1
0

runtime: publish BACKUP-FAILED-UNITS-SOURCE-217

This commit is contained in:
homelab-runtime-publisher
2026-07-22 15:35:58 +00:00
parent 744480384c
commit c250241459
4 changed files with 610 additions and 45 deletions
File diff suppressed because one or more lines are too long
@@ -0,0 +1,297 @@
CHAT_OUTPUT_BEGIN
COMMAND_ID=BACKUP-FAILED-UNITS-SOURCE-217
STATUS=OK
RC=0
HOST=pve01
MODE=read-only
COMPONENT=cluster-audit
REFERENCE_REGISTER_CHECK=OK
REFERENCE_SHA256=5763f2f8edc75fcf6f3951f8c95896d1cc112ceccb6cf5b64a383f540a4deb8e
ERROR_REGISTER_CHECK=OK
ERROR_REGISTER_SHA256=08b3a769b8e38eaa1acec915352fb328180d70caa7ff6cd9d1ec0da426262752
COMMAND_SHA256=2839e6de5b682a8f42432e3841c7b452be12e60f73cd670b0ac4412372ad4c86
DUPLICATE_FAILED_COMMAND_BLOCKED=false
EXECUTION_STARTED=true
CHANGES_MADE=false
SANITIZED=yes
SECRETS_INCLUDED=no
PRIVATE_ADDRESSES_INCLUDED=no
RAW_EVIDENCE_SHA256=96e72a8cec4b5e5cde52a48e6816f4ce162311b07530ee398c57efc67f3480b1
SANITIZED_OUTPUT_SHA256=dbe6bb8b0419312d9774755ff8668c435118bc102b49b0255667d04f06dda499
OUTPUT_BEGIN
PREVIOUS_COMMAND_ID=RUNTIME-PUBLISH-GAP-DIAG-216
===== UNIT homelab-backup-matrix-refresh =====
# /etc/systemd/system/homelab-backup-matrix-refresh.service
[Unit]
Description=Generate unified homelab backup matrix state
After=local-fs.target
[Service]
Type=oneshot
ExecStart=/usr/local/sbin/homelab-backup-matrix-refresh
TimeoutStartSec=5min
===== UNIT homelab-vm160-remote-backup =====
# /etc/systemd/system/homelab-vm160-remote-backup.service
[Unit]
Description=VM160 full-image backup streamed from pve02 to pve03
After=network-online.target pve-cluster.service
Wants=network-online.target
[Service]
Type=oneshot
TimeoutStartSec=6h
ExecStart=/usr/local/sbin/homelab-vm160-remote-backup
===== UNIT homelab-cluster-admin-pull =====
# /etc/systemd/system/homelab-cluster-admin-pull.service
[Unit]
Description=Pull cluster-admin observer health into pve01
[Service]
Type=oneshot
ExecStart=/usr/local/sbin/homelab-cluster-admin-pull
===== FILE /usr/local/sbin/homelab-backup-matrix-refresh =====
/usr/local/sbin/homelab-backup-matrix-refresh: Python script, ASCII text executable, with very long lines (592)
1 #!/usr/bin/env python3
2 from pathlib import Path
3 import shutil
4 import json,subprocess,os,time,sys,hashlib
5 M=Path("/etc/homelab-backup-matrix/matrix.json")
6 ROOT=Path("/var/lib/homelab-backup-matrix")
7 RENDER=ROOT/"rendered"
8 COMPAT=ROOT/"compat"
9 def cmd(args):
10 p=subprocess.run(args,stdout=subprocess.PIPE,stderr=subprocess.DEVNULL,text=True,check=False)
11 return p.stdout.strip() or "unknown"
12 def atomic(path,text,mode=0o640):
13 path.parent.mkdir(parents=True,exist_ok=True)
14 tmp=path.with_name(path.name+".tmp")
15 tmp.write_text(text,encoding="utf-8")
16 os.chmod(tmp,mode)
17 os.replace(tmp,path)
18 def parse_health(path):
19 if path.suffix.lower()==".json":
20 try:
21 data=json.loads(path.read_text(encoding="utf-8",errors="replace"))
22 return str(data.get("status",data.get("STATUS",data.get("result","UNKNOWN"))))
23 except Exception:
24 return "ERROR"
25 data={}
26 text=path.read_text(encoding="utf-8",errors="replace")
27 for token in text.replace(chr(10)," ").split():
28 if "=" in token:
29 key,value=token.split("=",1)
30 if key and key[0].isalpha():
31 data[key]=value
32 return str(data.get("STATUS",data.get("status","UNKNOWN")))
33 def normalize(value):
34 value=str(value).strip().upper()
35 if value in {"OK","PASS","PASSED","SUCCESS","HEALTHY"}:
36 return "OK"
37 if value in {"WARN","WARNING"}:
38 return "WARN"
39 if value in {"FAIL","FAILED","ERROR","CRITICAL","BAD"}:
40 return "FAIL"
41 return value or "UNKNOWN"
42 matrix=json.loads(M.read_text(encoding="utf-8"))
43 ROOT.mkdir(parents=True,exist_ok=True)
44 RENDER.mkdir(parents=True,exist_ok=True)
45 COMPAT.mkdir(parents=True,exist_ok=True)
46 now=int(time.time())
47 now_iso=time.strftime("%Y-%m-%dT%H:%M:%SZ",time.gmtime(now))
48 results=[]
49 compat_map=[]
50 for job in matrix.get("jobs",[]):
51 original_sources=list(job.get("evidence_sources_original",[]))
52 evidence_inputs=list(job.get("evidence_inputs",[]))
53 if original_sources and len(original_sources)!=len(evidence_inputs):
54 raise RuntimeError("evidence_mapping_length_mismatch:"+str(job.get("name","")))
55 for source_raw,destination_raw in zip(original_sources,evidence_inputs):
56 source_path=Path(source_raw)
57 destination_path=Path(destination_raw)
58 if not source_path.is_file():
59 continue
60 head=source_path.read_text(encoding="utf-8",errors="ignore")[:4096]
61 if "TYPE=backup-matrix-generated" in head or "CANONICAL_STATE=/var/lib/homelab-backup-matrix/state.json" in head:
62 continue
63 destination_path.parent.mkdir(parents=True,exist_ok=True)
