{ "schema_version": 1, "channel": "homelab-runtime", "command_id": "SUPPORT-260922-HOMELAB-BACKUP-1115-ROOT-RCA-1116R1", "status": "FAIL", "rc": 3, "host": "pve01", "mode": "read-only", "component": "homelab-backup-1115-root-rca", "started_at_utc": "2026-09-22T05:57:35Z", "finished_at_utc": "2026-09-22T05:57:35Z", "reference_register_checked": true, "reference_sha256": "5a3d8e5154c41cb582a4e0aca68090be1f0138918bf82131a948df326f9d8d66", "error_register_checked": true, "error_register_sha256": "3b09a553ec0f527ed3afeed4753f52a74ee3036045bbfb3c685e9f8af4ba7ba0", "command_sha256": "a475f4109c57c07002ec72ba49450725574b1c08d698b9878c58257e7cc704c8", "duplicate_failed_command_blocked": false, "block_reason": null, "execution_started": true, "change_declared": false, "result_contract_valid": true, "result_contract_status": null, "result_contract_error": null, "command_rc": 3, "changes_made": false, "rollback_started": false, "rollback_restored": null, "mutation_outcome": "NO_MUTATION", "sanitized": true, "secrets_included": false, "private_addresses_included": false, "raw_evidence_retained_locally": true, "raw_evidence_sha256": "17278f059b28a5b338724f4e22fad8aa2db50e23143eeee6635c075bf93a9691", "sanitized_output_sha256": "17278f059b28a5b338724f4e22fad8aa2db50e23143eeee6635c075bf93a9691", "output_truncated_in_json": false, "full_sanitized_output_url": "https://git.gram1.ru/.well-known/homelab-runtime/latest.txt", "output": "WORKERS2_BACKUP_1115_ROOT_RCA_BEGIN=1\nCOMMAND_ID=SUPPORT-260922-HOMELAB-BACKUP-1115-ROOT-RCA-1116R1\nMODE=read-only\nCOMPONENT=homelab-backup-1115-root-rca\nMUTATION_BOUNDARY=NONE;READ_ONLY_US_NETBIRD_FAILURE_AND_CANONICAL_SKLADCHIK_EVIDENCE_RCA;NO_SYSTEMCTL_ACTION;NO_BACKUP;NO_RESTORE_EXECUTION;NO_GIT_FETCH;NO_GIT_WRITE;NO_CLOUD_WRITE;NO_REMOTE_MUTATION\nERROR_REGISTER_CHECK=OK\nREFERENCE_CHECK=OK\nREFERENCE_SHA256=5a3d8e5154c41cb582a4e0aca68090be1f0138918bf82131a948df326f9d8d66\nERROR_REGISTER_SHA256=3b09a553ec0f527ed3afeed4753f52a74ee3036045bbfb3c685e9f8af4ba7ba0\nRUNNER_SHA256=b248a4c32c9cc64e5747e7dce6c7fc0a23f5124a77c71ce72e27a81aceae9d2d\nWRAPPER_SHA256=5077444065c732451cb84898680f62361b21a24ef9eb4f191253e82ead524ea2\nAUDIT_SHA256=5777bbb204708ffcebba0753506b819833b76370ecda59b4574e1aff3b9e4593\nSCHEDULER_RC=0\nSELFTEST_REPLACEMENT486=PASS\nSCHEDULER_LIVE_SELECTION=DEFER_NO_DUE_TARGET\nUS_NETBIRD_STATE_SAFE_BEGIN=1\nUS_NETBIRD_STATE_SHA256=12d13c0d07056469b1ad4b9328ce29737d1ab76698a4d6115ca1c5648b95ba26\n{\"schema\":1,\"logical_id\":\"us-netbird\",\"result\":\"FAIL\",\"generation\":\"v2-us-netbird-20260922T055316Z-d5c92acdd2e8\",\"updated_at_utc\":\"2026-09-22T05:53:31Z\",\"last_success_epoch\":1789704670,\"last_attempt_epoch\":1790056411,\"retry_after_epoch\":1790078011,\"consecutive_failures\":14,\"replica_count\":null}\n{\"path\":\"consecutive_failures\",\"value\":14}\n{\"path\":\"failure_backoff_seconds\",