feat: Update docker-compose configuration to use environment variables for sensitive data, enhance testing documentation with new LaTeX files for capacity planning, and refine README for clarity on compilation and testing processes.

2026-07-17 08:32:00 -05:00
parent e47ef4f8c6
commit dc3fcc2280
89 changed files with 5331 additions and 1210 deletions
@@ -2,35 +2,47 @@
 \centering
 \begin{tikzpicture}[
  font=\small,
-  box/.style={draw=ipsblue, rounded corners=2pt, align=center, inner sep=4pt, minimum height=0.85cm, minimum width=2.4cm},
+  box/.style={draw=ipsblue, rounded corners=2pt, align=center, inner sep=4pt, minimum height=0.85cm, minimum width=2.3cm},
  ext/.style={draw=ipsgrey, dashed, rounded corners=2pt, align=center, inner sep=4pt, minimum height=0.85cm},
-  arr/.style={-{Latex}, thick, ipsblue}
+  arr/.style={-{Latex}, thick, ipsblue},
+  dbarr/.style={-{Latex}, thick, ipsgrey}
 ]
  \node[box, fill=codebg] (desk) {Desktop\\Electron};
-  \node[box, fill=codebg, right=1.8cm of desk] (gw) {api\_gateway\\:8189};
-  \node[box, below=1.1cm of gw] (ext) {extraction\\:8001};
-  \node[box, left=0.8cm of ext] (clean) {cleaning\\:8002};
-  \node[box, right=0.8cm of ext] (count) {count-api\\:8080};
-  \node[box, below=1.1cm of ext] (bot) {bot-descarga\\:8200};
-  \node[box, below left=1.0cm and 0.2cm of bot] (pg) {PostgreSQL};
-  \node[box, below right=1.0cm and 0.2cm of bot] (redis) {Redis};
-  \node[ext, left=1.6cm of desk] (agent) {Agent PC\\Selenium};
-  \node[ext, below=2.6cm of agent] (minio) {MinIO/S3};
-  \node[ext, below=0.9cm of minio] (img) {Imaging};
+  \node[box, fill=codebg, right=1.7cm of desk] (gw) {api\_gateway\\:8189};
+  \node[box, below=1.05cm of gw] (ext) {extraction\\(+ workers)};
+  \node[box, left=0.7cm of ext] (clean) {cleaning\\:8002};
+  \node[box, right=0.7cm of ext] (count) {count-api\\:8080};
+  \node[box, below=1.05cm of ext] (bot) {bot-descarga\\orquestador};
+  \node[box, below left=1.15cm and 1.4cm of bot] (pg) {PostgreSQL};
+  \node[box, below right=1.15cm and 1.4cm of bot] (redis) {Redis};
+  \node[ext, left=1.5cm of desk] (agent) {Agent Selenium\\(PC externo)};
+  \node[ext, below=2.5cm of agent] (minio) {MinIO/S3};
+  \node[ext, below=0.85cm of minio] (img) {Imaging};

+  % HTTP via gateway
  \draw[arr] (desk) -- (gw);
  \draw[arr] (gw) -- (ext);
  \draw[arr] (gw) -- (clean);
  \draw[arr] (gw) -- (count);
  \draw[arr] (gw) -- (bot);
-  \draw[arr] (bot) -- (pg);
-  \draw[arr] (ext) -- (pg);
-  \draw[arr] (ext) -- (redis);
  \draw[arr] (agent) |- (gw);
+
+  % PostgreSQL: solo servicios de dominio (NO gateway, NO count-api)
+  \draw[dbarr] (ext) -- (pg);
+  \draw[dbarr] (clean) |- (pg);
+  \draw[dbarr] (bot) -- (pg);
+
+  % Redis: cola ARQ (extraction/workers)
+  \draw[dbarr] (ext) -- (redis);
+
+  % MinIO / Imaging
  \draw[arr] (agent) -- (minio);
  \draw[arr] (bot) -| (minio);
  \draw[arr] (agent) -- (img);
 \end{tikzpicture}
-\caption{Arquitectura IPSExtractor bajo prueba.}
+\caption{Arquitectura bajo prueba. El \textbf{gateway no habla con Postgres}: solo hace proxy HTTP.
+Usan la BD compartida: \texttt{extraction} (+ workers), \texttt{cleaning} y \texttt{bot-descarga}.