64 temporary=destination_path.with_name(destination_path.name+".tmp-ingest")
65 shutil.copy2(source_path,temporary)
66 os.replace(temporary,destination_path)
67 jid=str(job.get("id",""))
68 name=str(job.get("name",jid))
69 timer=str(job.get("timer",""))
70 service=str(job.get("service",""))
71 enabled=cmd(["systemctl","is-enabled",timer]) if timer else "missing"
72 active=cmd(["systemctl","is-active",timer]) if timer else "missing"
73 source_kind=str(job.get("source_kind","health_file"))
74 best=None
75 for raw in job.get("evidence_inputs",job.get("legacy_health_inputs",[])):
76 path=Path(raw)
77 if not path.is_file():
78 continue
79 mtime=int(path.stat().st_mtime)
80 if best is None or mtime>best[0]:
81 best=(mtime,str(path),parse_health(path))
82 service_result=cmd(["systemctl","show",service,"-p","Result","--value"]) if service else "missing"
83 service_exec_main_status=cmd(["systemctl","show",service,"-p","ExecMainStatus","--value"]) if service else "missing"
84 evidence_level="artifact_or_health"
85 if best is None and source_kind=="systemd":
86 stamp=Path("/var/lib/systemd/timers")/("stamp-"+timer)
87 mtime=int(stamp.stat().st_mtime) if stamp.is_file() else 0
88 source="systemd:"+service
89 source_status="OK" if service_result=="success" and service_exec_main_status=="0" else "FAIL"
90 evidence_level="execution_only"
91 else:
92 source=best[1] if best else ""
93 source_status=normalize(best[2]) if best else "UNMAPPED"
94 mtime=best[0] if best else 0
95 age=now-mtime if mtime else -1
96 max_age=int(job.get("max_age_seconds",172800))
97 critical=bool(job.get("critical",False))
98 if not critical and enabled!="enabled":
99 final="DISABLED"
100 elif critical and (enabled!="enabled" or active!="active"):
101 final="FAIL"
102 elif source_status=="UNMAPPED":
103 final="UNMAPPED"
104 elif source_status!="OK":
105 final="FAIL"
106 elif age<0:
107 final="MISSING"
108 elif age>max_age:
109 final="STALE"
110 else:
111 final="OK"
112 freshness="UNKNOWN" if age<0 else ("FRESH" if age<=max_age else "STALE")
113 source_mtime=time.strftime("%Y-%m-%dT%H:%M:%SZ",time.gmtime(mtime)) if mtime else ""
114 result={"id":jid,"name":name,"category":job.get("category","unknown"),"critical":critical,"timer":timer,"service":service,"timer_enabled":enabled,"timer_active":active,"status":final,"source_kind":source_kind,"evidence_level":evidence_level,"service_result":service_result,"service_exec_main_status":service_exec_main_status,"source_health":source,"source_status":source_status,"source_mtime":source_mtime,"age_seconds":age,"max_age_seconds":max_age,"freshness":freshness,"generated_health_output":job.get("generated_health_output","")}
115 results.append(result)
116 mode_value=matrix.get("mode","shadow"); lines=[f"STATUS={final}",f"TS={now_iso}","TYPE=backup-matrix-generated",f"JOB_ID={jid}",f"JOB_NAME={name}",f"MATRIX_MODE={mode_value}",f"TIMER={timer}",f"SERVICE={service}",f"TIMER_ENABLED={enabled}",f"TIMER_ACTIVE={active}",f"SOURCE_KIND={source_kind}",f"EVIDENCE_LEVEL={evidence_level}",f"SERVICE_RESULT={service_result}",f"SERVICE_EXEC_MAIN_STATUS={service_exec_main_status}",f"SOURCE_HEALTH={source}",f"SOURCE_STATUS={source_status}",f"SOURCE_MTIME_UTC={source_mtime}",f"AGE_SECONDS={age}",f"MAX_AGE_SECONDS={max_age}",f"FRESHNESS={freshness}"]
117 atomic(Path(job.get("generated_health_output",str(RENDER/(name+".txt")))),chr(10).join(lines)+chr(10))
118 for raw in job.get("compatibility_outputs",[]):
119 legacy=str(raw)
120 digest=hashlib.sha256(legacy.encode("utf-8")).hexdigest()[:12]
121 base_name=Path(legacy).name or "health"
122 safe=f"{jid}--{base_name}--{digest}.txt"
123 compat_path=COMPAT/safe
124 compat_lines=lines+[f"LEGACY_SOURCE_PATH={legacy}",f"CANONICAL_JOB_ID={jid}",f"CANONICAL_STATE={ROOT}/state.json"]
125 atomic(compat_path,chr(10).join(compat_lines)+chr(10))
126 atomic(Path(legacy),chr(10).join(compat_lines)+chr(10))
127 compat_map.append({"job_id":jid,"legacy_path":legacy,"compatibility_path":str(compat_path)})
128 bad=sum(1 for item in results if item["status"] in {"FAIL","STALE","MISSING"})
129 unmapped=sum(1 for item in results if item["status"]=="UNMAPPED")
130 disabled=sum(1 for item in results if item["status"]=="DISABLED")
131 ok=sum(1 for item in results if item["status"]=="OK")
132 mode=str(matrix.get("mode","shadow"))
133 overall="OK" if bad==0 and unmapped==0 else ("WARN" if mode=="shadow" else "FAIL")
134 compat_map=sorted(compat_map,key=lambda item:(item["legacy_path"],item["job_id"]))
135 atomic(ROOT/"compat-map.json",json.dumps(compat_map,ensure_ascii=False,indent=2,sort_keys=True)+chr(10))
