From fa4bb5ce925ba611dac09a3edb6be04031417760 Mon Sep 17 00:00:00 2001
From: Nicola Piga <nicolapiga@gmail.com>
Date: Tue, 4 Jul 2023 23:55:33 +0200
Subject: [PATCH] [handover] Remove docker compose README.md and move yml file
 to root/dockercompose

---
 .../docker-compose.yml                        |  0
 src/handover/dockercompose/README.md          | 88 -------------------
 2 files changed, 88 deletions(-)
 rename {src/handover/dockercompose => dockercompose}/docker-compose.yml (100%)
 delete mode 100644 src/handover/dockercompose/README.md

diff --git a/src/handover/dockercompose/docker-compose.yml b/dockercompose/docker-compose.yml
similarity index 100%
rename from src/handover/dockercompose/docker-compose.yml
rename to dockercompose/docker-compose.yml
diff --git a/src/handover/dockercompose/README.md b/src/handover/dockercompose/README.md
deleted file mode 100644
index 147d8f3..0000000
--- a/src/handover/dockercompose/README.md
+++ /dev/null
@@ -1,88 +0,0 @@
-## Setup up the GPU-enabled swarm (once)
-
-### Manager
-On the chosen manager machine please do:
-
-```console
-docker swarm init
-```
-
-### Workers
-
-For each worker to be added, on the `manager` please do:
-
-```console
-docker swarm join-token worker
-```
-
-The output will be something like
-```console
-docker swarm join --token <a_token> <manager_ip>:<manager_port>
-```
-
-Simply copy-paste, or use ssh to ease the process, the command **on the worker node**. However, please substitute the `<manager_ip>` with its hostname.
-
-After that, we must add labels to nodes devoted to run ROFT and ML modules.
-
-For each worker providing ROFT, please do on the `manager`:
-
-```console
-docker node update --label-add roft_deployer=true <node_hostname>
-```
-
-For those providing ML modules, please do on the `manager`:
-
-```console
-docker node update --label-add ycb_cv_deployer=true <node_hostname>
-```
-
-### Configure GPUs on all machines (manager and workers)
-
-For each machine, please do the following.
-
-Find the GPU-ID first:
-
-```console
-nvidia-smi -a | grep UUID | awk '{print substr($4,0,12)}'
-```
-
-You will get something like `GPU-xxxxxxx`.
-
-Then edit `/etc/docker/daemon.json` such that it looks like:
-```json
-{
-  "runtimes": {
-    "nvidia": {
-      "path": "/usr/bin/nvidia-container-runtime",
-      "runtimeArgs": []
-    }
-  },
-  "default-runtime": "nvidia",
-  "node-generic-resources": [
-    "NVIDIA-GPU=<gpu_id_0>",
-    "NVIDIA-GPU=<gpu_id_1>"
-    ]
-}
-```
-
-Please add as many GPUs as desired.
-
-Then enable GPU advertising by uncommenting the line `swarm-resource = "DOCKER_RESOURCE_GPU"` in  `/etc/nvidia-container-runtime/config.toml`.
-
-Finally, restart docker by issuing `sudo systemctl restart docker.service`.
-
-## Deploy the stack (for each run)
-
-```console
-cd roft-samples/dockercompose
-docker stack deploy -c docker-compose.yml roft-samples-handover-stack
-```
-
-After running the demo, stop all services and rm the stack:
-
-```console
-docker service update --replicas 0 roft-samples-handover-stack_roft
-docker service update --replicas 0 roft-samples-handover-stack_detectron2
-docker service update --replicas 0 roft-samples-handover-stack_dope
-docker stack rm roft-samples-handover-stack
-```