JSON from args + merged to one container

docker-compose.dev.yml

@@ -17,10 +17,4 @@ services:
       - --development
       - --port=14000
       - --dont-delete-pdfs
-  snmp-collector:
-    build:
-      context: .
-      dockerfile: ./collector/Dockerfile
-    ports:
-      - 5000:5000
-
+      - --config-json-path=/app/config/config.json

docker-compose.yml

@@ -13,10 +13,7 @@ services:
       - ~/.ssh/known_hosts:/app/known_hosts
       - "/etc/cups/ppd/:/etc/cups/ppd"
     tty: true
-  snmp-collector:
-    build:
-      context: .
-      dockerfile: ./collector/Dockerfile
+
 # we attach the print container to an external docker
 # network called "poweredge". we do this so a prometheus
 # container can pull metrics from the server over HTTP

printer/Dockerfile

@@ -1,20 +1,19 @@
 # Base image from https://github.com/DrPsychick/docker-cups-airprint
 # Docker images are here https://hub.docker.com/r/drpsychick/airprint-bridge/tags
-FROM drpsychick/airprint-bridge:latest
+FROM drpsychick/airprint-bridge:jammy
 
 WORKDIR /app
-
 RUN apt-get update
 
 RUN apt install -y python3 python3-pip python3-venv jq ssh
 
-# Create a virtual environment
+# Create the virtual environment with Python
 RUN python3 -m venv /opt/venv
 
 # Set the virtual environment as the default Python environment
 ENV PATH="/opt/venv/bin:$PATH"
 
-COPY ./printer/requirements.txt /app/printer/requirements.txt
+COPY ./printer/requirements.txt  /app/printer/requirements.txt
 
 RUN /opt/venv/bin/pip install -r /app/printer/requirements.txt
 
@@ -29,3 +28,4 @@ EXPOSE 9000
 # The below command runs the bash script that sets up the connection to the
 # printers and runs PrintHandler.js
 ENTRYPOINT [ "./printer/what.sh" ]
+

