Update Dockerfile

Dockerfile

@@ -1,35 +1,40 @@
-# Use correct Java 17 base image
-FROM openjdk:17-jdk-slim
+FROM ubuntu:22.04
 
-# Install dependencies
+# Install system dependencies
 RUN apt-get update && apt-get install -y \
     curl \
+    openjdk-11-jdk \
     python3 \
     python3-pip \
     wget \
+    apt-transport-https \
+    gnupg \
     && rm -rf /var/lib/apt/lists/*
 
-# Install Elasticsearch 8.8.0
+# Install Elasticsearch
 ENV ES_VERSION=8.8.0
-RUN curl -O https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-${ES_VERSION}-linux-x86_64.tar.gz && \
-    tar -xzf elasticsearch-${ES_VERSION}-linux-x86_64.tar.gz && \
-    mv elasticsearch-${ES_VERSION} /usr/share/elasticsearch && \
-    rm elasticsearch-${ES_VERSION}-linux-x86_64.tar.gz
-
-# Configure Elasticsearch properly
-RUN echo "discovery.type: single-node\n\
-xpack.security.enabled: false\n\
-network.host: 0.0.0.0" > /usr/share/elasticsearch/config/elasticsearch.yml
-
-# Create user and directories
-RUN useradd -m -u 1000 appuser && \
-    mkdir -p /app && \
+RUN curl -O https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-8.8.0-linux-x86_64.tar.gz && \
+    tar -xzf elasticsearch-8.8.0-linux-x86_64.tar.gz && \
+    mv elasticsearch-8.8.0 /usr/share/elasticsearch && \
+    rm elasticsearch-8.8.0-linux-x86_64.tar.gz
+
+# Create elasticsearch.yml with proper YAML format
+RUN echo "discovery.type: single-node" > /usr/share/elasticsearch/config/elasticsearch.yml && \
+    echo "xpack.security.enabled: false" >> /usr/share/elasticsearch/config/elasticsearch.yml && \
+    echo "network.host: 0.0.0.0" >> /usr/share/elasticsearch/config/elasticsearch.yml
+
+# Set Elasticsearch environment variables
+ENV ES_JAVA_OPTS="-Xms1g -Xmx1g"
+
+# Create non-root user for running the services
+RUN useradd -m -u 1000 appuser
+RUN mkdir -p /app /usr/share/elasticsearch/data && \
     chown -R appuser:appuser /app /usr/share/elasticsearch
 
-# Set working directory
+# Create app directory
 WORKDIR /app
 
-# Copy application files
+# Copy your project files
 COPY --chown=appuser:appuser app.py streamlit.py requirements.txt ./
 COPY --chown=appuser:appuser chunking ./chunking
 COPY --chown=appuser:appuser embeddings ./embeddings
@@ -38,26 +43,43 @@ COPY --chown=appuser:appuser elastic ./elastic
 COPY --chown=appuser:appuser file_processing.py ./
 COPY --chown=appuser:appuser ingestion.py ./
 
+# Copy ES data if needed - consider if this is actually necessary
+COPY --chown=appuser:appuser es_data /usr/share/elasticsearch/data
+
 # Install Python dependencies
-RUN pip3 install -r requirements.txt && \
-    pip3 install logback==0.2.1  # For SLF4J compatibility
-
-# Configure environment variables
-ENV ES_JAVA_OPTS="-Xms1g -Xmx1g" \
-    STREAMLIT_SERVER_PORT=7860 \
-    STREAMLIT_SERVER_HEADLESS=true \
-    ES_HOST=localhost \
-    ES_PORT=9200 \
-    ELASTICSEARCH_HOSTS=http://localhost:9200
-
-# Expose required ports
+RUN pip3 install -r requirements.txt
+
+# Set environment variables for Streamlit
+ENV STREAMLIT_SERVER_HEADLESS=true
+ENV STREAMLIT_SERVER_PORT=7860
+ENV STREAMLIT_SERVER_ENABLE_CORS=false
+ENV ES_HOST=localhost
+ENV ES_PORT=9200
+ENV ELASTICSEARCH_HOSTS="http://localhost:9200"
+
+# Expose required ports (Elasticsearch and Streamlit)
 EXPOSE 9200 7860
 
 # Switch to non-root user
 USER appuser
 
-# Create proper startup script
-COPY --chown=appuser:appuser start.sh /app/start.sh
-RUN chmod +x /app/start.sh
+# Create startup script
+RUN echo '#!/bin/bash\n\
+# Start Elasticsearch in the background\n\
+/usr/share/elasticsearch/bin/elasticsearch &\n\
+\n\
+# Wait for Elasticsearch to become available\n\
+echo "Waiting for Elasticsearch to start..."\n\
+until curl -s http://localhost:9200 > /dev/null; do\n\
+  sleep 2\n\
+  echo "Still waiting for Elasticsearch..."\n\
+done\n\
+echo "Elasticsearch is up and running!"\n\
+\n\
+# Start Streamlit\n\
+echo "Starting Streamlit application..."\n\
+streamlit run /app/streamlit.py\n\
+' > /app/start.sh && chmod +x /app/start.sh
 
+# Command to run
 CMD ["/app/start.sh"]