\"value\":21600}\n{\"path\":\"last_attempt_epoch\",\"value\":1790056411}\n{\"path\":\"last_success_epoch\",\"value\":1789704670}\n{\"path\":\"last_success_generation\",\"value\":\"v2-us-netbird-20260918T041032Z-d7648be4c16b\"}\n{\"path\":\"last_success_manifest_sha256\",\"value\":\"f82c761cafe8aa43d90d0b4b6124933799b5087d34daf880791d850d0d15cee5\"}\n{\"path\":\"last_success_payload_sha256\",\"value\":\"e41e3e012442ef41da1c667d7f5fbe4fd52786b0bb5ffca195660750c773efbf\"}\n{\"path\":\"reason\",\"value\":\"PIPE_FAILED\"}\n{\"path\":\"result\",\"value\":\"FAIL\"}\n{\"path\":\"retry_after_epoch\",\"value\":1790078011}\nUS_NETBIRD_STATE_SAFE_END=1\nUS_NETBIRD_SOURCE_HITS_BEGIN=1\nSOURCE_FILE=/opt/homelab-backup-v2/releases/backup-v2-accepted-486r1/legacy_engine.py SHA256=b6e79f087b9f1d21dac4f77a7b46b743299bbb85bc716e80d2ea67c2e038bcd7\n155- for node in ('pve01','pve02','pve03'):\n156- out.append({'logical_id':'node-'+node,'scope':'node','type':'pve_node_bundle','node':node,'due_seconds':7*d,'enabled':True,'green1':node=='pve01'})\n157- out += [\n158- {'logical_id':'md-relay','scope':'vps','type':'vps_md','host':'[PRIVATE_IP]','port':22,'user':'ops','hostkey_fingerprint':'SHA256:DXJGwunC5tEVvnC3nRxyayTN8FXmyJcVgf+Oa6udQsY','due_seconds':d,'enabled':True,'green1':True},\n159- {'logical_id':'spb-ingress','scope':'vps','type':'vps_spb','host':'[PRIVATE_IP]','port':22,'user':'edgeadmin','hostkey_fingerprint':'SHA256:Ef+HsZb8lIXoadvYBXuGKIL3H7NyJMmf9EYCoJRoSv8','due_seconds':d,'enabled':True,'green1':True},\n160: {'logical_id':'us-netbird','scope':'vps','type':'vps_us','host':'[PRIVATE_IP]','port':22,'user':'root','hostkey_fingerprint':'SHA256:805Q9eh/fePUci42dfY58I6q0lFiiPHn0/M+mckypmE','due_seconds':d,'enabled':True,'green1':True},\n161- {'logical_id':'router-netcraze','scope':'router','type':'router','due_seconds':d,'enabled':True,'green1':True},\n162- ]\n163- sites=[\n164- ('codevipe',160,'pve02','/var/www/forums/codevipe/public','xf_codevipe','single_transaction'),\n165- ('dsmods',160,'pve02','/var/www/forums/dsmods/public','xf_dsmods','single_transaction'),\n--\n216- except subprocess.TimeoutExpired:\n217- for q in (src,z):\n218- try:q.kill()\n219- except Exception:pass\n220- raise BackupError('PIPE_TIMEOUT')\n221: if src.returncode!=0 or z.returncode!=0: raise BackupError('PIPE_FAILED',f'{src.returncode}_{z.returncode}')\n222-\n223-def ensure_dir(p,mode=0o700):\n224- p=pathlib.Path(p); p.mkdir(parents=True,exist_ok=True); os.chmod(p,mode); return p\n225-\n226-def atomic_write(path,data,mode=0o600):\n--\n248- try:\n249- j=json.loads(p.read_text()); d[j['logical_id']]=j\n250- except Exception: continue\n251- return d\n252-\n253:def write_state(logical_id,result,**kw):\n254- if result not in RESULT_CLASSES: raise BackupError('RESULT_CLASS_INVALID')\n255- j={'schema':1,'logical_id':logical_id,'result':result,'updated_at_utc':utcnow(),**kw}\n256- if