From 9c9456216aefea79f5f106cb39beb71bbfdc707b Mon Sep 17 00:00:00 2001
From: Naved Ansari <naved001@gmail.com>
Date: Wed, 30 Oct 2024 12:33:21 -0400
Subject: [PATCH] Some updates to the kubernetes manifests so they work with
 the new code changes

I have added a couple of shell scripts to produce report and collect metrics.
The cronjobs now call them. The dockerfile has been updated to include the
shell files.

A configmap has been added that mounts the gpu node map file at the approprate
directory.
---
 Dockerfile                                    |  4 +++-
 bin/collect_metrics.sh                        |  5 +++++
 bin/produce_report.sh                         |  6 +++++
 ...y-openshift-metrics-collector-cronjob.yaml |  8 +------
 k8s/base/gpu-node-map-configmap.yaml          | 22 +++++++++++++++++++
 k8s/base/kustomization.yaml                   |  1 +
 k8s/base/produce-report-cronjob.yaml          |  8 ++++++-
 7 files changed, 45 insertions(+), 9 deletions(-)
 create mode 100755 bin/collect_metrics.sh
 create mode 100755 bin/produce_report.sh
 create mode 100644 k8s/base/gpu-node-map-configmap.yaml

diff --git a/Dockerfile b/Dockerfile
index c2ff164..2be712c 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -5,5 +5,7 @@ COPY requirements.txt ./
 RUN pip install -r requirements.txt
 
 COPY openshift_metrics/ /app/openshift_metrics
+COPY bin/collect_metrics.sh /app/collect_metrics.sh
+COPY bin/produce_report.sh /app/produce_report.sh
 
-CMD ["python", "openshift_metrics/openshift_prometheus_metrics.py", "--upload-to-s3"]
+CMD ["./collect_metrics.sh"]
diff --git a/bin/collect_metrics.sh b/bin/collect_metrics.sh
new file mode 100755
index 0000000..cf6cb34
--- /dev/null
+++ b/bin/collect_metrics.sh
@@ -0,0 +1,5 @@
+#!/usr/bin/env sh
+
+python -m openshift_metrics.openshift_prometheus_metrics \
+    --output-file /tmp/metrics.json \
+    --upload-to-s3
diff --git a/bin/produce_report.sh b/bin/produce_report.sh
new file mode 100755
index 0000000..dfc5d91
--- /dev/null
+++ b/bin/produce_report.sh
@@ -0,0 +1,6 @@
+#!/usr/bin/env sh
+
+python -m openshift_metrics.merge /data/*.json \
+    --invoice-file /tmp/invoice.csv \
+    --pod-report-file /tmp/pod-report.csv \
+    --upload-to-s3
diff --git a/k8s/base/daily-openshift-metrics-collector-cronjob.yaml b/k8s/base/daily-openshift-metrics-collector-cronjob.yaml
index 2951e5d..e7da03c 100644
--- a/k8s/base/daily-openshift-metrics-collector-cronjob.yaml
+++ b/k8s/base/daily-openshift-metrics-collector-cronjob.yaml
@@ -29,11 +29,5 @@ spec:
                 secretKeyRef:
                   name: openshift-metrics-b2-bucket
                   key: secret-access-key
-            volumeMounts:
-            - name: data-volume
-              mountPath: /data
-            command: ["/bin/sh", "-c", "cd /data && python /app/openshift_metrics/openshift_prometheus_metrics.py --upload-to-s3"]
-          volumes:
-          - name: data-volume
-            emptyDir: {}
+            command: ["./collect_metrics.sh"]
           restartPolicy: OnFailure
diff --git a/k8s/base/gpu-node-map-configmap.yaml b/k8s/base/gpu-node-map-configmap.yaml
new file mode 100644
index 0000000..b841af3
--- /dev/null
+++ b/k8s/base/gpu-node-map-configmap.yaml
@@ -0,0 +1,22 @@
+apiVersion: v1
+kind: ConfigMap
+metadata:
+  name: gpu-node-map
+data:
+  gpu_node_map.json: |
+    {
+        "wrk-88": "Tesla-V100-PCIE-32GB",
+        "wrk-89": "Tesla-V100-PCIE-32GB",
+        "wrk-94": "NVIDIA-A100-SXM4-40GB",
+        "wrk-95": "NVIDIA-A100-SXM4-40GB",
+        "wrk-97": "NVIDIA-A100-SXM4-40GB",
+        "wrk-98": "NVIDIA-A100-SXM4-40GB",
+        "wrk-99": "NVIDIA-A100-SXM4-40GB",
+        "wrk-102": "Tesla-V100-PCIE-32GB",
+        "wrk-103": "Tesla-V100-PCIE-32GB",
+        "wrk-104": "Tesla-V100-PCIE-32GB",
+        "wrk-105": "Tesla-V100-PCIE-32GB",
+        "wrk-106": "Tesla-V100-PCIE-32GB",
+        "wrk-107": "Tesla-V100-PCIE-32GB",
+        "wrk-108": "Tesla-V100-PCIE-32GB"
+    }
diff --git a/k8s/base/kustomization.yaml b/k8s/base/kustomization.yaml
index 5769c19..1fbae9b 100644
--- a/k8s/base/kustomization.yaml
+++ b/k8s/base/kustomization.yaml
@@ -2,3 +2,4 @@ resources:
   - daily-openshift-metrics-collector-cronjob.yaml
   - produce-report-cronjob.yaml
   - metrics-downloader-configmap.yaml
+  - gpu-node-map-configmap.yaml
diff --git a/k8s/base/produce-report-cronjob.yaml b/k8s/base/produce-report-cronjob.yaml
index b5c3319..d67e1f5 100644
--- a/k8s/base/produce-report-cronjob.yaml
+++ b/k8s/base/produce-report-cronjob.yaml
@@ -35,7 +35,10 @@ spec:
             volumeMounts:
             - name: data-volume
               mountPath: /data
-            command: ["/bin/sh", "-c", "cd /data && python /app/openshift_metrics/merge.py /data/*.json --upload-to-s3"]
+            - name: gpu-node-map
+              mountPath: /app/gpu_node_map.json
+              subPath: gpu_node_map.json
+            command: ["./produce_report.sh"]
           initContainers:
             - name: download-metrics
               image: amazon/aws-cli
@@ -63,4 +66,7 @@ spec:
             configMap:
               name: metrics-downloader
               defaultMode: 0555
+          - name: gpu-node-map
+            configMap:
+              name: gpu-node-map
           restartPolicy: OnFailure