feat(terraform): Optimize ETL execution

Signed-off-by: pmikus <peter.mikus@protonmail.ch>
Change-Id: I11dfe8f642c6d565d39b1e0fbc41361460a42320

1 files changed, 53 insertions, 0 deletions

diff --git a/fdio.infra.terraform/terraform-nomad-pyspark-etl/conf/nomad/etl-stats.hcl.tftpl b/fdio.infra.terraform/terraform-nomad-pyspark-etl/conf/nomad/etl-stats.hcl.tftpl
new file mode 100644
index 0000000000..86ca584de7
--- /dev/null
+++ b/fdio.infra.terraform/terraform-nomad-pyspark-etl/conf/nomad/etl-stats.hcl.tftpl
@@ -0,0 +1,53 @@
+job "${job_name}" {
+  datacenters = "${datacenters}"
+  type        = "${type}"
+  periodic {
+    cron             = "${cron}"
+    prohibit_overlap = "${prohibit_overlap}"
+    time_zone        = "${time_zone}"
+  }
+  group "${job_name}" {
+    restart {
+      mode = "fail"
+    }
+    constraint {
+      attribute       = "$${attr.cpu.arch}"
+      operator        = "!="
+      value           = "arm64"
+    }
+    constraint {
+      attribute      = "$${node.class}"
+      value          = "builder"
+    }
+    task "${job_name}" {
+      artifact {
+        source      = "git::https://github.com/FDio/csit"
+        destination = "local/csit"
+      }
+      driver = "docker"
+      config {
+        image   = "${image}"
+        command = "gluesparksubmit"
+        args = [
+          "--driver-memory", "10g",
+          "--executor-memory", "10g",
+          "stats.py"
+        ]
+        work_dir = "/local/csit/csit.infra.etl"
+      }
+      env {
+        AWS_ACCESS_KEY_ID         = "${aws_access_key_id}"
+        AWS_SECRET_ACCESS_KEY     = "${aws_secret_access_key}"
+        AWS_DEFAULT_REGION        = "${aws_default_region}"
+        OUT_AWS_ACCESS_KEY_ID     = "${out_aws_access_key_id}"
+        OUT_AWS_SECRET_ACCESS_KEY = "${out_aws_secret_access_key}"
+        OUT_AWS_DEFAULT_REGION    = "${out_aws_default_region}"
+        ${ envs }
+      }
+      resources {
+        cpu    = ${cpu}
+        memory = ${memory}
+      }
+    }
+  }
+}