Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Server crashes when you use more than three services #90

Closed
aninja5 opened this issue Mar 26, 2016 · 7 comments
Closed

Server crashes when you use more than three services #90

aninja5 opened this issue Mar 26, 2016 · 7 comments

Comments

@aninja5
Copy link

aninja5 commented Mar 26, 2016

I created four different services using thesample models. When I try to access all four of the services , the server exits with the message "killed". It happens when I am consequently hitting the services. It works fine with three services. But when I use the fourth one, the server crashes. I am running deepdetect on a Ubuntu 64 Bit VM with 8GB memory.

@beniz
Copy link
Collaborator

beniz commented Mar 26, 2016

Hi, this is very likely that your system is simply running out of memory and server gets killed by kernel. I guess you are using a CPU, and thus you need to increase your system's RAM.

@aninja5
Copy link
Author

aninja5 commented Mar 26, 2016

Working with 12GB now. Will let you know what happens.

@Isaacpm
Copy link

Isaacpm commented Mar 26, 2016

@aninja5 check the system logs, if the process is being killed by the OS there should be an entry there

@aninja5
Copy link
Author

aninja5 commented Mar 26, 2016

@beniz @Isaacpm You guys are right. I checked kern.log. There is an out of memory entry for deep detect process. What is the recommended configuration? Any work around? I am thinking perhaps I could move the system to AWS.

@Isaacpm
Copy link

Isaacpm commented Mar 26, 2016

There is this value when you start training a job: Memory required for data: 82504000, which should give you the amount of memory required so you can plan accordingly.
For AWS you can check this issue:
https://github.com/beniz/deepdetect/issues/5

What I did was to get a cheap ovh.co.uk server with 32gb of ram, which is cheaper than AWS for long term running, and it works for me. But my jobs aren't that resource intensive.

@beniz
Copy link
Collaborator

beniz commented Mar 26, 2016

User @revilokeb did an outstanding review of dd performances on AWS along with AMI and recommendations, I'd suggest to use that, it is in issue #5.

@aninja5 aninja5 closed this as completed Mar 26, 2016
@aninja5
Copy link
Author

aninja5 commented Mar 26, 2016

@beniz @Isaacpm thanks! I'll follow up if I find issues!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants