Compare commits

..

1 Commits

Author SHA1 Message Date
Andrew Stoltz
55729a24f9 monitoring: alert on remotedesktop longhorn pvc growth 2026-05-19 12:34:28 -05:00
2 changed files with 58 additions and 205 deletions

View File

@@ -656,15 +656,14 @@ data:
summary: "Print queue backlog on edge2 ({{ $value }} active jobs)" summary: "Print queue backlog on edge2 ({{ $value }} active jobs)"
description: "CUPS has {{ $value }} active jobs queued. Possible printer jam, USB disconnect, or paper out." description: "CUPS has {{ $value }} active jobs queued. Possible printer jam, USB disconnect, or paper out."
# Printer hardware and paper-roll lifecycle alerts. # Paper roll lifecycle alerts (XL Track I, 2026-04-26).
# print_printer_online: 1 when the transport is reachable/selected. # Source-of-truth gauge: print_paper_remaining_percent (Print.Web OTEL,
# print_printer_state enum: 0 unknown, 1 online, 2 offline, # hydrated on startup from the active PaperRoll row).
# 3 paper_depleted, 4 jam, 5 head_error, 6 cover_open. # alert_channel=thermal_print routes through irc-notify -> Print.Web
# Offline/jam/cover alerts stay IRC-only. Paper depleted and head # /api/print/alert so the printer announces its own paper-out warning
# error may route to the thermal digest only when the printer is # on its remaining paper. Self-referential humor + operator nudge.
# online enough to make that useful.
- alert: PrintPaperRollLow - alert: PrintPaperRollLow
expr: (print_paper_remaining_percent{job="printweb-otel"} < 10 and print_paper_remaining_percent{job="printweb-otel"} > 5) and print_printer_online{job="printweb-otel"} == 1 expr: print_paper_remaining_percent{job="printweb-otel"} < 10 and print_paper_remaining_percent{job="printweb-otel"} > 5
for: 5m for: 5m
labels: labels:
severity: warning severity: warning
@@ -673,59 +672,15 @@ data:
summary: "Print roll low on edge2 ({{ $value | printf \"%.1f\" }}% remaining)" summary: "Print roll low on edge2 ({{ $value | printf \"%.1f\" }}% remaining)"
description: "NuPrint 210 paper roll has {{ $value | printf \"%.1f\" }}% remaining. Operator should load a fresh roll soon. Run /api/paper/status for the precise mm + estimated jobs left." description: "NuPrint 210 paper roll has {{ $value | printf \"%.1f\" }}% remaining. Operator should load a fresh roll soon. Run /api/paper/status for the precise mm + estimated jobs left."
- alert: PrinterOfflineWarning
expr: print_printer_state{job="printweb-otel"} == 2
for: 2m
labels:
severity: warning
service: print-web
alert_channel: irc
annotations:
summary: "Print.Web printer offline on edge2"
description: "Print.Web reports the NuPrint 210 transport is offline or unreachable. IRC-only by design: do not thermal-print an alert when the thermal printer itself is offline."
- alert: PrintPaperRollCritical - alert: PrintPaperRollCritical
expr: print_printer_state{job="printweb-otel"} == 3 and print_printer_online{job="printweb-otel"} == 1 expr: print_paper_remaining_percent{job="printweb-otel"} <= 5
for: 2m for: 2m
labels: labels:
severity: critical severity: critical
alert_channel: thermal_print alert_channel: thermal_print
annotations: annotations:
summary: "Print paper depleted on edge2" summary: "Print roll critical on edge2 ({{ $value | printf \"%.1f\" }}% remaining)"
description: "NuPrint 210 reports paper depleted while the printer is still online. Load a new roll, drain the hardware buffer if needed, then replay DeadLetter jobs from /print-log." description: "NuPrint 210 paper roll at {{ $value | printf \"%.1f\" }}% — load a new roll NOW. The 50ft roll has a ~12% red-stripe zone; once paper passes that, the printer can run dry mid-job."
- alert: PrinterJamWarning
expr: print_printer_state{job="printweb-otel"} == 4
for: 2m
labels:
severity: warning
service: print-web
alert_channel: irc
annotations:
summary: "Print.Web printer jam on edge2"
description: "Print.Web reports a paper/cutter jam state. IRC-only: clear the jam, drain the hardware buffer if bytes were queued, then retry affected jobs."
- alert: PrinterHeadErrorCritical
expr: print_printer_state{job="printweb-otel"} == 5
for: 2m
labels:
severity: critical
service: print-web
alert_channel: thermal_print
annotations:
summary: "Print.Web printer head error on edge2"
description: "Print.Web reports a thermal head or unrecoverable printer error. Critical routing may enter the thermal digest per existing policy; IRC remains the primary triage stream."
- alert: PrinterCoverOpenWarning
expr: print_printer_state{job="printweb-otel"} == 6
for: 2m
labels:
severity: warning
service: print-web
alert_channel: irc
annotations:
summary: "Print.Web printer cover open on edge2"
description: "Print.Web reports the printer cover/lid is open. IRC-only: close the cover and verify /api/print/status before retrying jobs."
- alert: PrintJobDeadLetter - alert: PrintJobDeadLetter
expr: increase(print_jobs_dead_letter_total[15m]) > 0 expr: increase(print_jobs_dead_letter_total[15m]) > 0
@@ -869,6 +824,53 @@ data:
summary: "desktop.iamworkin.lan TLS cert expires within 2 days" summary: "desktop.iamworkin.lan TLS cert expires within 2 days"
description: "The desktop.iamworkin.lan cert is inside the 2-day renewal window and cert-manager has not renewed. Check cert-manager logs, step-ca reachability, and pfSense DNS overrides per the ACME DNS-01 gate." description: "The desktop.iamworkin.lan cert is inside the 2-day renewal window and cert-manager has not renewed. Check cert-manager logs, step-ca reachability, and pfSense DNS overrides per the ACME DNS-01 gate."
- alert: LonghornPVCGrowthRapid
expr: |
(
(
(
longhorn_volume_actual_size_bytes
- (longhorn_volume_actual_size_bytes offset 1h)
)
/ clamp_min(longhorn_volume_actual_size_bytes offset 1h, 1)
)
* on(volume) group_left(namespace, persistentvolumeclaim) (
(
label_replace(kube_persistentvolumeclaim_info{storageclass="longhorn"}, "volume", "$1", "volumename", "(.+)")
* on(namespace, persistentvolumeclaim) group_left()
kube_persistentvolumeclaim_labels{label_flowercore_io_managed_by="remotedesktop"}
)
or
label_replace(kube_persistentvolumeclaim_info{namespace="fc-desktop", storageclass="longhorn", persistentvolumeclaim=~"fc-profile-.*|remotedesktop-data"}, "volume", "$1", "volumename", "(.+)")
)
) > 0.20
or
(
(
longhorn_volume_actual_size_bytes
/ on(volume) clamp_min(longhorn_volume_capacity_bytes, 1)
)
* on(volume) group_left(namespace, persistentvolumeclaim) (
(
label_replace(kube_persistentvolumeclaim_info{storageclass="longhorn"}, "volume", "$1", "volumename", "(.+)")
* on(namespace, persistentvolumeclaim) group_left()
kube_persistentvolumeclaim_labels{label_flowercore_io_managed_by="remotedesktop"}
)
or
label_replace(kube_persistentvolumeclaim_info{namespace="fc-desktop", storageclass="longhorn", persistentvolumeclaim=~"fc-profile-.*|remotedesktop-data"}, "volume", "$1", "volumename", "(.+)")
)
) > 0.80
for: 5m
labels:
severity: warning
alert_channel: thermal_print
service: remotedesktop
annotations:
summary: "RemoteDesktop Longhorn PVC {{ $labels.namespace }}/{{ $labels.persistentvolumeclaim }} growing rapidly"
description: "Longhorn volume {{ $labels.volume }} backing RemoteDesktop PVC {{ $labels.namespace }}/{{ $labels.persistentvolumeclaim }} grew more than 20% in 1h or is over 80% capacity. Check for runaway SQLite/user-profile growth; this alert was added after the 2026-05-16 RemoteDesktop web SQLite Error 13 incident."
runbook: "1. kubectl -n {{ $labels.namespace }} describe pvc {{ $labels.persistentvolumeclaim }} 2. Open Longhorn UI volume {{ $labels.volume }} 3. Check RemoteDesktop web/user-volume SQLite files for permission or runaway growth 4. Expand PVC only after confirming the writer is healthy"
todo: "2026-05-19 metric gate: live noc1 Prometheus currently exposes kube_persistentvolumeclaim_info and kube_persistentvolumeclaim_resource_requests_storage_bytes, but not longhorn_volume_actual_size_bytes, longhorn_volume_capacity_bytes, kube_persistentvolumeclaim_labels, or kubelet_volume_stats_used_bytes. Keep the fc-desktop PVC fallback until kube-state-metrics label allowlist exposes flowercore.io/managed-by=remotedesktop."
- name: pi-fleet - name: pi-fleet
rules: rules:
- alert: PiManagerDown - alert: PiManagerDown
@@ -3680,146 +3682,6 @@ data:
relativeTimeRange: {from: 120, to: 0} relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__ datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [600], type: gt}}], refId: C} model: {type: threshold, expression: B, conditions: [{evaluator: {params: [600], type: gt}}], refId: C}
- orgId: 1
name: Print Services
folder: Print Alerts
interval: 1m
rules:
- uid: printer-offline-warning
title: PrinterOfflineWarning
condition: C
for: 2m
noDataState: OK
execErrState: OK
annotations:
summary: "Print.Web printer offline on edge2"
description: "Print.Web reports the NuPrint 210 transport is offline or unreachable. IRC-only by design: do not thermal-print an alert when the thermal printer itself is offline."
