From 50ab50d282761bac36003d27ff3c638ceeeb663a Mon Sep 17 00:00:00 2001
From: Abraham Toriz <categulario@gmail.com>
Date: Fri, 29 Apr 2022 10:39:20 +0800
Subject: [PATCH 1/2] proposal of a Containerfile for datapusher-plus

---
 Containerfile                     | 70 +++++++++++++++++++++++++++++++
 container/initialize-and-start.sh | 23 ++++++++++
 container/uwsgi.ini               | 16 +++++++
 datapusher/settings.py            |  2 +-
 4 files changed, 110 insertions(+), 1 deletion(-)
 create mode 100644 Containerfile
 create mode 100644 container/initialize-and-start.sh
 create mode 100644 container/uwsgi.ini

diff --git a/Containerfile b/Containerfile
new file mode 100644
index 00000000..b2eb0346
--- /dev/null
+++ b/Containerfile
@@ -0,0 +1,70 @@
+FROM ubuntu:20.04
+
+# Set timezone
+ENV TZ=UTC
+RUN ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
+
+# Setting the locale
+ENV LC_ALL=en_US.UTF-8
+RUN apt-get update && apt-get install --no-install-recommends -y locales
+RUN sed -i "/$LC_ALL/s/^# //g" /etc/locale.gen
+RUN dpkg-reconfigure --frontend=noninteractive locales
+RUN update-locale LANG=${LC_ALL}
+
+# Install required system packages
+RUN apt-get -q -y update \
+    && DEBIAN_FRONTEND=noninteractive apt-get -q -y upgrade \
+    && apt-get -q -y install \
+        python3 \
+        pip \
+        virtualenv \
+        postgresql-client \
+        unzip \
+        wget \
+    && apt-get -q clean \
+    && rm -rf /var/lib/apt/lists/*
+
+# Define environment variables
+ENV DATAPUSHER_HOME=/usr/lib/ckan/datapusher
+ENV DATAPUSHER_CODE=$DATAPUSHER_HOME/code
+ENV DATAPUSHER_VENV=$DATAPUSHER_HOME/venv
+ENV DATAPUSHER_CONFIG=/etc/ckan/datapusher
+
+# Create ckan user
+RUN useradd -r -u 900 -m -c "ckan account" -d $DATAPUSHER_HOME -s /bin/false ckan
+
+# Install qsv
+ENV QSV_RELEASE=0.43.0
+ENV QSV_ARCHIVE=qsv-$QSV_RELEASE-x86_64-unknown-linux-gnu.zip
+RUN cd /tmp && \
+    wget https://github.com/jqnatividad/qsv/releases/download/$QSV_RELEASE/$QSV_ARCHIVE && \
+    unzip $QSV_ARCHIVE && mv qsvlite /usr/local/bin/ && rm $QSV_ARCHIVE
+
+# Setup virtual environment for CKAN
+RUN mkdir -p $DATAPUSHER_CONFIG && \
+    virtualenv $DATAPUSHER_VENV && \
+    ln -s $DATAPUSHER_VENV/bin/pip3 /usr/local/bin/ckan-pip3 && \
+    ln -s $DATAPUSHER_VENV/bin/ckan /usr/local/bin/ckan
+
+# Virtual environment binaries/scripts to be used first
+ENV PATH=${DATAPUSHER_VENV}/bin:${PATH}
+
+# install the dependencies
+RUN ckan-pip3 install -U pip && \
+    CPUCOUNT=1 ckan-pip3 install --upgrade --no-cache-dir uwsgi psycopg2-binary
+
+# Copy the ckan code to the image
+COPY . $DATAPUSHER_CODE
+RUN cp $DATAPUSHER_CODE/container/initialize-and-start.sh / && \
+    chmod +x /initialize-and-start.sh
+
+# install datapusher-plus
+RUN ckan-pip3 install -e $DATAPUSHER_CODE
+
+# Set ownership of directories
+RUN chown -R ckan:ckan $DATAPUSHER_HOME $DATAPUSHER_CONFIG
+
+USER ckan
+EXPOSE 8800
+
+CMD ["/initialize-and-start.sh"]
diff --git a/container/initialize-and-start.sh b/container/initialize-and-start.sh
new file mode 100644
index 00000000..c82c1482
--- /dev/null
+++ b/container/initialize-and-start.sh
@@ -0,0 +1,23 @@
+#!/bin/sh
+set -e
+
+UWSGI_FILE=${DATAPUSHER_CONFIG}/uwsgi.ini
+
+abort () {
+  echo "$@" >&2
+  exit 1
+}
+
+# Fail if postgresql is not running
+if ! pg_isready -h "${POSTGRES_HOST}" -U "${POSTGRES_USER}"; then
+    abort "Postgresql not running"
+fi
+
+if [ ! -e $UWSGI_FILE ]; then
+    cp $DATAPUSHER_CODE/container/uwsgi.ini $UWSGI_FILE
+fi
+
+datapusher_initdb $DATAPUSHER_CODE/datapusher/settings.py
+
+# run ckan with uwsgi
+exec $DATAPUSHER_VENV/bin/uwsgi -i $UWSGI_FILE
diff --git a/container/uwsgi.ini b/container/uwsgi.ini
new file mode 100644
index 00000000..8aa04d74
--- /dev/null
+++ b/container/uwsgi.ini
@@ -0,0 +1,16 @@
+[uwsgi]
+
+http = 0.0.0.0:8800
+uid = ckan
+guid = ckan
+virtualenv = /usr/lib/ckan/datapusher/venv
+module = datapusher.wsgi:application
+master = true
+harakiri = 50
+max-requests = 5000
+vacuum = true
+callable = application
+buffer-size = 32768
+workers = 4
+threads = 2
+lazy-apps = true
diff --git a/datapusher/settings.py b/datapusher/settings.py
index 1e5a6898..4f692f30 100644
--- a/datapusher/settings.py
+++ b/datapusher/settings.py
@@ -19,7 +19,7 @@
 
 # PostgreSQL COPY settings
 # set this to the same value as your ckan.datastore.write_url
-WRITE_ENGINE_URL = os.environ.get('WRITE_ENGINE_URL', 'postgresql://datapusher:THEPASSWORD@localhost/datastore_default')
+WRITE_ENGINE_URL = os.environ.get('DATAPUSHER_WRITE_ENGINE_URL', 'postgresql://datapusher:THEPASSWORD@localhost/datastore_default')
 
 # qsv settings
 QSV_BIN = os.environ.get('DATAPUSHER_QSV_BIN', '/usr/local/bin/qsvlite')

From 609678df6af7673988c94e22a884bbbc640fd9c1 Mon Sep 17 00:00:00 2001
From: Abraham Toriz <categulario@gmail.com>
Date: Fri, 29 Apr 2022 10:43:18 +0800
Subject: [PATCH 2/2] mention in readme that all variables are tunable via
 environment

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 9a58be61..1839a5ee 100644
--- a/README.md
+++ b/README.md
@@ -248,7 +248,7 @@ CREATE ROLE datapusher LOGIN SUPERUSER PASSWORD 'thepassword';
 quit
 ```
 
-Most of the configuration options above can be also provided as environment
+All of the configuration options above can be also provided as environment
 variables prepending the name with `DATAPUSHER_`, eg
 `DATAPUSHER_SQLALCHEMY_DATABASE_URI`, `DATAPUSHER_PORT`, etc. For variables with
 boolean values you must use `1` or `0`.