136 state={"version":1,"generated_at":now_iso,"matrix_mode":mode,"matrix_sha256":hashlib.sha256(M.read_bytes()).hexdigest(),"summary":{"status":overall,"jobs_total":len(results),"jobs_ok":ok,"jobs_bad":bad,"jobs_unmapped":unmapped,"jobs_disabled":disabled},"jobs":results}
137 state["authority"]="canonical"
138 state["compatibility_map"]=str(ROOT/"compat-map.json")
139 state["summary"]["compatibility_outputs"]=len(compat_map)
140 atomic(ROOT/"state.json",json.dumps(state,ensure_ascii=False,indent=2,sort_keys=True)+chr(10))
===== FILE /usr/local/sbin/homelab-vm160-remote-backup =====
/usr/local/sbin/homelab-vm160-remote-backup: Bourne-Again shell script, ASCII text executable
1 #!/usr/bin/env bash
2 set -Eeuo pipefail
3 export SYSTEMD_PAGER=cat PAGER=cat SYSTEMD_LESS=FRXMK
4 VMID=160
5 SOURCE=root@[PRIVATE_IP]
6 TARGET=root@[PRIVATE_IP]
7 KH=[SENSITIVE_PATH]
8 PARENT=/mnt/staging/vzdump
9 PREFIX=vm160-pve02-
10 HEALTH=/var/lib/homelab-health/vm160-remote-backup.txt
11 LOCK=/run/lock/homelab-vm160-remote-backup.lock
12 SSH=(-o BatchMode=yes -o ConnectTimeout=12 -o StrictHostKeyChecking=yes -o UserKnownHostsFile="$KH")
13 TARGET_WORKER=/usr/local/libexec/homelab-vm160-target-worker
14 target_worker(){ ssh "${SSH[@]}" "$TARGET" bash -s -- "$@" < "$TARGET_WORKER"; }
15 install -d -m 0755 "$(dirname "$HEALTH")" "$(dirname "$LOCK")"
16 exec 9>"$LOCK"
17 flock -n 9
18 test "$(hostname -s)" = pve01
19 test -x "$TARGET_WORKER"
20 test "$(ssh "${SSH[@]}" "$SOURCE" "qm status $VMID" | awk '{print $2}')" = running
21 SOURCE_ACTIVE_RC=0
22 SOURCE_ACTIVE="$(ssh "${SSH[@]}" "$SOURCE" "pgrep -af '(^|/)(vzdump|qmrestore|pctrestore|vma|zstd)([[:space:]]|$)'")" || SOURCE_ACTIVE_RC=$?
23 test "$SOURCE_ACTIVE_RC" -eq 0 || test "$SOURCE_ACTIVE_RC" -eq 1
24 test -z "$SOURCE_ACTIVE"
25 TS="$(date -u +%Y%m%dT%H%M%SZ)"
26 DIR="$PARENT/${PREFIX}${TS}"
27 NAME="vzdump-qemu-${VMID}-${TS}.vma.zst"
28 PART="$DIR/.${NAME}.partial"
29 FINAL="$DIR/$NAME"
30 LOG="$(mktemp /tmp/vm160-remote-backup.XXXXXX.log)"
31 cleanup(){
32 rc=$?
33 trap - EXIT
34 if test "$rc" -ne 0; then
35 ssh "${SSH[@]}" "$TARGET" "rm -rf -- '$DIR'" >/dev/null 2>&1 || true
36 tmp="${HEALTH}.tmp"
37 printf 'STATUS=FAIL\nSERVICE=homelab-vm160-remote-backup\nTIMESTAMP_UTC=%s\nVMID=%s\nTARGET_DIR=%s\nRC=%s\n' "$TS" "$VMID" "$DIR" "$rc" > "$tmp"
38 chmod 0644 "$tmp"
39 mv "$tmp" "$HEALTH"
40 fi
41 rm -f "$LOG"
42 exit "$rc"
43 }
44 trap cleanup EXIT
45 REMOTE_FREE="$(ssh "${SSH[@]}" "$TARGET" "df -B1 --output=avail '$PARENT' | awk 'NR==2{print \$1}'")"
46 LATEST_SIZE="$(ssh "${SSH[@]}" "$TARGET" "find '$PARENT' -mindepth 2 -maxdepth 2 -type f -path '$PARENT/${PREFIX}*/vzdump-qemu-${VMID}-*.vma.zst' -printf '%s\n' | sort -nr | head -n1")"
47 LATEST_SIZE="${LATEST_SIZE:-0}"
48 if test "$LATEST_SIZE" -gt 0; then
49 REQUIRED="$((LATEST_SIZE + LATEST_SIZE / 4 + 10737418240))"
50 else
51 REQUIRED=161061273600
52 fi
53 test "$REMOTE_FREE" -ge "$REQUIRED"
54 ssh "${SSH[@]}" "$TARGET" "install -d -m 0700 '$DIR'"
55 ssh "${SSH[@]}" "$SOURCE" "vzdump $VMID --stdout 1 --mode snapshot --compress zstd --ionice 7" 2>"$LOG" | ssh "${SSH[@]}" "$TARGET" "cat > '$PART'"
56 target_worker finalize "$PARENT" "$PREFIX" "$PART" "$FINAL" "$DIR" "$NAME"
57 ssh "${SSH[@]}" "$TARGET" "cat > '$DIR/vzdump.log'" < "$LOG"
58 SIZE="$(ssh "${SSH[@]}" "$TARGET" "stat -c %s '$FINAL'")"
59 SHA="$(ssh "${SSH[@]}" "$TARGET" "sha256sum '$FINAL' | awk '{print \$1}'")"
60 target_worker promote "$PARENT" "$PREFIX" "$DIR" "$TS"
61 tmp="${HEALTH}.tmp"
62 printf 'STATUS=OK\nSERVICE=homelab-vm160-remote-backup\nTIMESTAMP_UTC=%s\nVMID=%s\nSOURCE_NODE=pve02\nTARGET_NODE=pve03\nTARGET_DIR=%s\nARCHIVE=%s\nSIZE_BYTES=%s\nSHA256=%s\nREMOTE_RETENTION_GENERATIONS=1\n' "$TS" "$VMID" "$DIR" "$FINAL" "$SIZE" "$SHA" > "$tmp"
63 chmod 0644 "$tmp"
64 mv "$tmp" "$HEALTH"
65 trap - EXIT
66 rm -f "$LOG"
67 printf 'VM160_REMOTE_BACKUP_STATUS=OK\nTARGET_DIR=%s\nARCHIVE=%s\nSIZE_BYTES=%s\nSHA256=%s\n' "$DIR" "$FINAL" "$SIZE" "$SHA"
===== FILE /usr/local/sbin/homelab-cluster-admin-pull =====
/usr/local/sbin/homelab-cluster-admin-pull: Bourne-Again shell script, ASCII text executable, with very long lines (424)
1 #!/usr/bin/env bash
2 set +e +u +o pipefail 2>/dev/null || true
3 TS=$(date -u +%Y-%m-%dT%H:%M:%SZ)
4 ROOT=/var/lib/homelab-cluster-admin
5 HEALTH_ROOT=/var/lib/homelab-health
6 mkdir -p "$ROOT" "$HEALTH_ROOT"
7 REMOTE=$(ssh -n -o BatchMode=yes -o ConnectTimeout=8 debian@[PRIVATE_IP] 'sudo /opt/homelab-cluster-admin/cluster-observer.sh >/dev/null 2>&1 || true; sudo /opt/homelab-cluster-admin/full-observer.sh >/dev/null 2>&1 || true; printf "__OBS__\\n"; cat /var/lib/homelab-health/cluster-admin-observer.txt; printf "__FULL__\\n"; cat /var/lib/homelab-health/cluster-admin-full-observer.txt' 2>/tmp/homelab-cluster-admin-pull.err)
