stanfordgpsusers

Hi, I try to run PageRank example according to the documentation, https://sites.google.com/site/

Hi Huangxun, No I haven't and I do not plan to. Actually I have also stopped maintaining GPS any

Hi, For my program, I need to store a double array for each node (vertex). For this I am creating

Hi all, I am slightly confused with executing a code in GPS. The documentation shows the execute the

Hi all, Recently I installed GPS by following its documentation. I was trying to run the example

Hi Raj, Just copy-pasting a reply I sent to another user: " By default GPS simply does random

Hi Sunita, By default GPS simply does random partitioning. So you don't have to do anything extra

thank you sir i found it helpful. :) On Wed, May 27, 2015 at 3:24 PM, Semih Salihoglu <

I'm not sure. This looks like a potential jar issue, where the system isn't able to to find

In the attachments there are the log files of the workers! I cannot find any output file neither in

Unfortunately, I am unable to develop the codebase further these days but if you send a patch, I

Hi Tung, I'm not sure. Somehow the phase is not moving to AVERAGE_COMPUTATION_STAGE. So for some

Thanks, Semih, It works smoothly. Regards, Tung, On 9/18/14, 4:09, Semih Salihoglu wrote: Hi Tung,

Hi! According to the paper linked in GPS home page, fault tolerance is implemented through

Hi Semih, I would like to know how many messages could be reduced if I turn on LALP. Specifically, I

Dear all, I've spend a whole day to debug my program, just now, I have caught the bug! I want to

Sorry for mistype the head. It's "About static partition" 在 2014年5月29日星期四UTC+8下午12时41分

Probably SCCBaseMaster. semih On Sun, May 18, 2014 at 8:03 PM, Yi Lu <luyi...@gmail.com> wrote

Hi Lyuwei, LALP doesn't decrease the overall memory consumed by GPS across all machines but it

Thank you very much. I have just found a file called debug_monitoring_runner_manifest.txt and the pre

Make the IDs contiguous from 0 to numVertices. So if you have 1000 vertices, they should have IDs

Hi Sujian, A good and easiest approximate is probably to look at the master's logs and find the

Did you update the Hadoop dependencies in ~/gps/trunk/debug_monitoring_runner_manifest.txt and

TwoSync was fine to use but after a while we started focusing on OneSync. So we may have introduced

Thanks for your answer. I find the default value is 1000ms, which is much smaller than 3 seconds. By

Hi Lyuwei, Looking at your double.cfg file, I expect that you have "master" and "

​You know that XMS indicates the start memory and XMX indicates maximum memory; so they can be

Hi, Semih, Thanks for this info, that seems in C++, it will be good if there's an example

Hi, All, Thanks, but I tried all the suggested port numbers, no one works yet, will check further.

Hi Wang, For jobs that already completed, you should use the offline monitoring website to look at