runbook: "1. Check edge2 power/network 2. Check USB/CUPS queue 3. Open https://print.iamworkin.lan/admin 4. Do not force thermal routing for offline alerts."
labels:
severity: warning
service: print-web
alert_channel: irc
data:
- refId: A
relativeTimeRange: {from: 120, to: 0}
datasourceUid: prometheus
model: {expr: 'print_printer_state{job="printweb-otel"} == 2', instant: true, refId: A}
- refId: B
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: reduce, expression: A, reducer: last, refId: B}
- refId: C
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [0], type: gt}}], refId: C}
- uid: print-paper-roll-critical
title: PrintPaperRollCritical
condition: C
for: 2m
noDataState: OK
execErrState: OK
annotations:
summary: "Print paper depleted on edge2"
description: "NuPrint 210 reports paper depleted while the printer is still online. Load a new roll, drain the hardware buffer if needed, then replay DeadLetter jobs from /print-log."
runbook: "1. Load a fresh roll 2. Drain the hardware buffer if paper-out happened mid-job 3. Open https://print.iamworkin.lan/print-log 4. Retry DeadLetter jobs after the state clears."
labels:
severity: critical
service: print-web
alert_channel: thermal_print
data:
- refId: A
relativeTimeRange: {from: 120, to: 0}
datasourceUid: prometheus
model: {expr: 'print_printer_state{job="printweb-otel"} == 3 and print_printer_online{job="printweb-otel"} == 1', instant: true, refId: A}
- refId: B
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: reduce, expression: A, reducer: last, refId: B}
- refId: C
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [0], type: gt}}], refId: C}
- uid: printer-jam-warning
title: PrinterJamWarning
condition: C
for: 2m
noDataState: OK
execErrState: OK
annotations:
summary: "Print.Web printer jam on edge2"
description: "Print.Web reports a paper/cutter jam state. IRC-only: clear the jam, drain the hardware buffer if bytes were queued, then retry affected jobs."
runbook: "1. Clear paper/cutter path 2. Drain hardware buffer if CUPS queued bytes 3. Verify /api/print/status 4. Retry affected jobs."
labels:
severity: warning
service: print-web
alert_channel: irc
data:
- refId: A
relativeTimeRange: {from: 120, to: 0}
datasourceUid: prometheus
model: {expr: 'print_printer_state{job="printweb-otel"} == 4', instant: true, refId: A}
- refId: B
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: reduce, expression: A, reducer: last, refId: B}
- refId: C
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [0], type: gt}}], refId: C}
- uid: printer-head-error-critical
title: PrinterHeadErrorCritical
condition: C
for: 2m
noDataState: OK
execErrState: OK
annotations:
summary: "Print.Web printer head error on edge2"
description: "Print.Web reports a thermal head or unrecoverable printer error. Critical routing may enter the thermal digest per existing policy; IRC remains the primary triage stream."
runbook: "1. Let the printer cool if overheated 2. Power-cycle only after checking queued jobs 3. Verify /api/print/status 4. Retry jobs after the state clears."
labels:
severity: critical
service: print-web
alert_channel: thermal_print
data:
- refId: A
relativeTimeRange: {from: 120, to: 0}
datasourceUid: prometheus
model: {expr: 'print_printer_state{job="printweb-otel"} == 5', instant: true, refId: A}
- refId: B
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: reduce, expression: A, reducer: last, refId: B}
- refId: C
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [0], type: gt}}], refId: C}
- uid: printer-cover-open-warning
title: PrinterCoverOpenWarning
condition: C
for: 2m
noDataState: OK
execErrState: OK
annotations:
summary: "Print.Web printer cover open on edge2"
description: "Print.Web reports the printer cover/lid is open. IRC-only: close the cover and verify /api/print/status before retrying jobs."
runbook: "1. Close the printer cover 2. Verify /api/print/status returns online 3. Retry affected jobs only after the state clears."
labels:
severity: warning
service: print-web
alert_channel: irc
data:
- refId: A
relativeTimeRange: {from: 120, to: 0}
datasourceUid: prometheus
model: {expr: 'print_printer_state{job="printweb-otel"} == 6', instant: true, refId: A}
- refId: B
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: reduce, expression: A, reducer: last, refId: B}
- refId: C
relativeTimeRange: {from: 120, to: 0}
datasourceUid: __expr__
model: {type: threshold, expression: B, conditions: [{evaluator: {params: [0], type: gt}}], refId: C}
- orgId: 1 - orgId: 1
name: CI Runners name: CI Runners
folder: CI Alerts folder: CI Alerts