8 RC=$?
9 OBS=$(printf "%s\n" "$REMOTE" | awk '/^__OBS__$/{f=1;next}/^__FULL__$/{f=0}f')
10 FULL=$(printf "%s\n" "$REMOTE" | awk '/^__FULL__$/{f=1;next}f')
11 OBS_BAD=0; FULL_BAD=0
12 printf "%s\n" "$OBS" | grep -q "^STATUS=OK" || OBS_BAD=1
13 printf "%s\n" "$FULL" | grep -q "^STATUS=OK" || FULL_BAD=1
14 printf "%s\n" "$FULL" | grep -q "BAD=0" || FULL_BAD=1
15 PROBE_OK=$(printf "%s\n" "$FULL" | sed -n 's/.*PROBE_OK=\([0-9][0-9]*\).*/\1/p' | head -n1)
16 PROBE_BAD=$(printf "%s\n" "$FULL" | sed -n 's/.*PROBE_BAD=\([0-9][0-9]*\).*/\1/p' | head -n1)
17 [ "${PROBE_BAD:-1}" = 0 ] || FULL_BAD=1
18 BAD=$((OBS_BAD+FULL_BAD)); [ "$RC" -eq 0 ] || BAD=$((BAD+1))
19 printf "%s\n" "$OBS" > "$ROOT/observer-remote-latest.txt"
20 printf "%s\n" "$FULL" > "$ROOT/full-observer-central-latest.md"
21 OBS_STATUS=$([ "$OBS_BAD" -eq 0 ] && [ "$RC" -eq 0 ] && echo OK || echo WARN)
22 FULL_STATUS=$([ "$FULL_BAD" -eq 0 ] && [ "$RC" -eq 0 ] && echo OK || echo WARN)
23 echo "STATUS=$OBS_STATUS TS=$TS TYPE=cluster-admin-central-pull BAD=$OBS_BAD VMID=180 IP=[PRIVATE_IP] SSH_RC=$RC OBSERVER_STATUS=$OBS_STATUS REPORT=$ROOT/observer-remote-latest.txt" > "$HEALTH_ROOT/cluster-admin-observer.txt"
24 echo "STATUS=$FULL_STATUS TS=$TS TYPE=cluster-admin-full-observer-central BAD=$FULL_BAD WARN=0 VMID=180 PULL_RC=$RC FULL_STATUS=$FULL_STATUS PROBE_OK=${PROBE_OK:-0} PROBE_BAD=${PROBE_BAD:-1} REPORT=$ROOT/full-observer-central-latest.md" > "$HEALTH_ROOT/cluster-admin-full-observer.txt"
25 echo "STATUS=$FULL_STATUS TS=$TS TYPE=cluster-admin-restricted-probes BAD=$FULL_BAD VMID=180 MODE=FORCED_COMMAND_READONLY_CURRENT PROBE_OK=${PROBE_OK:-0} PROBE_BAD=${PROBE_BAD:-1} TARGETS=pve01,pve02,pve03,edge SOURCE=full-observer" > "$HEALTH_ROOT/cluster-admin-restricted-probes.txt"
26 cat "$HEALTH_ROOT/cluster-admin-observer.txt"
27 cat "$HEALTH_ROOT/cluster-admin-full-observer.txt"
28 cat "$HEALTH_ROOT/cluster-admin-restricted-probes.txt"
29 [ "$BAD" -eq 0 ]
CHANGES_MADE=NO
OUTPUT_END
CHAT_OUTPUT_END
+7 -7
View File
File diff suppressed because one or more lines are too long
+276 -38
View File
@@ -1,5 +1,5 @@
CHAT_OUTPUT_BEGIN
COMMAND_ID=RUNTIME-PUBLISH-GAP-DIAG-216
COMMAND_ID=BACKUP-FAILED-UNITS-SOURCE-217
STATUS=OK
RC=0
HOST=pve01
@@ -9,50 +9,288 @@ REFERENCE_REGISTER_CHECK=OK
REFERENCE_SHA256=5763f2f8edc75fcf6f3951f8c95896d1cc112ceccb6cf5b64a383f540a4deb8e
ERROR_REGISTER_CHECK=OK
ERROR_REGISTER_SHA256=08b3a769b8e38eaa1acec915352fb328180d70caa7ff6cd9d1ec0da426262752
COMMAND_SHA256=7e59c8be0d1806c75e6a70e6272d23228e685ed0f7ce29ff23655e5cb6cab201
COMMAND_SHA256=2839e6de5b682a8f42432e3841c7b452be12e60f73cd670b0ac4412372ad4c86
DUPLICATE_FAILED_COMMAND_BLOCKED=false
EXECUTION_STARTED=true
CHANGES_MADE=false
SANITIZED=yes
SECRETS_INCLUDED=no
PRIVATE_ADDRESSES_INCLUDED=no
RAW_EVIDENCE_SHA256=d960ffb487a4fd038b366d566f09939731854b6a723f7c2255698cd5d5b758f0
SANITIZED_OUTPUT_SHA256=d960ffb487a4fd038b366d566f09939731854b6a723f7c2255698cd5d5b758f0
RAW_EVIDENCE_SHA256=96e72a8cec4b5e5cde52a48e6816f4ce162311b07530ee398c57efc67f3480b1
SANITIZED_OUTPUT_SHA256=dbe6bb8b0419312d9774755ff8668c435118bc102b49b0255667d04f06dda499
OUTPUT_BEGIN
PREVIOUS_CONFIRMED=BACKUP-OFFHOST-DELTA-212
=== LEDGER 213-215 ===
2026-07-22T15:16:25Z BACKUP-FAILED-UNITS-ROOTCAUSE-213 fb59f580fefab96bf42d4c9f720847f498158079695bebebca527795287c0435 2 FAIL backup-audit
2026-07-22T15:19:41Z BACKUP-FAILED-UNITS-ROOTCAUSE-214 19d600f59a44903ae388de5924789d15c9dda0ede9ed5f3966600f8f808c43c1 0 OK backup-audit
=== LOCAL FILES 213-215 ===
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z.json
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z-latest.txt
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z.txt
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z.json
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z-latest.txt
/var/lib/homelab-chat-run/public/BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z.txt
/var/lib/homelab-chat-run/raw/BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z.log
/var/lib/homelab-chat-run/raw/BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z.log
=== RECENT PUBLIC RESULTS ===
