AWS Setup md

Signed-off-by: Effi-S <effi.szt@gmail.com>

Author: Effi-S

cmd/benchmarking/plotly_plot_node.py

@@ -1,5 +1,5 @@
 import re
-import sys
+import os
 from pathlib import Path
 from typing import Any
 import pandas as pd
@@ -10,7 +10,7 @@
 
 IGNORE_COLS = {"bench", "workers", "tps",
                "iterations", "ns/op", "B/op", "allocs/op"}
-DEFAULT_BENCH_DIR = "bench2"
+DEFAULT_BENCH_DIR = "bench"
 
 
 def _is_number(s: str) -> bool:
@@ -75,6 +75,10 @@ def simple_parser(path: Path) -> pd.DataFrame:
         df[f"{label} (ms)"] = df[col] / 1e6
         df = df.drop(columns=[col])
 
+    if 'tps' in df.columns and 'workers' in df.columns:
+        # Little's Law (latency = concurrency / throughput)
+        df['avg (ms)'] = df['workers'] * 1000 / df['tps']
+
     return df
 
 
@@ -102,13 +106,21 @@ def _parse_ms(s: str) -> float:
     raise ValueError(f"unknown unit: {unit}")
 
 
-LATENCY_KEYS = {'P5': 'p5 (ms)', 'P95': 'p95 (ms)', 'P99': 'p99 (ms)'}
+LATENCY_KEYS = {
+    'P50 (Median)': 'p50 (ms)',
+    'P5': 'p5 (ms)',
+    'P95': 'p95 (ms)',
+    'P99': 'p99 (ms)',
+}
+DISPLAY_LATENCY = {'avg (ms)', 'p50 (ms)', 'p95 (ms)'}
+WORKERS_RE = re.compile(r'^Workers\s+(\d+)')
 
 
-def parse_parallel_log(path: Path) -> pd.DataFrame:
+def parse_parallel_log(path: Path, default_bench: str | None = None) -> pd.DataFrame:
     text = ANSI_ESCAPE_RE.sub('', path.read_text())
     rows: list[dict] = []
     current: dict = {}
+    has_run_lines = bool(PARALLEL_RUN_RE.search(text))
 
     for line in text.splitlines():
         line = line.strip()
@@ -129,6 +141,18 @@ def parse_parallel_log(path: Path) -> pd.DataFrame:
                         current[k] = v
             continue
 
+        if not has_run_lines:
+            m = WORKERS_RE.match(line)
+            if m:
+                workers = int(m.group(1))
+                if current:
+                    rows.append(current)
+                current = {
+                    'bench': default_bench or path.stem,
+                    'workers': workers,
+                }
+                continue
+
         m = PARALLEL_THROUGHPUT_RE.match(line)
         if m:
             current['tps'] = float(m.group(1))
@@ -157,31 +181,30 @@ def has_multi_nc(df: pd.DataFrame) -> bool:
 
 def _tps_fig(df, color_col, title):
     fig = px.line(df, x='workers', y='tps', color=color_col,
+                  symbol=color_col,
                   markers=True, title=title,
-                  labels={'workers': 'Workers', 'tps': 'TPS', color_col: ''})
-    # fig.update_xaxes(
-    #     showgrid=True,
-    #     # griddash='dash',
-    # )
+                  labels={'workers': 'Workers', 'tps': 'TPS', color_col: ''},
+                  symbol_sequence=SHAPES)
     fig.update_yaxes(
-        # showgrid=True,
         nticks=25,
-        # griddash='dash',
     )
     fig.update_layout(template='plotly_white',
                       hovermode='x unified')
     return fig
 
 
 def _latency_fig(df, color_col, dash_col, title):
-    latency_cols = [c for c in df.columns if c.endswith('(ms)')]
+    latency_cols = [c for c in df.columns if c in DISPLAY_LATENCY]
     if not latency_cols:
         return None
     melted = df.melt(id_vars=[color_col, 'workers'], value_vars=latency_cols,
                      var_name='percentile', value_name='latency')
+    dash_map = {'avg (ms)': 'dash', 'p50 (ms)': 'solid', 'p95 (ms)': 'dot'}
     fig = px.line(melted, x='workers', y='latency',
                   color=color_col, line_dash=dash_col, markers=True, title=title,
-                  labels={'workers': 'Workers', 'latency': 'Latency (ms)', color_col: ''})
+                  labels={'workers': 'Workers',
+                          'latency': 'Latency (ms)', color_col: ''},
+                  line_dash_map=dash_map)
     fig.update_layout(template='plotly_white', hovermode='x unified')
     return fig
 
@@ -233,6 +256,9 @@ def parse_combined(dfs: dict[str, pd.DataFrame]):
     return dct
 
 
+SHAPES = ['circle', 'square', 'diamond', 'cross', 'hexagon', 'star']
+
+
 def make_combined_figures(dfs, local_dfs):
     figs = {"_All": None}
     all_dfs = []
@@ -265,12 +291,28 @@ def make_combined_figures(dfs, local_dfs):
     return figs
 
