viktor/infra
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

[ci skip] Fix Uptime Kuma false-down reports: use bulk heartbeat API instead of per-monitor calls

Browse source
This commit is contained in:
Viktor Barzin 2026-02-22 01:37:28 +00:00
parent 0ff2aaec60
commit 00dc78e0d2
2 changed files with 45 additions and 37 deletions
Download patch file Download diff file Expand all files Collapse all files

43
.claude/cluster-health.sh
View file

@ -626,7 +626,7 @@ except ImportError:
sys.exit(0) sys.exit(0)
try: try:
api = UptimeKumaApi("https://uptime.viktorbarzin.me") api = UptimeKumaApi("https://uptime.viktorbarzin.me", timeout=30)
password = os.environ.get("UPTIME_KUMA_PASSWORD", "") password = os.environ.get("UPTIME_KUMA_PASSWORD", "")
if not password: if not password:
print("ERROR:UPTIME_KUMA_PASSWORD not set") print("ERROR:UPTIME_KUMA_PASSWORD not set")
@ -634,31 +634,36 @@ try:
api.login("admin", password) api.login("admin", password)
monitors = api.get_monitors() monitors = api.get_monitors()
down = [] # Build id->name map and track active/paused
up_count = 0 id_to_name = {}
paused_count = 0 paused_count = 0
for m in monitors: for m in monitors:
mid = m.get("id")
name = m.get("name", "unknown") name = m.get("name", "unknown")
active = m.get("active", True) active = m.get("active", True)
if not active: if not active:
paused_count += 1 paused_count += 1
continue
# Check heartbeat list for latest status
try:
hb = api.get_monitor_beats(m["id"], 1)
if hb and len(hb) > 0:
status = hb[-1].get("status", 0)
else:
status = m.get("status", 0)
except Exception:
status = m.get("status", 0)
# status: 0=DOWN, 1=UP, 2=PENDING, 3=MAINTENANCE
if status == 1:
up_count += 1
elif status == 3:
paused_count += 1
else: else:
id_to_name[mid] = name
# Use bulk heartbeat fetch (single API call) instead of per-monitor calls
heartbeats = api.get_heartbeats()
down = []
up_count = 0
for mid, name in id_to_name.items():
beats = heartbeats.get(mid, [])
if beats:
status = beats[-1].get("status", 0)
# Handle both enum and int (MonitorStatus.UP == 1)
if status == 1:
up_count += 1
elif status == 3:
paused_count += 1
else:
down.append(name)
else:
# No heartbeats = unknown, treat as down
down.append(name) down.append(name)
api.disconnect() api.disconnect()

39
scripts/cluster_healthcheck.sh
View file

@ -594,34 +594,37 @@ except ImportError:
sys.exit(0) sys.exit(0)
try: try:
api = UptimeKumaApi("https://uptime.viktorbarzin.me") api = UptimeKumaApi("https://uptime.viktorbarzin.me", timeout=30)
api.login("admin", "EUxhLr4w4NFsGehy") api.login("admin", "EUxhLr4w4NFsGehy")
monitors = api.get_monitors() monitors = api.get_monitors()
down = [] # Build id->name map and track active/paused
up_count = 0 id_to_name = {}
paused_count = 0 paused_count = 0
for m in monitors: for m in monitors:
mid = m.get("id")
name = m.get("name", "unknown") name = m.get("name", "unknown")
active = m.get("active", True) active = m.get("active", True)
if not active: if not active:
paused_count += 1 paused_count += 1
continue else:
# Check heartbeat list for latest status id_to_name[mid] = name
try:
hb = api.get_monitor_beats(m["id"], 1) # Use bulk heartbeat fetch (single API call) instead of per-monitor calls
if hb and len(hb) > 0: heartbeats = api.get_heartbeats()
status = hb[-1].get("status", 0)
down = []
up_count = 0
for mid, name in id_to_name.items():
beats = heartbeats.get(mid, [])
if beats:
status = beats[-1].get("status", 0)
if status == 1:
up_count += 1
elif status == 3:
paused_count += 1
else: else:
status = m.get("status", 0) down.append(name)
except Exception:
status = m.get("status", 0)
# status: 0=DOWN, 1=UP, 2=PENDING, 3=MAINTENANCE
if status == 1:
up_count += 1
elif status == 3:
paused_count += 1
else: else:
down.append(name) down.append(name)