Describe your issue here.
List relevant info:
- OS distribution, version
- TensorHive and API version
- Versions of installed dependencies
- Essential hardware specs (GPU model)
Tell us how to reproduce this issue from the ground up.
commands to execute
You can also put a bunch of screenshots, GIF or Asciinema recording
Tell us what should happen
Tell us what happens instead
Dear Micmarty and Tensorhive Authors,
Thank you for managing the issue report on Tensorhive. Currently, I am trying to establish a connection of Tensorhive to GPU servers using 2 AWS Ubuntu instances 18.04 ( For example named as A and B). I have installed Tensorhive version 18.04 and tried to generate the Tensorhive key in instance A and copy over to other instance B for SSH establishment and also define the host_name, user_name and port in hosts_config.ini file . However, somehow, I wouldn't be able to connect through SSH. May I know should I install Tensorhive into both instances or only one instance A which is a management instance to monitor the GPU statistic data from instance B and others?
Please let me share our GPU servers architecture where we have 15 GPU servers and 6 DGX stations where we hope Tensorhive could help to manage.
Could I have one more question here that does it need us to install Tensorhive into all of our GPU servers as distributed network?
Thank you and look forward to hearing from you.
Best Regards,
William Le
William Le