View File

@@ -304,7 +304,7 @@ public sealed class FleetManifestLintTests
} }
[Fact] [Fact]
public void Monitoring_MustIncludeRequiredAlertRoutingGuards() public void Monitoring_MustAlertWhenLinuxRunnerDeploymentIsUnavailable()
{ {
var monitoring = File.ReadAllText(Path.Combine(Inventory.BluejayRoot, "apps", "monitoring", "noc-monitoring.yaml")); var monitoring = File.ReadAllText(Path.Combine(Inventory.BluejayRoot, "apps", "monitoring", "noc-monitoring.yaml"));
@@ -315,15 +315,6 @@ public sealed class FleetManifestLintTests
monitoring.Should().Contain("folder: CI Alerts"); monitoring.Should().Contain("folder: CI Alerts");
monitoring.Should().Contain("uid: linux-runner-offline"); monitoring.Should().Contain("uid: linux-runner-offline");
monitoring.Should().Contain("alert_channel: irc"); monitoring.Should().Contain("alert_channel: irc");
monitoring.Should().Contain("PrinterOfflineWarning");
monitoring.Should().Contain("expr: print_printer_state{job=\"printweb-otel\"} == 2");
monitoring.Should().Contain("IRC-only by design: do not thermal-print an alert when the thermal printer itself is offline.");
monitoring.Should().Contain("PrintPaperRollCritical");
monitoring.Should().Contain("expr: print_printer_state{job=\"printweb-otel\"} == 3 and print_printer_online{job=\"printweb-otel\"} == 1");
monitoring.Should().Contain("PrinterJamWarning");
monitoring.Should().Contain("PrinterHeadErrorCritical");
monitoring.Should().Contain("PrinterCoverOpenWarning");
} }
[Fact] [Fact]