printer/collector.py

@@ -0,0 +1,106 @@
+import time
+import enum
+import logging
+import json
+import asyncio
+
+from pysnmp.hlapi import *
+
+from metrics import MetricsHandler
+
+metrics_handler = MetricsHandler.instance()
+
+
+logging.basicConfig(
+    # in mondo we trust
+    format="%(asctime)s.%(msecs)03dZ %(levelname)s:%(name)s:%(message)s",
+    datefmt="%Y-%m-%dT%H:%M:%S",
+    level=logging.INFO,
+)
+
+
+class SnmpOid(enum.Enum):
+    INK_LEVEL = ("ink_level", "1.3.6.1.2.1.43.11.1.1.9.1.1")
+    INK_CAPACITY = ("ink_capacity", "1.3.6.1.2.1.43.11.1.1.8.1.1")
+    PAGE_COUNT = ("page_count", "1.3.6.1.2.1.43.10.2.1.4.1.1")
+    TRAY_EMPTY = ("tray_empty", "1.3.6.1.2.1.43.18.1.1.8.1.13", True)
+    # we observed each printer emitting a different SNMP OID for
+    # an empty paper tray, the below accounts for this second OID.
+    # the _2 at the end of this metric does imply that the printer has
+    # 2 trays. tray_empty_2 is an indication of the same exact issue
+    # as tray_empty: an empty paper tray.
+    TRAY_EMPTY_2 = ("tray_empty_2", "1.3.6.1.2.1.43.18.1.1.8.1.2", True)
+
+    def __init__(self, metric_name, metric_value, is_error=False):
+        self.metric_name = metric_name
+        self.metric_value = metric_value
+        self.is_error = is_error
+
+
+def fetch_ips_from_config(config_file_path):
+    try:
+        with open(config_file_path, "r") as f:
+            config = json.load(f)
+            printer_configs = config.get("PRINTING")
+            if not printer_configs:
+                raise Exception("No printers defined in config file")
+
+            ip_list = []
+            for printer in printer_configs:
+                if isinstance(printer_configs[printer], dict):
+                    ip = printer_configs[printer]["IP"]
+                    logging.info(f"Adding printer {printer} with IP {ip}")
+                    ip_list.append(ip)
+            return ip_list
+
+    except Exception as e:
+        logging.error(f"error opening config file: {e}")
+
+
+def scrape_snmp(ip_list, sleep_duration_minutes=5):
+    while True:
+        for ip in ip_list:
+            get_snmp_data(ip)
+        time.sleep(sleep_duration_minutes * 60)
+
+
+def get_snmp_data(ip):
+    for oid in SnmpOid:
+        with metrics_handler.snmp_request_duration.time():
+            errorIndication, errorStatus, errorIndex, varBinds = next(
+                getCmd(
+                    SnmpEngine(),
+                    CommunityData("public", mpModel=0),
+                    UdpTransportTarget((ip, 161)),
+                    ContextData(),
+                    ObjectType(ObjectIdentity(oid.metric_value)),
+                )
+            )
+        if errorIndication:
+            logging.error(
+                f"Error indication from {ip} for metric {oid.metric_value}: {errorIndication}"
+            )
+            metrics_handler.device_unreachable.set(1)
+            continue
+        if errorStatus:
+            logging.error(
+                f"Error status from {ip} for metric {oid.metric_value}: {errorStatus.prettyPrint()}"
+            )
+            # SNMP OIDs related to errors often dissappear when
+            # the associated issue that the metric refers to is
+            # no longer present (i.e. an empty tray now has
+            # paper). To avoid leaving an error metric as 1
+            # which would create a false positive, set the metric
+            # to zero if the associated SNMP OID was not found
+            if oid.is_error:
+                metrics_handler.snmp_error.labels(name=oid.metric_name, ip=ip).set(0)
+            continue
+
+        metrics_handler.device_unreachable.set(0)
+        if not varBinds:
+            continue
+        res = varBinds[0]
+        if oid.is_error:
+            metrics_handler.snmp_error.labels(name=oid.metric_name, ip=ip).set(1)
+            continue
+        metrics_handler.snmp_metric.labels(name=oid.metric_name, ip=ip).set(res[1])

printer/metrics.py

@@ -23,6 +23,31 @@ class Metrics(enum.Enum):
         "total bytes of files pointed to by cache",
         prometheus_client.Gauge,
     )
+    SNMP_METRIC = (
+        "snmp_metric",
+        "ex: Number of pages printed",
+        prometheus_client.Gauge,
+        ["name", "ip"],
+    )
+
+    SNMP_ERROR = (
+        "snmp_error",
+        "Error metrics",
+        prometheus_client.Gauge,
+        ["name", "ip"],
+    )
+
+    SNMP_REQ_DURATION = (
+        "snmp_request_duration",
+        "Time it took for SNMP request",
+        prometheus_client.Summary,
+    )
+
+    DEVICE_UNREACHABLE = (
+        "device_unreachable",
+        "set to 1 when error",
+        prometheus_client.Gauge,
+    )
 
     def __init__(self, title, description, prometheus_type, labels=()):
         # we use the above default value for labels because it matches what's used
@@ -38,8 +63,8 @@ class MetricsHandler:
     _instance = None
 
     def __init__(self):
-        raise RuntimeError('Call MetricsHandler.instance() instead')
-    
+        raise RuntimeError("Call MetricsHandler.instance() instead")
+
     def init(self) -> None:
         for metric in Metrics:
             setattr(

printer/requirements.txt

@@ -4,3 +4,5 @@ py-grpc-prometheus==0.7.0
 python-multipart==0.0.9
 httpx==0.28.1
 requests==2.32.3
+pysnmp==4.4.12
+pyasn1==0.4.8