result=='SUCCESS': j['last_success_epoch']=epoch()\n257- atomic_write(STATE_ROOT/'state'/(logical_id+'.json'),canonical_json(j)+b'\\n',0o600)\n258- return j\n--\n384- if not names: raise BackupError('TAR_ZST_EMPTY')\n385- if required_member and required_member not in names: raise BackupError('REQUIRED_MEMBER_MISSING')\n386-\n387-def validate_payload(t,path):\n388- typ=t['type']\n389: if typ in ('pve_node_bundle','vps_md','vps_spb','vps_us','router'):\n390- validate_tar_zst(path)\n391- elif typ=='pve_guest':\n392- if t['guest_type']=='lxc': validate_tar_zst(path)\n393- else:\n394- # Green set is intentionally CT-only. Scheduled QEMU backups still get dual remote byte readback.\n--\n405-def cluster_inventory():\n406- p=run(['pvesh','get','/cluster/resources','--type','vm','--output-format','json'],timeout=60)\n407- try:return json.loads(p.stdout)\n408- except Exception: raise BackupError('PVE_INVENTORY_JSON')\n409-\n410:def ssh_remote(node,argv,input=None,timeout=1800,check=True):\n411- cmd='PATH=/usr/sbin:/usr/bin:/sbin:/bin; export PATH; exec '+shlex.join([str(x) for x in argv])\n412- if node=='pve01': return run([str(x) for x in argv],input=input,timeout=timeout,check=check)\n413- return run([SSH,'-o','BatchMode=yes','-o','StrictHostKeyChecking=yes','root@'+node,cmd],input=input,timeout=timeout,check=check)\n414-\n415-def _pve_dumpdir_spec(guest_type):\n--\n471- if i+1>=len(argv) or argv[i+1]!=str(scratch): raise BackupError('PROVEN_DUMP_ARGV_TMPDIR_BINDING')\n472- elif '--tmpdir='+str(scratch) not in argv: raise BackupError('PROVEN_DUMP_ARGV_NO_TMPDIR')\n473- return argv,job\n474-\n475-def _default_unprivileged_lxc(node,vmid):\n476: q=ssh_remote(node,['pct','config',str(vmid)],timeout=30)\n477- lines=[x.strip() for x in q.stdout.decode('utf-8','replace').splitlines()]\n478- if 'unprivileged: 1' not in lines: raise BackupError('LXC_NOT_UNPRIVILEGED')\n479- if any(x.startswith('lxc.idmap:') or x.startswith('lxc.idmap =') for x in lines): raise BackupError('LXC_CUSTOM_IDMAP_UNSUPPORTED')\n480-\n481-def mapped_scratch_probe(node,scratch):\n482- tool=shutil.which('lxc-usernsexec') or '/usr/bin/lxc-usernsexec'\n483- pos=[tool,'-m','u:0:100000:65536','-m','g:0:100000:65536','--','/usr/bin/test','-x',str(scratch)]\n484: q=ssh_remote(node,pos,timeout=30,check=False)\n485- if q.returncode!=0: raise BackupError('MAPPED_UID_ACCESS_PROBE_FAILED')\n486-\n487-def _stream_argv_to_local(node,argv,dest,timeout=7200):\n488- dest=pathlib.Path(dest)\n489- if node=='pve01': cmd=argv\n--\n509- if typ!=t['guest_type']: raise BackupError('PVE_GUEST_TYPE_DRIFT')\n510- if typ=='lxc':\n511- _require_exact_source(P270,P270_SHA,'P270_IDENTITY');_require_exact_source(P282,P282_SHA,'P282_IDENTITY')\n512- _default_unprivileged_lxc(node,vmid)\n513- scratch=f'/var/tmp/homelab-vzdump-v2-{gen}'\n514: ssh_remote(node,['install','-d','-m','710','-o','0','-g','100000',scratch],timeout=30)\n515: q=ssh_remote(node,['stat','-c','%a:%u:%g',scratch],timeout=30)\n516- if q.stdout.decode().strip()!