1784733581.6637746080 BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z-latest.txt
1784733581.6615274220 BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z.json
1784733581.6428190850 BACKUP-FAILED-UNITS-ROOTCAUSE-214-20260722T151941Z.txt
1784733385.6042597560 BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z-latest.txt
1784733385.6012319060 BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z.json
1784733385.5822310630 BACKUP-FAILED-UNITS-ROOTCAUSE-213-20260722T151625Z.txt
1784732405.9184316600 BACKUP-OFFHOST-DELTA-212-20260722T150005Z-latest.txt
1784732405.9148605780 BACKUP-OFFHOST-DELTA-212-20260722T150005Z.json
1784732405.8929104080 BACKUP-OFFHOST-DELTA-212-20260722T150005Z.txt
1784732035.7581781390 BACKUP-OFFHOST-CURRENT-STATE-212-20260722T145351Z-latest.txt
1784732035.7560127120 BACKUP-OFFHOST-CURRENT-STATE-212-20260722T145351Z.json
1784732035.7377798780 BACKUP-OFFHOST-CURRENT-STATE-212-20260722T145351Z.txt
1784730859.9011235460 HOMELAB-TAILS-CLOSEOUT-211-20260722T143419Z-latest.txt
1784730859.8981025540 HOMELAB-TAILS-CLOSEOUT-211-20260722T143419Z.json
1784730859.8809036410 HOMELAB-TAILS-CLOSEOUT-211-20260722T143419Z.txt
1784730606.0033666560 HOMELAB-TAILS-CLOSEOUT-210-20260722T143005Z-latest.txt
1784730606.0015773800 HOMELAB-TAILS-CLOSEOUT-210-20260722T143005Z.json
1784730605.9797819180 HOMELAB-TAILS-CLOSEOUT-210-20260722T143005Z.txt
1784730456.7690908780 HOMELAB-TAILS-CLOSEOUT-209-20260722T142736Z-latest.txt
1784730456.7669407010 HOMELAB-TAILS-CLOSEOUT-209-20260722T142736Z.json
PREVIOUS_COMMAND_ID=RUNTIME-PUBLISH-GAP-DIAG-216
===== UNIT homelab-backup-matrix-refresh =====
# /etc/systemd/system/homelab-backup-matrix-refresh.service
[Unit]
Description=Generate unified homelab backup matrix state
After=local-fs.target
[Service]
Type=oneshot
ExecStart=/usr/local/sbin/homelab-backup-matrix-refresh
TimeoutStartSec=5min
===== UNIT homelab-vm160-remote-backup =====
# /etc/systemd/system/homelab-vm160-remote-backup.service
[Unit]
Description=VM160 full-image backup streamed from pve02 to pve03
After=network-online.target pve-cluster.service
Wants=network-online.target
[Service]
Type=oneshot
TimeoutStartSec=6h
ExecStart=/usr/local/sbin/homelab-vm160-remote-backup
===== UNIT homelab-cluster-admin-pull =====
# /etc/systemd/system/homelab-cluster-admin-pull.service
[Unit]
Description=Pull cluster-admin observer health into pve01
[Service]
Type=oneshot
ExecStart=/usr/local/sbin/homelab-cluster-admin-pull
===== FILE /usr/local/sbin/homelab-backup-matrix-refresh =====
/usr/local/sbin/homelab-backup-matrix-refresh: Python script, ASCII text executable, with very long lines (592)
1 #!/usr/bin/env python3
2 from pathlib import Path
3 import shutil
4 import json,subprocess,os,time,sys,hashlib
5 M=Path("/etc/homelab-backup-matrix/matrix.json")
6 ROOT=Path("/var/lib/homelab-backup-matrix")
7 RENDER=ROOT/"rendered"
8 COMPAT=ROOT/"compat"
9 def cmd(args):
10 p=subprocess.run(args,stdout=subprocess.PIPE,stderr=subprocess.DEVNULL,text=True,check=False)
11 return p.stdout.strip() or "unknown"
12 def atomic(path,text,mode=0o640):
13 path.parent.mkdir(parents=True,exist_ok=True)
14 tmp=path.with_name(path.name+".tmp")
15 tmp.write_text(text,encoding="utf-8")
16 os.chmod(tmp,mode)
17 os.replace(tmp,path)
18 def parse_health(path):
19 if path.suffix.lower()==".json":
20 try:
21 data=json.loads(path.read_text(encoding="utf-8",errors="replace"))
22 return str(data.get("status",data.get("STATUS",data.get("result","UNKNOWN"))))
23 except Exception:
24 return "ERROR"
25 data={}
26 text=path.read_text(encoding="utf-8",errors="replace")
27 for token in text.replace(chr(10)," ").split():
28 if "=" in token:
29 key,value=token.split("=",1)
30 if key and key[0].isalpha():
31 data[key]=value
32 return str(data.get("STATUS",data.get("status","UNKNOWN")))
33 def normalize(value):
34 value=str(value).strip().upper()
35 if value in {"OK","PASS","PASSED","SUCCESS","HEALTHY"}:
36 return "OK"
37 if value in {"WARN","WARNING"}:
38 return "WARN"
39 if value in {"FAIL","FAILED","ERROR","CRITICAL","BAD"}:
40 return "FAIL"
41 return value or "UNKNOWN"
42 matrix=json.loads(M.read_text(encoding="utf-8"))
