postgres shadow

Dockers/postgres-shadow/Dockerfile

@@ -1,8 +1,9 @@
-FROM postgres:15
+FROM pgvector/pgvector:pg15
 
 # Install additional tools
 RUN apt-get update && apt-get install -y \
     postgresql-client \
+    cron \
     && rm -rf /var/lib/apt/lists/*
 
 # Create directory for scripts
@@ -12,9 +13,31 @@ RUN mkdir -p /docker-entrypoint-initdb.d
 COPY init.sh /docker-entrypoint-initdb.d/
 RUN chmod +x /docker-entrypoint-initdb.d/init.sh
 
+# Copy refresh script
+COPY refresh.sh /usr/local/bin/
+RUN chmod +x /usr/local/bin/refresh.sh
+
 # Environment variables
 ENV SOURCE_DB_URL=""
 ENV TARGET_DB_NAME="shadow_db"
+ENV TARGET_DB_USER="postgres"
+ENV TARGET_DB_PASSWORD=""
+ENV REFRESH_TIME="00:00"
+
+# Define volume for persistent data
+VOLUME ["/var/lib/postgresql/data"]
 
 # Expose PostgreSQL port
 EXPOSE 5432
+
+# Add crontab file
+RUN echo "0 0 * * * /usr/local/bin/refresh.sh >> /var/log/cron.log 2>&1" > /etc/cron.d/refresh-cron
+RUN chmod 0644 /etc/cron.d/refresh-cron
+
+# Create the log file to be able to run tail
+RUN touch /var/log/cron.log
+
+# Start cron and PostgreSQL
+COPY docker-entrypoint.sh /
+RUN chmod +x /docker-entrypoint.sh
+ENTRYPOINT ["/docker-entrypoint.sh"]

Dockers/postgres-shadow/docker-entrypoint.sh

@@ -0,0 +1,14 @@
+#!/bin/bash
+set -e
+
+# Start cron
+service cron start
+
+# Update crontab with the specified refresh time
+hour=$(echo $REFRESH_TIME | cut -d':' -f1)
+minute=$(echo $REFRESH_TIME | cut -d':' -f2)
+echo "$minute $hour * * * /usr/local/bin/refresh.sh >> /var/log/cron.log 2>&1" > /etc/cron.d/refresh-cron
+chmod 0644 /etc/cron.d/refresh-cron
+
+# Start PostgreSQL
+exec docker-entrypoint.sh postgres

Dockers/postgres-shadow/init.sh

@@ -1,25 +0,0 @@
-#!/bin/bash
-set -e
-
-# Wait for PostgreSQL to be ready
-until pg_isready; do
-    echo "Waiting for PostgreSQL to be ready..."
-    sleep 2
-done
-
-# Create the target database
-psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$POSTGRES_DB" <<-EOSQL
-    CREATE DATABASE $TARGET_DB_NAME;
-EOSQL
-
-# Copy the database if SOURCE_DB_URL is provided
-if [ ! -z "$SOURCE_DB_URL" ]; then
-    echo "Copying database from $SOURCE_DB_URL to $TARGET_DB_NAME..."
-
-    # Dump the source database
-    pg_dump "$SOURCE_DB_URL" | psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$TARGET_DB_NAME"
-
-    echo "Database copy completed successfully"
-else
-    echo "No SOURCE_DB_URL provided, skipping database copy"
-fi

Dockers/postgres-shadow/refresh.sh

@@ -0,0 +1,64 @@
+#!/bin/bash
+set -e
+
+echo "Starting database refresh at $(date)"
+
+# Wait for PostgreSQL to be ready
+until pg_isready; do
+    echo "Waiting for PostgreSQL to be ready..."
+    sleep 2
+done
+
+# Create user if it doesn't exist
+psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$POSTGRES_DB" <<-EOSQL
+    DO
+    \$do\$
+    BEGIN
+        IF NOT EXISTS (SELECT FROM pg_catalog.pg_roles WHERE rolname = '$POSTGRES_USER') THEN
+            CREATE USER $POSTGRES_USER WITH PASSWORD '$POSTGRES_PASSWORD' SUPERUSER;
+        END IF;
+    END
+    \$do\$;
+EOSQL
+
+# Drop and recreate the target database
+psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$POSTGRES_DB" <<-EOSQL
+    DROP DATABASE IF EXISTS $TARGET_DB_NAME;
+    CREATE DATABASE $TARGET_DB_NAME OWNER $POSTGRES_USER;
+EOSQL
+
+# Copy the database if SOURCE_DB_URL is provided
+if [ ! -z "$SOURCE_DB_URL" ]; then
+    echo "Copying database from $SOURCE_DB_URL to $TARGET_DB_NAME..."
+
+    # Dump the source database and filter out transaction_timeout
+    pg_dump "$SOURCE_DB_URL" | grep -v "transaction_timeout" | psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$TARGET_DB_NAME"
+
+    # Update ownership of all objects
+    psql -v ON_ERROR_STOP=1 --username "$POSTGRES_USER" --dbname "$TARGET_DB_NAME" <<-EOSQL
+        DO \$\$
+        DECLARE
+            r record;
+        BEGIN
+            FOR r IN SELECT tablename FROM pg_tables WHERE schemaname = 'public'
+            LOOP
+                EXECUTE 'ALTER TABLE public.' || quote_ident(r.tablename) || ' OWNER TO $POSTGRES_USER';
+            END LOOP;
+
+            FOR r IN SELECT sequence_name FROM information_schema.sequences WHERE sequence_schema = 'public'
+            LOOP
+                EXECUTE 'ALTER SEQUENCE public.' || quote_ident(r.sequence_name) || ' OWNER TO $POSTGRES_USER';
+            END LOOP;
+
+            FOR r IN SELECT typname FROM pg_type WHERE typnamespace = 'public'::regnamespace
+            LOOP
+                EXECUTE 'ALTER TYPE public.' || quote_ident(r.typname) || ' OWNER TO $POSTGRES_USER';
+            END LOOP;
+        END
+        \$\$;
+EOSQL
+
+    echo "Database refresh completed successfully at $(date)"
+else
+    echo "No SOURCE_DB_URL provided, skipping database refresh"
+fi

Dockers/postgres-shadow/version

@@ -0,0 +1 @@
+latest