Re: Graph job stuck after map 100% completed

2016-02-17 Thread khaled . ammar
Hi Ni‎shant,It looks like there is no enough memory to hold the data. How much aggregate memory available in your cluster? Thanks, Khaled

Re: Giraph out-of-core feature is not helping!

2015-12-10 Thread Khaled Ammar
8:56 PM, Roman Shaposhnik wrote: > I remember seeing a discussion that discouraged users from > utilizing OOC. I am vague on the details but you can try > searching the archives. > > Thanks, > Roman. > > On Fri, Dec 4, 2015 at 4:34 PM, Khaled Ammar > wrote: > > Hi

Giraph out-of-core feature is not helping!

2015-12-04 Thread Khaled Ammar
Hi, I am using Giraph-1.1.0 to do large graph processing. I was trying to do a hashMin (WCC) algorithm on a large graph but it failed with out of memory error. I thought the out-of-core option may help, but it did not. Is there any advice about how to enable out-of-core processing? I followed th

Why failed jobs do not exit !

2015-09-23 Thread Khaled Ammar
Hi all, I am using Giraph 1.1.0. When a Giraph job fails, e.g. due to not enough memory, all mappers fail except one. Therefore, the job do not exit. I have to kill it with "hadoop job -kill " command. Does any one know why this happens and how to avoid that? -- Thanks, -Khaled

Re: How to accelerate performance of Giraph Job?

2015-07-24 Thread Khaled Ammar
Hi Udbhav, Please tell us more about your cluster setup, and your Hadoop version and configuration. Thanks, -Khaled On Fri, Jul 24, 2015 at 9:16 AM, Udbhav Agarwal wrote: > Hi, > > Am running Giraph Job for 1,00,000 connected vertices for shortest path > calculation. It is taking around 50

Re: InMemoryVertex Format(s)

2015-06-04 Thread Khaled Ammar
ZT6R8t4Vw1hrBcWWTghG_XgaC%3DYqrg%40mail.gmail.com%3E > > In my experience, Blocks framework is much easier to use and it naturally > suit your needs. > > > > > On Wed, Jun 3, 2015 at 11:02 AM, Khaled Ammar > wrote: > >> Thank you Sergey, >> >> This is exactly what

Re: InMemoryVertex Format(s)

2015-06-03 Thread Khaled Ammar
te more on why do you want to use I/O formats? You can use > different computation classes within one application and you don't need to > do I/O between them. All intermediate results can be kept in vertex and > edge data. > > Regards, > Sergey Edunov > > On Tue, Jun 2, 20

InMemoryVertex Format(s)

2015-06-02 Thread Khaled Ammar
Hi all, There are InMemory input and output format for giraph. These could be useful when a specific computation should be executed until convergence and then another computation is needed. Instead of writing intermediate results to HDFS and read it again, InMemoryVertex format sounds very conveni

Re: Is it possible to traverse a linear list (line graph) parallely in giraph?

2014-12-30 Thread Khaled Ammar
I don't think Giraph is suitable for this task because in this case you probably want to visit graph vertices in order which leaves no chance for parallelization. In fact, even running a shortest path query on such graph will not perform as good as it would in web or social network graphs. These

looking for a User guide

2014-09-18 Thread Khaled Ammar
Hi all, I was looking for a user guide for Giraph 1.0.0 or 1.1.0 but I could not find one with an example about writing your first Giraph application. Could anyone point me to one if exist? -- Thanks, -Khaled

Dynamic graphs vs Graph Mutation

2014-09-15 Thread Khaled Ammar
Hi all, I understand that Pregel supports graph mutations when vertices issue requests to make changes in the graph topology (add/remove vertex/edge). I have two questions here: 1- Does the same concept applies to supporting dynamic graphs? (Graphs that change over time where the add/remove verte