43 ROOT.mkdir(parents=True,exist_ok=True)
44 RENDER.mkdir(parents=True,exist_ok=True)
45 COMPAT.mkdir(parents=True,exist_ok=True)
46 now=int(time.time())
47 now_iso=time.strftime("%Y-%m-%dT%H:%M:%SZ",time.gmtime(now))
48 results=[]
49 compat_map=[]
50 for job in matrix.get("jobs",[]):
51 original_sources=list(job.get("evidence_sources_original",[]))
52 evidence_inputs=list(job.get("evidence_inputs",[]))
53 if original_sources and len(original_sources)!=len(evidence_inputs):
54 raise RuntimeError("evidence_mapping_length_mismatch:"+str(job.get("name","")))
55 for source_raw,destination_raw in zip(original_sources,evidence_inputs):
56 source_path=Path(source_raw)
57 destination_path=Path(destination_raw)
58 if not source_path.is_file():
59 continue
60 head=source_path.read_text(encoding="utf-8",errors="ignore")[:4096]
61 if "TYPE=backup-matrix-generated" in head or "CANONICAL_STATE=/var/lib/homelab-backup-matrix/state.json" in head:
62 continue
63 destination_path.parent.mkdir(parents=True,exist_ok=True)
64 temporary=destination_path.with_name(destination_path.name+".tmp-ingest")
65 shutil.copy2(source_path,temporary)
66 os.replace(temporary,destination_path)
67 jid=str(job.get("id",""))
68 name=str(job.get("name",jid))
69 timer=str(job.get("timer",""))
70 service=str(job.get("service",""))
71 enabled=cmd(["systemctl","is-enabled",timer]) if timer else "missing"
72 active=cmd(["systemctl","is-active",timer]) if timer else "missing"
73 source_kind=str(job.get("source_kind","health_file"))
74 best=None
75 for raw in job.get("evidence_inputs",job.get("legacy_health_inputs",[])):
76 path=Path(raw)
77 if not path.is_file():
78 continue
79 mtime=int(path.stat().st_mtime)
80 if best is None or mtime>best[0]:
81 best=(mtime,str(path),parse_health(path))
82 service_result=cmd(["systemctl","show",service,"-p","Result","--value"]) if service else "missing"
83 service_exec_main_status=cmd(["systemctl","show",service,"-p","ExecMainStatus","--value"]) if service else "missing"
84 evidence_level="artifact_or_health"
85 if best is None and source_kind=="systemd":
86 stamp=Path("/var/lib/systemd/timers")/("stamp-"+timer)
87 mtime=int(stamp.stat().st_mtime) if stamp.is_file() else 0
88 source="systemd:"+service
89 source_status="OK" if service_result=="success" and service_exec_main_status=="0" else "FAIL"
90 evidence_level="execution_only"
91 else:
92 source=best[1] if best else ""
93 source_status=normalize(best[2]) if best else "UNMAPPED"
94 mtime=best[0] if best else 0
95 age=now-mtime if mtime else -1
96 max_age=int(job.get("max_age_seconds",172800))
97 critical=bool(job.get("critical",False))
98 if not critical and enabled!="enabled":
99 final="DISABLED"
100 elif critical and (enabled!="enabled" or active!="active"):
101 final="FAIL"
102 elif source_status=="UNMAPPED":
103 final="UNMAPPED"
104 elif source_status!="OK":
105 final="FAIL"
106 elif age<0:
107 final="MISSING"
108 elif age>max_age:
109 final="STALE"
110 else:
111 final="OK"
112 freshness="UNKNOWN" if age<0 else ("FRESH" if age<=max_age else "STALE")
113 source_mtime=time.strftime("%Y-%m-%dT%H:%M:%SZ",time.gmtime(mtime)) if mtime else ""
114 result={"id":jid,"name":name,"category":job.get("category","unknown"),"critical":critical,"timer":timer,"service":service,"timer_enabled":enabled,"timer_active":active,"status":final,"source_kind":source_kind,"evidence_level":evidence_level,"service_result":service_result,"service_exec_main_status":service_exec_main_status,"source_health":source,"source_status":source_status,"source_mtime":source_mtime,"age_seconds":age,"max_age_seconds":max_age,"freshness":freshness,"generated_health_output":job.get("generated_health_output","")}
115 results.append(result)
116 mode_value=matrix.get("mode","shadow"); lines=[f"STATUS={final}",f"TS={now_iso}","TYPE=backup-matrix-generated",f"JOB_ID={jid}",f"JOB_NAME={name}",f"MATRIX_MODE={mode_value}",f"TIMER={timer}",f"SERVICE={service}",f"TIMER_ENABLED={enabled}",f"TIMER_ACTIVE={active}",f"SOURCE_KIND={source_kind}",f"EVIDENCE_LEVEL={evidence_level}",f"SERVICE_RESULT={service_result}",f"SERVICE_EXEC_MAIN_STATUS={service_exec_main_status}",f"SOURCE_HEALTH={source}",f"SOURCE_STATUS={source_status}",f"SOURCE_MTIME_UTC={source_mtime}",f"AGE_SECONDS={age}",f"MAX_AGE_SECONDS={max_age}",f"FRESHNESS={freshness}"]