='710:0:100000': raise BackupError('PVE_MAPPED_SCRATCH_ACCESS_CONTRACT')\n517- try:\n518- mapped_scratch_probe(node,scratch)\n519- argv,job=proven_lxc_dump_argv(t,row,node,scratch)\n520- local=work/'payload.bin'\n521- _stream_argv_to_local(node,argv,local,timeout=7200)\n522- if local.stat().st_size>int(job['archive_cap_bytes']): raise BackupError('VZDUMP_ARCHIVE_CAP_EXCEEDED')\n523- except Exception:\n524: ssh_remote(node,['rm','-rf','--',scratch],timeout=60,check=False);raise\n525- cleanup={'kind':'pve_source_dir','node':node,'path':scratch}\n526- return local,{'generation':gen,'source_class':'pve_guest','vmid':vmid,'guest_type':typ,'node':node,'status':row.get('status'),'producer_contract':'proven-270-282-stream'},cleanup\n527- # QEMU path remains unchanged: bounded dedicated dumpdir with later cloud readback/reconstruction.\n528- srcdir=f'/var/tmp/homelab-vzdump-v2-{gen}'\n529: ssh_remote(node,['install','-d','-m','700','-o','0','-g','0',srcdir],timeout=30)\n530- try:\n531: ssh_remote(node,['vzdump',str(vmid),'--mode','snapshot','--compress','zstd','--dumpdir',srcdir],timeout=7200)\n532: p=ssh_remote(node,['find',srcdir,'-maxdepth','1','-type','f','-printf','%p\\n'],timeout=30)\n533- files=[x for x in p.stdout.decode().splitlines() if x.strip()]\n534- if len(files)!=1: raise BackupError('VZDUMP_OUTPUT_NOT_UNIQUE')\n535- src=files[0];local=work/'payload.bin'\n536- if node=='pve01': local=pathlib.Path(src)\n537- else: run([SCP,'-q','-o','BatchMode=yes','-o','StrictHostKeyChecking=yes','root@'+node+':'+src,str(local)],timeout=7200)\n538- if not pathlib.Path(local).is_file() or pathlib.Path(local).stat().st_size<=0: raise BackupError('VZDUMP_PAYLOAD_MISSING')\n539- except Exception:\n540: ssh_remote(node,['rm','-rf','--',srcdir],timeout=60,check=False);raise\n541- return pathlib.Path(local),{'generation':gen,'source_class':'pve_guest','vmid':vmid,'guest_type':typ,'node':node,'status':row.get('status')},{'kind':'pve_source_dir','node':node,'path':srcdir}\n542-\n543-def operator_key_fingerprint():\n544- if not OPERATOR_KEY.is_file() or OPERATOR_KEY.is_symlink(): raise BackupError('OPERATOR_KEY_MISSING')\n545- p=run(['ssh-keygen','-lf',str(OPERATOR_KEY),'-E','sha256'],timeout=30)\n--\n563- return good[0]\n564-\n565-def ensure_vps_known_hosts(targets):\n566- lines=[]\n567- for t in targets:\n568: if t['type'] not in ('vps_md','vps_spb','vps_us'): continue\n569- lines.append(scan_hostkey_line(t['host'],t['port'],t['hostkey_fingerprint']))\n570- atomic_write(VPS_KNOWN_HOSTS,('\\n'.join(lines)+'\\n').encode(),0o600)\n571-\n572-def vps_ssh_argv(t, *, use_agent=False):\n573- a=[SSH,'-p',str(t['port']),'-o','BatchMode=yes','-o','StrictHostKeyChecking=yes','-o','UserKnownHostsFile='+str(VPS_KNOWN_HOSTS),'-o','ConnectTimeout=15']\n--\n598- tag=p.stdout.decode().strip()\n599- if tag not in ('ADDED','ALREADY'): raise BackupError('VPS_AUTHORIZED_KEY_BOOTSTRAP_RESULT')\n600- vps_run(t,['true'],timeout=30,use_agent=False)\n601- results[t['logical_id']]=tag\n602- # US was already operator-key based; prove it agentless too.