Available computers and servers

Greiner has two GPUs.

login to greiner:

• ip : ssh -X login@greiner.enstb.org
• login : ton_login_ecole

check the sate of the gpu with command

 nvidia-smi

For geiner, the output is :

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 396.24                 Driver Version: 396.24                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla K80           Off  | 00000000:06:00.0 Off |                    0 |
| N/A   50C    P0    58W / 149W |      0MiB / 11441MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   1  Tesla K80           Off  | 00000000:07:00.0 Off |                    0 |
| N/A   36C    P0    74W / 149W |      0MiB / 11441MiB |    100%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|    0     93895      C   python3                                    10941MiB |
|    1     93895      C   python3                                    10863MiB |
+-----------------------------------------------------------------------------+

Par défaut, Tensorflow alloue l'ensemble de la mémoire GPU au programme, même s'il n'en nécessite que 10%. Il existe une option permettant d'allouer seulement la mémoire GPU nécessaire. Il faut ajouter les lignes suivantes:

   config = tf.ConfigProto()
   config.gpu_options.allow_growth = True
   sess = tf.Session(config=config)

Login to gazebo:

• you must first connect to the School VPN. Official instructions are on https://intranet.telecom-bretagne.eu/page.php?idContenu=5061 :
  • On linux you can install `vpnc` and use the following configuration in `/etc/vpnc/tb.conf`:

    • IKE DH Group dh2
      IPSec gateway 192.108.116.206
      IPSec ID vpnrire
      IPSec secret rire
      Xauth username

    • Then run `sudo vpnc tb` to start the vpn
  • On macos, Open up your System Preferences and select “Network”. Click on the little + button at the bottom of the window to create a new connection.Pick “VPN” for the Interface and set the VPN type to “Cisco IPSec”. It doesn't matter what you set as the service name. Click on “create”. As Server address, input “192.108.116.206” and in the account name and passwords use your login and password for the school identification. As authentication settings, input the group name “vpnrire” and the shared secret is “rire”.
• You can now ssh to gazebo like this:
  • `ssh -X login@gazebo.enstb.org`
  • where login: ton_login_ecole

Update 20180802: Where is the desktop??

To login, you need first to send Jerome or Mai your ssh public key. Then, you can connect by ssh laborobo@backup1.enstb.org

Contents:

• VM (Virtual machines)
  • UbuntuRobot : Ubuntu virtual machine with gazebo (turtlebot) and ros installed. Description in Simulateur Gazebo
• Keraal