117 atomic(Path(job.get("generated_health_output",str(RENDER/(name+".txt")))),chr(10).join(lines)+chr(10))
118 for raw in job.get("compatibility_outputs",[]):
119 legacy=str(raw)
120 digest=hashlib.sha256(legacy.encode("utf-8")).hexdigest()[:12]
121 base_name=Path(legacy).name or "health"
122 safe=f"{jid}--{base_name}--{digest}.txt"
123 compat_path=COMPAT/safe
124 compat_lines=lines+[f"LEGACY_SOURCE_PATH={legacy}",f"CANONICAL_JOB_ID={jid}",f"CANONICAL_STATE={ROOT}/state.json"]
125 atomic(compat_path,chr(10).join(compat_lines)+chr(10))
126 atomic(Path(legacy),chr(10).join(compat_lines)+chr(10))
127 compat_map.append({"job_id":jid,"legacy_path":legacy,"compatibility_path":str(compat_path)})
128 bad=sum(1 for item in results if item["status"] in {"FAIL","STALE","MISSING"})
129 unmapped=sum(1 for item in results if item["status"]=="UNMAPPED")
130 disabled=sum(1 for item in results if item["status"]=="DISABLED")
131 ok=sum(1 for item in results if item["status"]=="OK")
132 mode=str(matrix.get("mode","shadow"))
133 overall="OK" if bad==0 and unmapped==0 else ("WARN" if mode=="shadow" else "FAIL")
134 compat_map=sorted(compat_map,key=lambda item:(item["legacy_path"],item["job_id"]))
135 atomic(ROOT/"compat-map.json",json.dumps(compat_map,ensure_ascii=False,indent=2,sort_keys=True)+chr(10))
136 state={"version":1,"generated_at":now_iso,"matrix_mode":mode,"matrix_sha256":hashlib.sha256(M.read_bytes()).hexdigest(),"summary":{"status":overall,"jobs_total":len(results),"jobs_ok":ok,"jobs_bad":bad,"jobs_unmapped":unmapped,"jobs_disabled":disabled},"jobs":results}
137 state["authority"]="canonical"
138 state["compatibility_map"]=str(ROOT/"compat-map.json")
139 state["summary"]["compatibility_outputs"]=len(compat_map)
140 atomic(ROOT/"state.json",json.dumps(state,ensure_ascii=False,indent=2,sort_keys=True)+chr(10))
===== FILE /usr/local/sbin/homelab-vm160-remote-backup =====
/usr/local/sbin/homelab-vm160-remote-backup: Bourne-Again shell script, ASCII text executable
1 #!/usr/bin/env bash
2 set -Eeuo pipefail
3 export SYSTEMD_PAGER=cat PAGER=cat SYSTEMD_LESS=FRXMK
4 VMID=160
5 SOURCE=root@[PRIVATE_IP]
6 TARGET=root@[PRIVATE_IP]
7 KH=[SENSITIVE_PATH]
8 PARENT=/mnt/staging/vzdump
9 PREFIX=vm160-pve02-
10 HEALTH=/var/lib/homelab-health/vm160-remote-backup.txt
11 LOCK=/run/lock/homelab-vm160-remote-backup.lock
12 SSH=(-o BatchMode=yes -o ConnectTimeout=12 -o StrictHostKeyChecking=yes -o UserKnownHostsFile="$KH")
13 TARGET_WORKER=/usr/local/libexec/homelab-vm160-target-worker
14 target_worker(){ ssh "${SSH[@]}" "$TARGET" bash -s -- "$@" < "$TARGET_WORKER"; }
15 install -d -m 0755 "$(dirname "$HEALTH")" "$(dirname "$LOCK")"
16 exec 9>"$LOCK"
17 flock -n 9
18 test "$(hostname -s)" = pve01
19 test -x "$TARGET_WORKER"
20 test "$(ssh "${SSH[@]}" "$SOURCE" "qm status $VMID" | awk '{print $2}')" = running
21 SOURCE_ACTIVE_RC=0
22 SOURCE_ACTIVE="$(ssh "${SSH[@]}" "$SOURCE" "pgrep -af '(^|/)(vzdump|qmrestore|pctrestore|vma|zstd)([[:space:]]|$)'")" || SOURCE_ACTIVE_RC=$?
23 test "$SOURCE_ACTIVE_RC" -eq 0 || test "$SOURCE_ACTIVE_RC" -eq 1
24 test -z "$SOURCE_ACTIVE"
25 TS="$(date -u +%Y%m%dT%H%M%SZ)"
26 DIR="$PARENT/${PREFIX}${TS}"
27 NAME="vzdump-qemu-${VMID}-${TS}.vma.zst"
28 PART="$DIR/.${NAME}.partial"
29 FINAL="$DIR/$NAME"
30 LOG="$(mktemp /tmp/vm160-remote-backup.XXXXXX.log)"
31 cleanup(){
32 rc=$?
33 trap - EXIT
34 if test "$rc" -ne 0; then
35 ssh "${SSH[@]}" "$TARGET" "rm -rf -- '$DIR'" >/dev/null 2>&1 || true
36 tmp="${HEALTH}.tmp"
37 printf 'STATUS=FAIL\nSERVICE=homelab-vm160-remote-backup\nTIMESTAMP_UTC=%s\nVMID=%s\nTARGET_DIR=%s\nRC=%s\n' "$TS" "$VMID" "$DIR" "$rc" > "$tmp"
38 chmod 0644 "$tmp"
39 mv "$tmp" "$HEALTH"
40 fi
41 rm -f "$LOG"
42 exit "$rc"
43 }
44 trap cleanup EXIT
45 REMOTE_FREE="$(ssh "${SSH[@]}" "$TARGET" "df -B1 --output=avail '$PARENT' | awk 'NR==2{print \$1}'")"
46 LATEST_SIZE="$(ssh "${SSH[@]}" "$TARGET" "find '$PARENT' -mindepth 2 -maxdepth 2 -type f -path '$PARENT/${PREFIX}*/vzdump-qemu-${VMID}-*.vma.zst' -printf '%s\n' | sort -nr | head -n1")"