\n603: us=[x for x in targets if x['type']=='vps_us'][0]\n604- vps_run(us,['true'],timeout=30,use_agent=False)\n605- results[us['logical_id']]='ALREADY_US_OPERATOR_KEY'\n606- return results\n607-\n608-def remote_hashset(t,command):\n609- p=vps_run(t,['sh','-c',command],timeout=120,use_agent=False)\n610- return sha256_bytes(p.stdout)\n611-\n612:def capture_vps_md(t,gen,work):\n613- files=['etc/netbird/install.conf','var/lib/netbird/state.json','var/lib/netbird/service.json','var/lib/netbird/active_profile.json','var/lib/netbird/default.json']\n614- cmd=\"sudo -n sha256sum -- \"+' '.join(shlex.quote('/'+x) for x in files)\n615- pre=remote_hashset(t,cmd)\n616- remote=shlex.join(['sudo','-n','tar','--numeric-owner','-C','/','-cf','-']+files)\n617- argv=vps_ssh_argv(t,use_agent=False)+[remote]\n618- payload=work/'payload.tar.zst'; pipe_to_zstd(argv,payload,timeout=600)\n619- post=remote_hashset(t,cmd)\n620- if pre!=post: raise BackupError('MD_SOURCE_CHANGED_DURING_CAPTURE')\n621- return payload,{'generation':gen,'source_class':'vps_md','consistency':'FILESET_PRE_POST_HASH_STABILITY','hashset_sha256':pre},None\n622-\n623:def capture_vps_spb(t,gen,work):\n624- hcmd=\"find /etc/nginx -xdev -type f -print0 | sort -z | xargs -0 sha256sum\"\n625- pre=remote_hashset(t,hcmd)\n626- remote=shlex.join(['tar','--numeric-owner','-C','/','-cf','-','etc/nginx'])\n627- argv=vps_ssh_argv(t,use_agent=False)+[remote]\n628- payload=work/'payload.tar.zst'; pipe_to_zstd(argv,payload,timeout=900)\n--\n657- if dst.execute('pragma quick_check').fetchone()[0]!='ok': raise SystemExit(33)\n658- b=dst.serialize(); dst.close(); src.close()\n659- ti=tarfile.TarInfo(name=('sqlite-backups/'+name)); ti.size=len(b); ti.mode=0o600; ti.mtime=int(time.time()); tf.addfile(ti,io.BytesIO(b))\n660-'''\n661-\n662:def capture_vps_us(t,gen,work):\n663- remote=shlex.join(['python3','-I','-B','-c',US_REMOTE_SCRIPT])\n664- argv=vps_ssh_argv(t,use_agent=False)+[remote]\n665- payload=work/'payload.tar.zst'; pipe_to_zstd(argv,payload,timeout=1200)\n666: return payload,{'generation':gen,'source_class':'vps_us','consistency':'SQLITE_ONLINE_BACKUP_PLUS_NONDB_FILES','sqlite_count':4},None\n667-\n668-def _simple_shell_assignments(text):\n669- vals={}\n670- for line in text.splitlines():\n671- m=re.match(r'^\\s*(?:export\\s+)?([A-Z_][A-Z0-9_]*)=(.