Topics

1

Python and Java API / [Cluster] Does the node save the project when it's task is finished ?

« on: June 14, 2017, 06:13:26 PM »

Hello,

I am trying the run exportDenseCloud through the cluster.

When a node finished this task, does the node save the project even if this task does not modify the current state of the project ?


More specifically, I am trying to parallize the export dense cloud task with the next build ortho task, so I need to be sure that the export dense cloud will not save the project, but only the build ortho task will do

Thanks in advance
 

2

Python and Java API / [Cluster] RunScript task error message

« on: April 18, 2017, 11:37:43 AM »

Hello admin,

I am currently trying to use the RunScript task to run some customized script in the cluster. When I get an error, I try to make an exception in the python script, but when I check the task status, the task's last error is always "Can't run script".

So I would like to know how can I get the exact error message when I check the batch/task status if there's an error ?

for info : I am using version 1.2.6

Thanks in advance

3

General / Performance 1.2.5 vs 1.3.0 and gpu usage

« on: March 16, 2017, 12:52:29 PM »

Hello, admin

I was trying to test the performance between 1.2.5 and 1.3.0.  I have a dataset with 200 images, and I ran Alignement and buildDenseCloud.

I ran my tests on AWS EC2 instances with 4 different types of instances:

ec2Type    cpu  gpu      gpuType        ram( G)

g2.2xlarge   8   1   NVIDIA GRID K25     15   
g2.8xlarge   32   4   NVIDIA GRID K25     60

p2.xlarge   4   1   NVIDIA Tesla K80    61
p2.8xlarge   8   8   NVIDIA Tesla K80    488



I found that alignment is much faster on v1.3.0 than v1.2.5, which is very exciting !

But for the build dense cloud, I get some wired results, I found that the v1.3 is faster than v1.2.5 on g2.2xlarge and p2.xlarge (both with one GPU),  but much slower on p2.8xlarge (8 gpus) and g2.8xlarge (4 gpus).

I observed the gpu usage on these instances, I found that, for v1.3.0, only 10 - 15% of gpu is used on p2.8xlarge (8 gpus) and only 30 - 40 % gpu is used on g2.8xlarge (4 gpus).  But the gpu usage are higher on v1.2.5 on these 2 instances.

4

Python and Java API / [cpu_enable] means using all the cpu when gpu are working ?

« on: March 13, 2017, 03:58:13 PM »

Hello, Admin,

I saw the change of 1.3 about the [cpu_enable] changed from [cpu_cores_inactive].

I am wondering [cpu_enable] means that photoscan will use all the cpu when gpu is working ? so it does not leave any cpu to operate with the gpu ? Or it means anything else ?

Thanks

5

General / All the available tasks in Network mode ?

« on: February 16, 2017, 07:38:35 PM »

Hi,

Could someone give me a complete list of all the tasks we can run in the network mode ?


Thanks in advance !

6

Python and Java API / Could i pass an error message when i do app.quit() ?

« on: January 15, 2017, 12:41:44 AM »

I am recently launching Photoscan GUI in my nodejs program (by using child_process in nodejs) to make some process. If the process finished or failed, it will use app.quit() to exit Photoscan GUI and continue my nodejs program.

But the thing is that i want to pass the error message when i do app.quit(), like app.quit('My error message'). I tried using sys.exit('my error message'), bug it did not work.


Regards.

7

General / [Python] Could I pass a error message when i do "app.quit()" ?

« on: January 15, 2017, 12:40:44 AM »

I am recently launching Photoscan GUI in my nodejs program (by using child_process in nodejs) to make some process. If the process finished or failed, it will use app.quit() to exit Photoscan GUI and continue my nodejs program.

But the thing is that i want to pass the error message when i do app.quit(), like app.quit('My error message'). I tried using sys.exit('my error message'), bug it did not work.


Regards.

8

Feature Requests / [Cluster] Distribute nodes equally to each batches when priority is the same

« on: October 26, 2016, 01:14:12 AM »

Hello dear admin,

I am working on the photoscan cluster project. I made a test in which I tried to launch 2 batches (the same priority) to the server with 2 nodes, and each batches can be cut to 10 small jobs (the waiting nodes are 10 accroding to photoscan network monitor).

So I observed that the 2 nodes were always distributed to the first batch and the second batch was always in status "pending", until that the waiting nodes of the first batch down to 0 and then the second batch began to work ...

I am wondering if it is possible to ajust this nodes distributions mechanism to distribute nodes equally to each batches (if possible) when each batches' priority are the same ?

Thanks in advance
Best regards 

9

Feature Requests / [Network] add the set node priority option in the node command line

« on: June 20, 2016, 11:24:16 AM »

Hello,

I am wondering if there's a way to set the node priority easily without using monitor or python script ?

thanks a lot.