+\texttt{count-api} hace OCR y \textbf{devuelve JSON} (no escribe en PG); el Desktop/extraction
+persiste esos resultados con \texttt{POST /count-jobs} en Postgres.}
 \label{fig:arquitectura}
 \end{figure}
@@ -0,0 +1,40 @@
+\begin{figure}[H]
+\centering
+\resizebox{\textwidth}{!}{%
+\begin{tikzpicture}[
+  font=\footnotesize,
+  node distance=0.45cm and 0.55cm,
+  box/.style={draw=ipsblue, rounded corners=3pt, align=center, inner sep=4pt,
+    minimum height=0.95cm, minimum width=2.55cm, fill=codebg, text=black},
+  title/.style={font=\small\bfseries, text=ipsblue, anchor=west},
+  arr/.style={-{Latex}, thick, draw=ipsblue}
+]
+  % --- 1: API Gateway ---
+  \node[title] (t1) at (0,0) {API Gateway};
+  \node[box, below=0.3cm of t1.west, anchor=west] (client) {Desktop / Agent};
+  \node[box, right=of client] (gw) {Gateway :8189};
+  \node[box, right=of gw] (svc) {Microservicios};
+  \draw[arr] (client) -- (gw);
+  \draw[arr] (gw) -- (svc);
+
+  % --- 2: ARQ ---
+  \node[title] (t2) at (0,-2.15) {Cola async (ARQ)};
+  \node[box, below=0.3cm of t2.west, anchor=west] (api) {extraction API};
+  \node[box, right=of api] (redis) {Redis / ARQ};
+  \node[box, right=of redis] (w) {Workers $\times N$};
+  \draw[arr] (api) -- (redis);
+  \draw[arr] (redis) -- (w);
+
+  % --- 3: Claim ---
+  \node[title] (t3) at (0,-4.3) {Claim / lease (orquestaci\'on)};
+  \node[box, below=0.3cm of t3.west, anchor=west] (store) {bot-descarga\\(orquestador)};
+  \node[box, right=of store] (claim) {claim at\'omico};
+  \node[box, right=of claim] (agent) {Agent Selenium\\(PC)};
+  \draw[arr] (store) -- (claim);
+  \draw[arr] (claim) -- (agent);
+\end{tikzpicture}%
+}
+\caption{Patrones: Gateway, cola ARQ y claim. El servicio \texttt{bot-descarga} \textbf{orquesta};
+el \textbf{agent Selenium} en el PC ejecuta la descarga.}
+\label{fig:patrones}
+\end{figure}
@@ -0,0 +1,30 @@
+\begin{figure}[ht]
+\centering
+\begin{tikzpicture}[
+  font=\small,
+  stage/.style={draw=ipsblue, rounded corners=3pt, align=center, inner sep=5pt,
+    minimum height=1.0cm, minimum width=2.15cm, fill=codebg},
+  arr/.style={-{Latex}, thick, ipsblue}
+]
+  \node[stage] (dl) {1. Descarga\\agent Selenium};
+  \node[stage, right=0.5cm of dl] (cnt) {2. Conteo\\OCR/ONNX};
+  \node[stage, right=0.5cm of cnt] (ext) {3. Extracci\'on\\LLM + ARQ};
+  \node[stage, right=0.5cm of ext] (cln) {4. Limpieza\\CSV/reglas};
+
+  \draw[arr] (dl) -- (cnt);
+  \draw[arr] (cnt) -- (ext);
+  \draw[arr] (ext) -- (cln);
+
+  \node[below=0.35cm of dl, font=\footnotesize\itshape, text=ipsgrey, align=center]
+    {orquestado por\\bot-descarga};
+  \node[below=0.35cm of cnt, font=\footnotesize\itshape, text=ipsgrey, align=center]
+    {filtro};
+  \node[below=0.35cm of ext, font=\footnotesize\itshape, text=ipsgrey, align=center]
+    {filas PG};
+  \node[below=0.35cm of cln, font=\footnotesize\itshape, text=ipsgrey, align=center]
+    {lote limpio};
+\end{tikzpicture}
+\caption{Pipeline punta a punta. La descarga la hace el \textbf{agent Selenium} (PC);
+el servicio \texttt{bot-descarga} solo orquesta jobs/claim/heartbeats.}
+\label{fig:pipeline-negocio}
+\end{figure}