feat: add kernel firewall log parsing to agent with blocked traffic detection and dashboard suspicious traffic enhancement

Add parse_firewall_log_line to extract SRC/DST/PROTO/SPT/DPT/LEN from kernel log lines with drop/reject/accept decision classification, implement collect_firewall_log_flows to parse journalctl -k output from last 5 minutes with flow aggregation by 5-tuple+decision, add merge_flow_sources to combine packet flows and firewall log flows with deduplication, extend agent config with
This commit is contained in:
2026-07-09 21:07:03 +02:00
parent 68c11eba57
commit 35ffcb6768
3 changed files with 142 additions and 12 deletions
+115 -4
View File
@@ -18,9 +18,10 @@ from pathlib import Path
from typing import Any
VERSION = "0.2.0"
VERSION = "0.2.1"
VM_INTERFACE_RE = re.compile(r"(?:tap|fwbr|fwln|fwpr)(\d+)")
VM_INTERFACE_DETAIL_RE = re.compile(r"(?:tap|fwbr|fwln|fwpr)(\d+)i(\d+)")
LOG_FIELD_RE = re.compile(r"\b([A-Z]+)=([^\s]+)")
ETH_P_IP = 0x0800
ETH_P_ALL = 0x0003
ETH_P_8021Q = 0x8100
@@ -326,6 +327,100 @@ def collect_flows(limit: int = 500) -> list[dict[str, Any]]:
return flows
def parse_firewall_log_line(line: str) -> dict[str, Any] | None:
fields = {key: value for key, value in LOG_FIELD_RE.findall(line)}
source_ip = fields.get("SRC")
destination_ip = fields.get("DST")
protocol = fields.get("PROTO", "").lower()
if not source_ip or not destination_ip or protocol not in {"tcp", "udp", "icmp"}:
return None
lowered = line.lower()
decision = None
if any(token in lowered for token in ("drop", "reject", "deny", "blocked")):
decision = "blocked"
elif any(token in lowered for token in ("accept", "allow")):
decision = "allowed"
if not decision:
return None
source_port = fields.get("SPT")
destination_port = fields.get("DPT")
length = fields.get("LEN")
return {
"source_ip": source_ip,
"destination_ip": destination_ip,
"protocol": protocol,
"source_port": int(source_port) if source_port and source_port.isdigit() else None,
"destination_port": int(destination_port) if destination_port and destination_port.isdigit() else None,
"packets": 1,
"bytes": int(length) if length and length.isdigit() else 0,
"state": decision,
"decision": decision,
"collector": "firewall-log",
"log_excerpt": line[-500:],
}
def collect_firewall_log_flows(since_minutes: int = 5, limit: int = 500) -> tuple[list[dict[str, Any]], dict[str, Any]]:
diagnostics = {"collector": "journalctl-kernel", "since_minutes": since_minutes, "errors": []}
code, output = run_command(
["journalctl", "-k", "--since", f"-{max(since_minutes, 1)} min", "--no-pager", "-o", "cat"],
timeout=10,
)
if code != 0 or not output:
diagnostics["errors"].append(output or "journalctl returned no firewall log output")
return [], diagnostics
flows: dict[tuple[object, ...], dict[str, Any]] = {}
for line in output.splitlines():
if "SRC=" not in line or "DST=" not in line:
continue
flow = parse_firewall_log_line(line)
if not flow:
continue
key = (
flow["source_ip"],
flow["destination_ip"],
flow["protocol"],
flow.get("source_port"),
flow.get("destination_port"),
flow.get("decision"),
)
current = flows.get(key)
if current:
current["packets"] += 1
current["bytes"] += int(flow.get("bytes") or 0)
continue
flows[key] = flow
if len(flows) >= limit:
diagnostics["truncated"] = True
break
diagnostics["flow_count"] = len(flows)
return sorted(flows.values(), key=lambda item: int(item.get("packets") or 0), reverse=True), diagnostics
def merge_flow_sources(*sources: list[dict[str, Any]], limit: int = 500) -> list[dict[str, Any]]:
flows: dict[tuple[object, ...], dict[str, Any]] = {}
for source in sources:
for flow in source:
key = (
flow.get("source_ip"),
flow.get("destination_ip"),
flow.get("protocol"),
flow.get("source_port"),
flow.get("destination_port"),
flow.get("decision") or flow.get("state") or "observed",
)
current = flows.get(key)
if current:
current["packets"] = int(current.get("packets") or 0) + int(flow.get("packets") or 0)
current["bytes"] = int(current.get("bytes") or 0) + int(flow.get("bytes") or 0)
continue
flows[key] = dict(flow)
return sorted(flows.values(), key=lambda item: int(item.get("bytes") or 0), reverse=True)[:limit]
def collect_conntrack() -> dict[str, Any]:
code, output = run_command(["conntrack", "-C"])
if code == 0 and output.isdigit():
@@ -339,11 +434,17 @@ def collect_conntrack() -> dict[str, Any]:
return {"count": None, "source": "unavailable"}
def collect_flow_diagnostics(conntrack: dict[str, Any], flow_count: int, packet_diagnostics: dict[str, Any] | None = None) -> dict[str, Any]:
def collect_flow_diagnostics(
conntrack: dict[str, Any],
flow_count: int,
packet_diagnostics: dict[str, Any] | None = None,
firewall_log_diagnostics: dict[str, Any] | None = None,
) -> dict[str, Any]:
code, conntrack_path = run_command(["sh", "-c", "command -v conntrack"])
return {
"flow_count": flow_count,
"packet_collector": packet_diagnostics,
"firewall_log_collector": firewall_log_diagnostics,
"conntrack_binary": conntrack_path if code == 0 else None,
"conntrack_count": conntrack.get("count"),
"conntrack_source": conntrack.get("source"),
@@ -372,13 +473,20 @@ def collect_payload(config: dict[str, Any]) -> dict[str, Any]:
flow_limit = int(config.get("flow_limit", 500))
packet_flows: list[dict[str, Any]] = []
packet_diagnostics: dict[str, Any] | None = None
firewall_log_flows: list[dict[str, Any]] = []
firewall_log_diagnostics: dict[str, Any] | None = None
if bool(config.get("packet_flow_collector", True)):
packet_flows, packet_diagnostics = collect_packet_flows(
interfaces,
int(config.get("packet_flow_window_seconds", 10)),
flow_limit,
)
flows = packet_flows or collect_flows(flow_limit)
if bool(config.get("firewall_log_collector", True)):
firewall_log_flows, firewall_log_diagnostics = collect_firewall_log_flows(
int(config.get("firewall_log_window_minutes", 5)),
flow_limit,
)
flows = merge_flow_sources(packet_flows or collect_flows(flow_limit), firewall_log_flows, limit=flow_limit)
conntrack = collect_conntrack()
return {
"version": VERSION,
@@ -393,7 +501,10 @@ def collect_payload(config: dict[str, Any]) -> dict[str, Any]:
"flows": flows,
"conntrack": conntrack,
"firewall": collect_firewall(),
"extra": {"platform": platform.platform(), "flow_diagnostics": collect_flow_diagnostics(conntrack, len(flows), packet_diagnostics)},
"extra": {
"platform": platform.platform(),
"flow_diagnostics": collect_flow_diagnostics(conntrack, len(flows), packet_diagnostics, firewall_log_diagnostics),
},
}