47 LATEST_SIZE="${LATEST_SIZE:-0}"
48 if test "$LATEST_SIZE" -gt 0; then
49 REQUIRED="$((LATEST_SIZE + LATEST_SIZE / 4 + 10737418240))"
50 else
51 REQUIRED=161061273600
52 fi
53 test "$REMOTE_FREE" -ge "$REQUIRED"
54 ssh "${SSH[@]}" "$TARGET" "install -d -m 0700 '$DIR'"
55 ssh "${SSH[@]}" "$SOURCE" "vzdump $VMID --stdout 1 --mode snapshot --compress zstd --ionice 7" 2>"$LOG" | ssh "${SSH[@]}" "$TARGET" "cat > '$PART'"
56 target_worker finalize "$PARENT" "$PREFIX" "$PART" "$FINAL" "$DIR" "$NAME"
57 ssh "${SSH[@]}" "$TARGET" "cat > '$DIR/vzdump.log'" < "$LOG"
58 SIZE="$(ssh "${SSH[@]}" "$TARGET" "stat -c %s '$FINAL'")"
59 SHA="$(ssh "${SSH[@]}" "$TARGET" "sha256sum '$FINAL' | awk '{print \$1}'")"
60 target_worker promote "$PARENT" "$PREFIX" "$DIR" "$TS"
61 tmp="${HEALTH}.tmp"
62 printf 'STATUS=OK\nSERVICE=homelab-vm160-remote-backup\nTIMESTAMP_UTC=%s\nVMID=%s\nSOURCE_NODE=pve02\nTARGET_NODE=pve03\nTARGET_DIR=%s\nARCHIVE=%s\nSIZE_BYTES=%s\nSHA256=%s\nREMOTE_RETENTION_GENERATIONS=1\n' "$TS" "$VMID" "$DIR" "$FINAL" "$SIZE" "$SHA" > "$tmp"
63 chmod 0644 "$tmp"
64 mv "$tmp" "$HEALTH"
65 trap - EXIT
66 rm -f "$LOG"
67 printf 'VM160_REMOTE_BACKUP_STATUS=OK\nTARGET_DIR=%s\nARCHIVE=%s\nSIZE_BYTES=%s\nSHA256=%s\n' "$DIR" "$FINAL" "$SIZE" "$SHA"
===== FILE /usr/local/sbin/homelab-cluster-admin-pull =====
/usr/local/sbin/homelab-cluster-admin-pull: Bourne-Again shell script, ASCII text executable, with very long lines (424)
1 #!/usr/bin/env bash
2 set +e +u +o pipefail 2>/dev/null || true
3 TS=$(date -u +%Y-%m-%dT%H:%M:%SZ)
4 ROOT=/var/lib/homelab-cluster-admin
5 HEALTH_ROOT=/var/lib/homelab-health
6 mkdir -p "$ROOT" "$HEALTH_ROOT"
7 REMOTE=$(ssh -n -o BatchMode=yes -o ConnectTimeout=8 debian@[PRIVATE_IP] 'sudo /opt/homelab-cluster-admin/cluster-observer.sh >/dev/null 2>&1 || true; sudo /opt/homelab-cluster-admin/full-observer.sh >/dev/null 2>&1 || true; printf "__OBS__\\n"; cat /var/lib/homelab-health/cluster-admin-observer.txt; printf "__FULL__\\n"; cat /var/lib/homelab-health/cluster-admin-full-observer.txt' 2>/tmp/homelab-cluster-admin-pull.err)
8 RC=$?
9 OBS=$(printf "%s\n" "$REMOTE" | awk '/^__OBS__$/{f=1;next}/^__FULL__$/{f=0}f')
10 FULL=$(printf "%s\n" "$REMOTE" | awk '/^__FULL__$/{f=1;next}f')
11 OBS_BAD=0; FULL_BAD=0
12 printf "%s\n" "$OBS" | grep -q "^STATUS=OK" || OBS_BAD=1
13 printf "%s\n" "$FULL" | grep -q "^STATUS=OK" || FULL_BAD=1
14 printf "%s\n" "$FULL" | grep -q "BAD=0" || FULL_BAD=1
15 PROBE_OK=$(printf "%s\n" "$FULL" | sed -n 's/.*PROBE_OK=\([0-9][0-9]*\).*/\1/p' | head -n1)
16 PROBE_BAD=$(printf "%s\n" "$FULL" | sed -n 's/.*PROBE_BAD=\([0-9][0-9]*\).*/\1/p' | head -n1)
17 [ "${PROBE_BAD:-1}" = 0 ] || FULL_BAD=1
18 BAD=$((OBS_BAD+FULL_BAD)); [ "$RC" -eq 0 ] || BAD=$((BAD+1))
19 printf "%s\n" "$OBS" > "$ROOT/observer-remote-latest.txt"
20 printf "%s\n" "$FULL" > "$ROOT/full-observer-central-latest.md"
21 OBS_STATUS=$([ "$OBS_BAD" -eq 0 ] && [ "$RC" -eq 0 ] && echo OK || echo WARN)
22 FULL_STATUS=$([ "$FULL_BAD" -eq 0 ] && [ "$RC" -eq 0 ] && echo OK || echo WARN)
23 echo "STATUS=$OBS_STATUS TS=$TS TYPE=cluster-admin-central-pull BAD=$OBS_BAD VMID=180 IP=[PRIVATE_IP] SSH_RC=$RC OBSERVER_STATUS=$OBS_STATUS REPORT=$ROOT/observer-remote-latest.txt" > "$HEALTH_ROOT/cluster-admin-observer.txt"
24 echo "STATUS=$FULL_STATUS TS=$TS TYPE=cluster-admin-full-observer-central BAD=$FULL_BAD WARN=0 VMID=180 PULL_RC=$RC FULL_STATUS=$FULL_STATUS PROBE_OK=${PROBE_OK:-0} PROBE_BAD=${PROBE_BAD:-1} REPORT=$ROOT/full-observer-central-latest.md" > "$HEALTH_ROOT/cluster-admin-full-observer.txt"
25 echo "STATUS=$FULL_STATUS TS=$TS TYPE=cluster-admin-restricted-probes BAD=$FULL_BAD VMID=180 MODE=FORCED_COMMAND_READONLY_CURRENT PROBE_OK=${PROBE_OK:-0} PROBE_BAD=${PROBE_BAD:-1} TARGETS=pve01,pve02,pve03,edge SOURCE=full-observer" > "$HEALTH_ROOT/cluster-admin-restricted-probes.txt"
26 cat "$HEALTH_ROOT/cluster-admin-observer.txt"
27 cat "$HEALTH_ROOT/cluster-admin-full-observer.txt"
28 cat "$HEALTH_ROOT/cluster-admin-restricted-probes.txt"
29 [ "$BAD" -eq 0 ]
CHANGES_MADE=NO
OUTPUT_END