 
-def main():
+def _get_dir():
+    def_dir = Path(os.environ.get("DEF_BENCH", DEFAULT_BENCH_DIR))
+
     with st.sidebar:
         directory = st.text_input(
-            "Directory", value=DEFAULT_BENCH_DIR, key="benchdir")
+            "Directory", value=str(def_dir) if def_dir.exists() else "", key="benchdir")
+        os.environ["DEF_BENCH"] = directory
+
+    if not directory:
+        st.info(f"Please input a Folder in the Sidebar")
+        return None
     directory = Path(directory)
+    if not directory.exists():
+        st.error(f"Directory `{directory}` does not exist")
+        return None
+    return directory
 
+
+def main():
+    directory = _get_dir()
+    if not directory:
+        return
     single = {}
 
     for path in sorted(directory.glob("*.log")):

token/core/zkatdlog/nogh/v1/validator/bench/transfer_service/aws_setup_2_machines.md

@@ -0,0 +1,112 @@
+## SSH 
+Add ssh pubkey of server to client `known_hosts`
+
+## Server
+```bash
+GOGC=10000 go run ./server/
+```
+
+## Client 
+1. Rsync:
+
+```bash
+rsync -avz AWS-128:~/effi/fabric-token-sdk/token/core/zkatdlog/nogh/v1/validator/bench/transfer_service/out/ ./out
+```
+
+2. replacing ip
+
+```bash
+sed 's#127.0.0.1#ec2-54-90-141-176.compute-1.amazonaws.com#g' ./out/testdata/fsc/nodes/test-node.0/client-config.yaml -i
+```
+
+```bash
+GOGC=10000 nohup go run ./client/ -benchtime=30s -count=5 -workloads=transfer-service -cpu=1,2,4,8,16,32,48,64 -numConn=1,2,4,8 2>&1 | tee out.txt &
+```
+
+
+# Prometheus Setup 
+1. Download
+
+```bash
+sudo yum update -y && sudo yum install -y wget
+wget https://github.com/prometheus/prometheus/releases/download/v2.52.0/prometheus-2.52.0.linux-amd64.tar.gz
+tar xvf prometheus-2.52.0.linux-amd64.tar.gz
+sudo mv prometheus-2.52.0.linux-amd64/prometheus /usr/local/bin/
+sudo mv prometheus-2.52.0.linux-amd64/promtool /usr/local/bin/
+```
+
+2. Make config dir 
+
+```bash
+sudo mkdir /etc/prometheus
+sudo mkdir /var/lib/prometheus
+sudo cp prometheus-2.52.0.linux-amd64/prometheus.yml /etc/prometheus/
+```
+
+3. Configure Prometheus for node exporter
+```bash
+sudo nano /etc/prometheus/prometheus.yml
+```
+```yaml
+scrape_configs:
+  - job_name: "node"
+    static_configs:
+      - targets: ["localhost:9100"]
+```    
+4. Add Prometheus user
+```bash
+sudo useradd --no-create-home --shell /bin/false prometheus
+sudo chown -R prometheus:prometheus /etc/prometheus
+sudo chown -R prometheus:prometheus /var/lib/prometheus
+sudo chown -R $(whoami):$(whoami) /var/lib/prometheus
+```
+
+4. Run Prometheus
+```bash
+prometheus \
+  --config.file=/etc/prometheus/prometheus.yml \
+  --storage.tsdb.path=/var/lib/prometheus
+```
+
+# Install Node Exporter 
+```bash
+wget https://github.com/prometheus/node_exporter/releases/latest/download/node_exporter-1.10.2.linux-amd64.tar.gz
+tar xvf node_exporter-1.10.2.linux-amd64.tar.gz
+cd node_exporter-1.10.2.linux-amd64
+sudo mv node_exporter /usr/local/bin/
+```
+Run it
+```bash 
+node_exporter # see http://<EC2-IP>:9100/metrics
+```
+(verify scraping at: `http://<EC2-IP>:9090/targets`)
+
+
+# Install Graphana
+```bash
+sudo yum install grafana -y
+```
+Start Graphana
+```bash
+sudo systemctl daemon-reexec
+sudo systemctl start grafana-server
+sudo systemctl enable grafana-server
+```
+
+Now a vailable at: `http://<EC2-IP>:3000`
+user/pwd: `admin/admin`
+
+## Setup dashboard
+In Grafana UI > Go to Dashboards > Click Import  
+
+Enter dashboard ID:`1860`  
+
+In `http://<EC2-IP>:3000/connections/datasources`:  
+Add prometheus > Save and Test 
+
+if you get:
+```bash
+Post "http://127.0.0.1:9090/api/v1/query": dial tcp 127.0.0.1:9090: connect: permission denied - There was an error returned querying the Prometheus API.
+```
+
+Do `sudo setenforce 0`