll see this >>> message when running on a cluster that's not been deployed to use the >>> native libraries. If I follow what you wrote, most likely your work >>> project cluster is so configured. Unless you actively expect to have >>> the native libraries loaded, I wouldn't be concerned. >>> >>> >>> On Tue, Feb 7, 2012 at 5:46 PM, David Garcia >>> wrote: I am running into a weird error that I haven't seen yet (I suppose I've been lucky). I see the following in the logging: org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable In the job definition, the property "mapreduce.map.class" is not even defined. For Giraph, this is usually set to "mapreduce.map.class=org.apache.giraph.graph.GraphMapper" I'm building my project with hadoop 0.20.204. When I build the GiraphProject myself (and run my own tests with the projects dependencies), I have no problems. The main difference is that I'm using a Giraph dependency in my work project. All input is welcome. Thx!! -David >
Re: running job with giraph dependency anomaly
If you're using GiraphJob, the mapper class should be set for you. That's weird. Avery On 2/7/12 5:58 PM, David Garcia wrote: That's interesting. Yes, I don't need native libraries. The problem I'm having is that after I run job.waitForCompletion(..), The job runs a mapper that is something other than GraphMapper. It doesn't complain that a Mapper isn't defined or anything. It runs something else. As I mentioned below, the map-class doesn't appear to be defined. On 2/7/12 7:50 PM, "Jakob Homan" wrote: That's not necessarily a bad thing. Hadoop (not Giraph) has native code library it can use for improved performance. You'll see this message when running on a cluster that's not been deployed to use the native libraries. If I follow what you wrote, most likely your work project cluster is so configured. Unless you actively expect to have the native libraries loaded, I wouldn't be concerned. On Tue, Feb 7, 2012 at 5:46 PM, David Garcia wrote: I am running into a weird error that I haven't seen yet (I suppose I've been lucky). I see the following in the logging: org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable In the job definition, the property "mapreduce.map.class" is not even defined. For Giraph, this is usually set to "mapreduce.map.class=org.apache.giraph.graph.GraphMapper" I'm building my project with hadoop 0.20.204. When I build the GiraphProject myself (and run my own tests with the projects dependencies), I have no problems. The main difference is that I'm using a Giraph dependency in my work project. All input is welcome. Thx!! -David
Re: running job with giraph dependency anomaly
That's interesting. Yes, I don't need native libraries. The problem I'm having is that after I run job.waitForCompletion(..), The job runs a mapper that is something other than GraphMapper. It doesn't complain that a Mapper isn't defined or anything. It runs something else. As I mentioned below, the map-class doesn't appear to be defined. On 2/7/12 7:50 PM, "Jakob Homan" wrote: >That's not necessarily a bad thing. Hadoop (not Giraph) has native >code library it can use for improved performance. You'll see this >message when running on a cluster that's not been deployed to use the >native libraries. If I follow what you wrote, most likely your work >project cluster is so configured. Unless you actively expect to have >the native libraries loaded, I wouldn't be concerned. > > >On Tue, Feb 7, 2012 at 5:46 PM, David Garcia >wrote: >> I am running into a weird error that I haven't seen yet (I suppose I've >> been lucky). I see the following in the logging: >> >> org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop >> library for your platform... using builtin-java classes where applicable >> >> >> In the job definition, the property "mapreduce.map.class" is not even >> defined. For Giraph, this is usually set to >> "mapreduce.map.class=org.apache.giraph.graph.GraphMapper" >> >> I'm building my project with hadoop 0.20.204. >> >> When I build the GiraphProject myself (and run my own tests with the >> projects dependencies), I have no problems. The main difference is that >> I'm using a Giraph dependency in my work project. All input is welcome. >> Thx!! >> >> -David >>
Re: running job with giraph dependency anomaly
That's not necessarily a bad thing. Hadoop (not Giraph) has native code library it can use for improved performance. You'll see this message when running on a cluster that's not been deployed to use the native libraries. If I follow what you wrote, most likely your work project cluster is so configured. Unless you actively expect to have the native libraries loaded, I wouldn't be concerned. On Tue, Feb 7, 2012 at 5:46 PM, David Garcia wrote: > I am running into a weird error that I haven't seen yet (I suppose I've > been lucky). I see the following in the logging: > > org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop > library for your platform... using builtin-java classes where applicable > > > In the job definition, the property "mapreduce.map.class" is not even > defined. For Giraph, this is usually set to > "mapreduce.map.class=org.apache.giraph.graph.GraphMapper" > > I'm building my project with hadoop 0.20.204. > > When I build the GiraphProject myself (and run my own tests with the > projects dependencies), I have no problems. The main difference is that > I'm using a Giraph dependency in my work project. All input is welcome. > Thx!! > > -David >
running job with giraph dependency anomaly
I am running into a weird error that I haven't seen yet (I suppose I've been lucky). I see the following in the logging: org.apache.hadoop.util.NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable In the job definition, the property "mapreduce.map.class" is not even defined. For Giraph, this is usually set to "mapreduce.map.class=org.apache.giraph.graph.GraphMapper" I'm building my project with hadoop 0.20.204. When I build the GiraphProject myself (and run my own tests with the projects dependencies), I have no problems. The main difference is that I'm using a Giraph dependency in my work project. All input is welcome. Thx!! -David
Re: the slides for my talk @ FOSDEM
Hi Sebastian, thanks for your feedback on the slides. As a matter of fact I'm aware of the pegasus matrix-based optimization as of the shimmy technique by Jimmy Lin. I thought that this kind of technique is general enough for all iterative graph algorithms, not just PR, and mostly using the naive algorithm would just help me out explaining the presentation. Messaging the adjacent vertices from the Mapper by iterating over them and emitting (otherVertex, myPartialPR) maps easily to our messaging paradigm. I'll maybe make it more clear in the next presentation. Thanks! On Mon, Feb 6, 2012 at 2:54 PM, Sebastian Schelter wrote: > Hi Claudio, > > nice job with the slides! I have only one small point to criticize: > > When PageRank is implemented with MapReduce, it's not necessary to have > the graph passed through in each iteration. Mahout for example uses > power iterations where the adjacency matrix is multiplied by the > pagerank vector and only that vector has to be sent over the network. > Pegasus uses a similar approach. > > /s > > > > On 06.02.2012 12:24, Claudio Martella wrote: >> Hello guys, >> >> for those interested, here are the "slides" for my talk at FOSDEM. >> >> http://prezi.com/9ake_klzwrga/apache-giraph-distributed-graph-processing-in-the-cloud/ >> >> The event was very nice, a tight community and a great interest in >> Giraph. Isabel Drost, one of the organizers of Berlin Buzzwords, >> invited the talk there. Jakob, are you still planning to talk there? >> Maybe we can split Kafka/Giraph talks? >> >> Best, >> Claudio >> > -- Claudio Martella claudio.marte...@gmail.com