Merge pull request #9742 from ycedres/promtail-container-adjustments

Adjust promtail container exec script

health-check/src/uyuni_health_check/containers/promtail/Dockerfile

@@ -4,10 +4,5 @@ FROM opensuse/leap:latest
 COPY promtail-linux-amd64 /usr/bin/promtail
 RUN chmod a+x /usr/bin/promtail
 
-RUN zypper -n in python311 python311-requests
-COPY complete_checker.py /usr/bin/complete_checker.py
-
-COPY run.py /usr/bin/run.py
-RUN chmod +x /usr/bin/run.py
-
-ENTRYPOINT ["/usr/bin/run.py"]
+ENTRYPOINT ["promtail"]
+CMD ["--config.file=/etc/promtail/config.yml"]

health-check/src/uyuni_health_check/loki/loki_manager.py

@@ -91,7 +91,6 @@ def run_loki(supportconfig_path=None, verbose=False):
     ]
 
     podman(podman_args, verbose)
-    wait_promtail_init()
 
 
 def render_promtail_cfg(supportconfig_path=None, promtail_template=None):
@@ -130,167 +129,3 @@ def render_promtail_cfg(supportconfig_path=None, promtail_template=None):
 
     # Write rendered promtail configuration file
     config.write_config("promtail", "config.yaml", promtail_template.render(**opts))
-
-
-def check_series_in_loki(
-    loki_url,
-    job_name="promtail-complete-job",
-    flag="complete",
-    message="Promtail finished!d",
-):
-    query = f'{{job="{job_name}", flag="{flag}"}} |= "{message}"'
-    end = int(time.time())
-    start = end - 60 * 60
-
-    response = requests.get(
-        f"{loki_url}/loki/api/v1/query",
-        params={
-            "query": query,
-            "start": start * 1_000_000_000,
-            "end": end * 1_000_000_000,
-        },
-    )
-
-    if response.status_code == 200:
-        data = response.json()
-        return len(data["data"]["result"]) > 0
-    else:
-        print("Failed to query Loki:", response.text)
-        return False
-
-
-def wait_promtail_init():
-    loki_url = "http://localhost:3100"
-    start_time = time.time()
-    timeout = 60
-    console.log("[bold]Waiting for Promtail to process logs")
-
-    while not check_series_in_loki(loki_url):
-        elapsed_time = time.time() - start_time
-        if elapsed_time >= timeout:
-            console.log("Timeout waiting for promtail to finish!")
-            break
-        time.sleep(10)
-    console.log("Promtail finished processing logs")
-
-
-def wait_loki_init(verbose=False):
-    """
-    Try to figure out when loki is ready to answer our requests.
-    There are two things to wait for:
-      - loki to be up
-      - promtail to have read the logs and the loki ingester having handled them
-    """
-    metrics = {}
-    timeout = False
-    request_message_bytes_sum = 0
-    loki_ingester_chunk_entries_count = 0
-    start_time = time.time()
-    ready = False
-
-    # Wait for promtail to be ready
-    # TODO Add a timeout here in case something went really bad
-    # TODO checking the lags won't work when working on older logs,
-    # we could try to compare the positions with the size of the files in such a case
-    while (
-        not metrics
-        or metrics["active"] < PROMTAIL_TARGETS
-        or (not metrics["lags"] and metrics["active_files"] == 0)
-        or any([v >= 10 for v in metrics["lags"].values()])
-        or (metrics["lags"] and metrics["active_files"])
-        or (metrics["encoded_bytes_total"] == metrics["sent_bytes_total"] == 0)
-        or (metrics["encoded_bytes_total"] != metrics["sent_bytes_total"] != 0)
-        or (
-            metrics["encoded_bytes_total"]
-            != metrics["sent_bytes_total"]
-            != request_message_bytes_sum
-        )
-        or loki_ingester_chunk_entries_count == 0
-        or not ready
-        and not timeout
-    ):
-        if verbose:
-            console.log("Waiting for promtail metrics to be collected")
-        time.sleep(5)
-        response = requests.get("http://localhost:3100/metrics")
-        if verbose:
-            console.log("loki metrics 3100 status code", response.status_code)
-        if response.status_code == 200:
-            content = response.content.decode()
-            request_message_bytes_sum = re.findall(
-                r'loki_request_message_bytes_sum{.*"loki_api_v1_push"} (\d+)', content
-            )
-            request_message_bytes_sum = (
-                int(request_message_bytes_sum[0]) if request_message_bytes_sum else 0
-            )
-            loki_ingester_chunk_entries_count = re.findall(
-                r"loki_ingester_chunk_entries_count (\d+)", content
-            )
-            loki_ingester_chunk_entries_count = (
-                int(loki_ingester_chunk_entries_count[0])
-                if loki_ingester_chunk_entries_count
-                else 0
-            )
-
-        response = requests.get("http://localhost:9081/metrics")
-        if verbose:
-            console.log("promtail metrics 9081 status code", response.status_code)
-        if response.status_code == 200:
-            content = response.content.decode()
-            active = re.findall(r"promtail_targets_active_total (\d+)", content)
-            encoded_bytes_total = re.findall(
-                r"promtail_encoded_bytes_total{.*} (\d+)", content
-            )
-            sent_bytes_total = re.findall(
-                r"promtail_sent_bytes_total{.*} (\d+)", content
-            )
-            active_files = re.findall(r"promtail_files_active_total (\d+)", content)
-            lags = re.findall(
-                'promtail_stream_lag_seconds{filename="([^"]+)".*} ([0-9.]+)', content
-            )
-            metrics = {
-                "lags": {row[0]: float(row[1]) for row in lags},
-                "active": int(active[0]) if active else 0,
-                "active_files": int(active_files[0]) if active_files else 0,
-                "encoded_bytes_total": (
-                    int(encoded_bytes_total[0]) if encoded_bytes_total else 0
-                ),
-                "sent_bytes_total": int(sent_bytes_total[0]) if sent_bytes_total else 0,
-            }
-
-        # check if loki is ready
-        if verbose:
-            console.log("Waiting for loki to be ready")
-        response = requests.get("http://localhost:3100/ready")
-        if verbose:
-            console.log("loki 3100 status code", response.status_code)
-        if response.status_code == 200:
-            content = response.content.decode()
-            if content == "ready\n":
-                ready = True
-
-        # check if promtail is ready
-        if verbose:
-            console.log("Waiting for promtail to be ready")
-        response = requests.get("http://localhost:9081/ready")
-        if verbose:
-            console.log("promtail ready 9081 status code", response.status_code)
-        if response.status_code == 200:
-            content = response.content.decode()
-            if content == "Ready":
-                ready = True
-            else:
-                ready = False
-
-        # check timeout
-        if (time.time() - start_time) > LOKI_WAIT_TIMEOUT:
-            timeout = True
-    if timeout:
-        raise HealthException(
-            "[red bold]Timeout has been reached waiting for Loki and promtail. Something unexpected may happen. Please check and try again."
-        )
-    else:
-        console.print(metrics)
-        console.print(loki_ingester_chunk_entries_count)
-        console.print(request_message_bytes_sum)
-        console.log("[bold]Loki and promtail are now ready to receive requests")