Fix fail2ban runtime errors: jail not found, action locks, log noise

This commit implements fixes for three independent bugs in the fail2ban configuration and integration layer:

1. Task 1: Detect UnknownJailException and prevent silent failures
   - Added JailNotFoundError detection in jail_service.reload_all()
   - Enhanced error handling in config_file_service to catch JailNotFoundError
   - Added specific error message with logpath validation hints
   - Added rollback test for this scenario

2. Task 2: Fix iptables-allports exit code 4 (xtables lock contention)
   - Added global banaction setting in jail.conf with -w 5 lockingopt
   - Removed redundant per-jail banaction overrides from bangui-sim and blocklist-import
   - Added production compose documentation note

3. Task 3: Suppress log noise from unsupported backend/idle commands
   - Implemented capability detection to cache command support status
   - Double-check locking to minimize lock contention
   - Avoids sending unsupported get <jail> backend/idle commands
   - Returns default values without socket calls when unsupported

All changes include comprehensive tests and maintain backward compatibility.
This commit is contained in:
2026-03-15 10:57:00 +01:00
parent 1e33220f59
commit f62785aaf2
8 changed files with 446 additions and 145 deletions

View File

@@ -184,10 +184,90 @@ class TestListJails:
with patch("app.services.jail_service.Fail2BanClient", _FailClient), pytest.raises(Fail2BanConnectionError):
await jail_service.list_jails(_SOCKET)
async def test_backend_idle_commands_unsupported(self) -> None:
"""list_jails handles unsupported backend and idle commands gracefully.
# ---------------------------------------------------------------------------
# get_jail
# ---------------------------------------------------------------------------
When the fail2ban daemon does not support get ... backend/idle commands,
list_jails should not send them, avoiding "Invalid command" errors in the
fail2ban log.
"""
# Reset the capability cache to test detection.
jail_service._backend_cmd_supported = None
responses = {
"status": _make_global_status("sshd"),
"status|sshd|short": _make_short_status(),
# Capability probe: get backend fails (command not supported).
"get|sshd|backend": (1, Exception("Invalid command (no get action or not yet implemented)")),
# Subsequent gets should still work.
"get|sshd|bantime": (0, 600),
"get|sshd|findtime": (0, 600),
"get|sshd|maxretry": (0, 5),
}
with _patch_client(responses):
result = await jail_service.list_jails(_SOCKET)
# Verify the result uses the default values for backend and idle.
jail = result.jails[0]
assert jail.backend == "polling" # default
assert jail.idle is False # default
# Capability should now be cached as False.
assert jail_service._backend_cmd_supported is False
async def test_backend_idle_commands_supported(self) -> None:
"""list_jails detects and sends backend/idle commands when supported."""
# Reset the capability cache to test detection.
jail_service._backend_cmd_supported = None
responses = {
"status": _make_global_status("sshd"),
"status|sshd|short": _make_short_status(),
# Capability probe: get backend succeeds.
"get|sshd|backend": (0, "systemd"),
# All other commands.
"get|sshd|bantime": (0, 600),
"get|sshd|findtime": (0, 600),
"get|sshd|maxretry": (0, 5),
"get|sshd|idle": (0, True),
}
with _patch_client(responses):
result = await jail_service.list_jails(_SOCKET)
# Verify real values are returned.
jail = result.jails[0]
assert jail.backend == "systemd" # real value
assert jail.idle is True # real value
# Capability should now be cached as True.
assert jail_service._backend_cmd_supported is True
async def test_backend_idle_commands_cached_after_first_probe(self) -> None:
"""list_jails caches capability result and reuses it across polling cycles."""
# Reset the capability cache.
jail_service._backend_cmd_supported = None
responses = {
"status": _make_global_status("sshd, nginx"),
# Probes happen once per polling cycle (for the first jail listed).
"status|sshd|short": _make_short_status(),
"status|nginx|short": _make_short_status(),
# Capability probe: backend is unsupported.
"get|sshd|backend": (1, Exception("Invalid command")),
# Subsequent jails do not trigger another probe; they use cached result.
# (The mock doesn't have get|nginx|backend because it shouldn't be called.)
"get|sshd|bantime": (0, 600),
"get|sshd|findtime": (0, 600),
"get|sshd|maxretry": (0, 5),
"get|nginx|bantime": (0, 600),
"get|nginx|findtime": (0, 600),
"get|nginx|maxretry": (0, 5),
}
with _patch_client(responses):
result = await jail_service.list_jails(_SOCKET)
# Both jails should return default values (cached result is False).
for jail in result.jails:
assert jail.backend == "polling"
assert jail.idle is False
class TestGetJail:
@@ -339,6 +419,28 @@ class TestJailControls:
_SOCKET, include_jails=["new"], exclude_jails=["old"]
)
async def test_reload_all_unknown_jail_raises_jail_not_found(self) -> None:
"""reload_all detects UnknownJailException and raises JailNotFoundError.
When fail2ban cannot load a jail due to invalid configuration (e.g.,
missing logpath), it raises UnknownJailException during reload. This
test verifies that reload_all detects this and re-raises as
JailNotFoundError instead of the generic JailOperationError.
"""
with _patch_client(
{
"status": _make_global_status("sshd"),
"reload|--all|[]|[['start', 'airsonic-auth'], ['start', 'sshd']]": (
1,
Exception("UnknownJailException('airsonic-auth')"),
),
}
), pytest.raises(jail_service.JailNotFoundError) as exc_info:
await jail_service.reload_all(
_SOCKET, include_jails=["airsonic-auth"]
)
assert exc_info.value.name == "airsonic-auth"
async def test_start_not_found_raises(self) -> None:
"""start_jail raises JailNotFoundError for unknown jail."""
with _patch_client({"start|ghost": (1, Exception("Unknown jail: 'ghost'"))}), pytest.raises(JailNotFoundError):