*)$',line)\n--\n801-\n802-def qga_exec(node,vmid,guest_argv,input_bytes=None,timeout=600):\n803- args=['qm','guest','exec',str(vmid),'--synchronous','1','--timeout',str(int(timeout))]\n804- if input_bytes is not None: args += ['--pass-stdin','1']\n805- args += ['--']+[str(x) for x in guest_argv]\n806: p=ssh_remote(node,args,input=input_bytes,timeout=timeout+60)\n807- try:j=json.loads(p.stdout)\n808- except Exception: raise BackupError('QGA_JSON')\n809- rc=j.get('exitcode',j.get('exit-code'))\n810- if rc is None: rc=0 if j.get('exited') else None\n811- if rc not in (0,None): raise BackupError('QGA_GUEST_COMMAND_FAILED',str(rc))\n--\n854-\n855-def capture_node_bundle(t,gen,work):\n856- node=t['node']; raw=node_bundle_paths(node); paths=[]\n857- for rel in raw:\n858- if node=='pve01': ok=(pathlib.Path('/')/rel).exists()\n859: else: ok=(ssh_remote(node,['test','-e','/'+rel],timeout=20,check=False).returncode==0)\n860- if ok: paths.append(rel)\n861- required={'etc/network/interfaces','etc/hosts','etc/hostname','etc/pve/nodes/'+node}\n862- if not required.issubset(set(paths)): raise BackupError('NODE_BUNDLE_REQUIRED_PATH_MISSING')\n863- payload=work/'payload.tar.zst'\n864- argv=['tar','--numeric-owner','--xattrs','--acls','-C','/','-cf','-']+paths\n--\n870-\n871-def capture_target(t,gen,work):\n872- typ=t['type']\n873- if typ=='pve_guest': return capture_pve_guest(t,gen,work)\n874- if typ=='pve_node_bundle': return capture_node_bundle(t,gen,work)\n875: if typ=='vps_md': return capture_vps_md(t,gen,work)\n876: if typ=='vps_spb': return capture_vps_spb(t,gen,work)\n877: if typ=='vps_us': return capture_vps_us(t,gen,work)\n878- if typ=='router': return capture_router(t,gen,work)\n879- if typ=='xenforo': return capture_xenforo(t,gen,work)\n880- raise BackupError('CAPTURE_TYPE_UNKNOWN')\n881-\n882-def cleanup_source(cleanup):\n883- if not cleanup: return\n884- if cleanup.get('kind')=='pve_source_dir':\n885: ssh_remote(cleanup['node'],['rm','-rf','--',cleanup['path']],timeout=60)\n886-\n887-def control_verify(t):\n888- engine=pathlib.Path('/usr/local/sbin/homelab-control-backup')\n889- if sha256_file(engine)!='0b3a0f685d4e67cb86250df89a4918aff2e6c67799b7d35b0e54b1bbc016ee57': raise BackupError('CONTROL_ENGINE_IDENTITY')\n890- root=pathlib.Path('/var/lib/homelab-backup/cloud-quorum/control-bootstrap-v2')\n--\n895- if j.get('status')=='SEALED' and j.get('generation'): rows.append((p.stat().st_mtime,j['generation']))\n896- except Exception: continue\n897- if not rows: raise BackupError('CONTROL_SEALED_GENERATION_MISSING')\n898- gen=sorted(rows)[-1][1]\n899- p=run([str(engine),'verify',gen],timeout=1800)\n900: write_state(t['logical_id'],'SUCCESS',generation=gen,verification='existing_control_engine_verify',last_success_epoch=epoch())\n901- return {'logical_id':t['logical_id'],'generation':gen,'result':'SUCCESS','payload_bytes':0,'replicas':2}\n902-\n903:def backup_target(t,green_verify=False):\n904- if t['type']=='control': return