metrics: HonchoDeriverBacklog and include honcho in target-down
This commit is contained in:
@@ -162,6 +162,13 @@ in
|
||||
labels.severity = "page";
|
||||
annotations.summary = "tether client-disconnect / converted > 50% over 15m";
|
||||
}
|
||||
{
|
||||
alert = "HonchoDeriverBacklog";
|
||||
expr = ''max(deriver_queue_items_pending) > 100 and max(deriv(deriver_queue_oldest_pending_age_seconds[30m])) > 0'';
|
||||
for = "10m";
|
||||
labels.severity = "page";
|
||||
annotations.summary = "Honcho pending queue > 100 and oldest age rising";
|
||||
}
|
||||
{
|
||||
alert = "LlamaServerRestarting";
|
||||
expr = ''increase(node_systemd_service_restart_total{name=~"llama-server.service|llama-vision.service"}[15m]) > 0'';
|
||||
@@ -202,7 +209,7 @@ in
|
||||
}
|
||||
{
|
||||
alert = "FleetTargetDown";
|
||||
expr = ''up{job=~"vm|node|nvidia-dixie|llama-dixie-helper|llama-dixie-vision|tether|sift"} == 0'';
|
||||
expr = ''up{job=~"vm|node|nvidia-dixie|llama-dixie-helper|llama-dixie-vision|tether|sift|honcho-api|honcho-deriver"} == 0'';
|
||||
for = "2m";
|
||||
labels.severity = "page";
|
||||
annotations.summary = "scrape target down ({{ $labels.job }} {{ $labels.instance }})";
|
||||
|
||||
Reference in New Issue
Block a user