control_verify(t)\n905- gen=generation_id(t['logical_id']); work=ensure_dir(WORK_ROOT/gen,0o700)\n906- payload=None; cleanup=None; sealed=False\n907- try:\n908- payload,meta,cleanup=capture_target(t,gen,work); meta['generation']=gen\n--\n916- ha=reconstruct_replica(manifest,manifest['replicas'][0],a); validate_payload(t,a)\nSOURCE_FILE=/opt/homelab-backup-v2/releases/backup-v2-accepted-486r1/wrapper.py SHA256=2b5eb4fc0f6f466a0d3e5a007af987092b6c17baa3ff9241a6afc2c2e97df1d5\n2-from __future__ import annotations\n3-import errno,fcntl,importlib.util,os,sys\n4-from pathlib import Path\n5-ROOT=Path(__file__).resolve().parent\n6-CT_ID='guest-pve01-lxc-112'\n7:LEGACY_IDS=('node-pve01','md-relay','spb-ingress','us-netbird','router-netcraze','xf-codevipe','xf-zakrutim','control-bootstrap-v2')\n8-LOCK=Path('/run/homelab-backup-heavy.lock')\n9-\n10-def _load(name,path):\n11- spec=importlib.util.spec_from_file_location(name,str(path))\n12- if spec is None or spec.loader is None: raise RuntimeError('MODULE_SPEC')\n--\n14-\n15-legacy=_load('_backupv2_legacy486',ROOT/'legacy_engine.py')\n16-ct112=_load('_backupv2_ct112_486',ROOT/'ct112_handler.py')\n17-driver=_load('_backupv2_driver454_479',ROOT/'representative_driver.py')\n18-_legacy_load_targets=legacy.load_targets\n19:_legacy_backup_target=legacy.backup_target\n20-\n21-def canonical_targets():\n22- rows=_legacy_load_targets()\n23- if not isinstance(rows,list) or len(rows)!=28: raise legacy.BackupError('TARGET_REGISTRY_COUNT')\n24- out=[dict(x) for x in rows]\n--\n35- 'guest_type':'lxc','due_seconds':due,'enabled':True,'green1':True}\n36-\n37-def load_targets():\n38- base=canonical_targets();return base+[_ct_descriptor(base)]\n39-\n40:def backup_target(t,green_verify=False):\n41- if not isinstance(t,dict) or t.get('logical_id')!=CT_ID:\n42: return _legacy_backup_target(t,green_verify)\n43- try:\n44: r=ct112.run_target(t,state_writer=legacy.write_state)\n45- return {'logical_id':CT_ID,'generation':r['generation'],'result':'SUCCESS','payload_bytes':r['payload_bytes'],\n46- 'replicas':2,'manifest_sha256':r['manifest_sha256'],'payload_sha256':r['payload_sha256']}\n47- except BaseException:\n48: try: legacy.write_state(CT_ID,'FAIL',reason='CT112_PIPELINE_FAILED')\n49- except BaseException: pass\n50- raise legacy.BackupError('CT112_PIPELINE_FAILED') from None\n51-\n52-legacy.load_targets=load_targets\n53:legacy.backup_target=backup_target\n54-\n55-def _legacy_reused_boundary_selftest():\n56- if legacy.PART_SIZE!=536870912: raise RuntimeError('LEGACY_PART_SIZE')\n57- if len(legacy.REMOTES)!=13 or len({p for _,p in legacy.REMOTES})!=13: raise RuntimeError('LEGACY_REMOTES')\n58- d=legacy.default_targets()\nSOURCE_FILE=/opt/homelab-backup-v2/releases/backup-v2-overlay-504r1-ca6876fa9c3b/entry504.py SHA256=ca6876fa9c3b1465b045344a8ad46962de16972c7dadc4ac6257c7d225047793\n57- if gp==legacy_path:\n58- found[id(g)]=g\n59- if len(found)!=1:\n60- raise RuntimeError('LEGACY_PROVENANCE_CARDINALITY')\n61- g=next(iter(found.values()))\n62: required=('capture_pve_guest','backup_target','load_targets','cluster_inventory','ssh_remote',\n63- '_default_unprivileged_lxc','mapped_scratch_probe','_stream_argv_to_local','run',\n64: 'choose_due_target','write_state','epoch')\n65- if any(not callable(g.get(n)) for n in required):\n66- raise RuntimeError('LEGACY_PROVENANCE_API')\n67- original=g['capture_pve_guest']\n68- proxy=_GlobalsProxy(g)\n69- if not isinstance(g.get('SCP'),str) or not g['SCP'].startswith('/'):\n--\n73- raise RuntimeError('QEMU_ADAPTER_BIND_RESULT')\n74- patched=lxc_adapter.bind(proxy,qemu_patched)\n75- if not callable(patched):\n76- raise RuntimeError('LXC_ADAPTER_BIND_RESULT')\n77- g['capture_pve_guest']=patched\n78: scheduler_wired=scheduler_adapter.bind(proxy,g['choose_due_target'],g['write_state'])\n79: if g.get('choose_due_target') is not scheduler_wired['choose_due_target'] or g.get('write_state') is not scheduler_wired['write_state']:\n80- raise RuntimeError('SCHEDULER_ADAPTER_BIND_RESULT')\n81- return {'legacy_globals':g,'legacy':proxy,'original_capture':original,'qemu_patched_capture':qemu_patched,'patched_capture':patched,'scheduler_wired':scheduler_wired}\n82-\n83-def _load():\n84- _verify_base()\n--\n105- wrapper,lxc_adapter,qemu_adapter,scheduler_adapter,wired=_load()\n106- if wired['legacy_globals'].get('capture_pve_guest') is not wired['patched_capture']:\n107- raise RuntimeError('PATCH_BINDING')\n108- if wired['legacy_globals'].get('choose_due_target') is not wired['scheduler_wired']['choose_due_target']:\n109- raise RuntimeError('SCHEDULER_CHOOSE_BINDING')\n110: if wired['legacy_globals'].get('write_state') is not wired['scheduler_wired']['write_state']:\n111- raise RuntimeError('SCHEDULER_WRITE_BINDING')\n112- if [scheduler_adapter.backoff_seconds(n) for n in range(1,7)]!=[1800,3600,7200,14400,21600,21600]:\n113- raise RuntimeError('SCHEDULER_BACKOFF_CONTRACT')\n114: if not callable(wrapper.backup_target) or not callable(wrapper.main) or not callable(wrapper.load_targets):\n115- raise RuntimeError('WRAPPER_PUBLIC_API')\n116- print('SELFTEST_OVERLAY504=PASS',flush=True)\n117- return 0\n118-\n119-def main(argv=None):\n--\n131- try:\n132- with tempfile.TemporaryDirectory(prefix='homelab-scheduler-proof-',dir='/run') as td:\n133- g['STATE_ROOT']=pathlib.Path(td); (g['STATE_ROOT']/'state').mkdir(mode=0o700)\n134- g['epoch']=lambda: now[0]\n135- common={'payload_sha256':'a'*64,'manifest_sha256':'b'*64,'replicas':['proof-r1','proof-r2']}\n136: legacy.write_state('proof-a','SUCCESS',generation='proof-a-good',**common)\n137: legacy.write_state('proof-b','SUCCESS',generation='proof-b-good',**common)\n138- now[0]=100000\n139: legacy.write_state('proof-a','FAIL',generation='proof-a-bad',reason='PROOF_FAILURE',sealed=False)\n140- state=legacy.load_state_all(); a=state['proof-a']\n141- targets=[{'logical_id':'proof-a','enabled':True,'due_seconds':100},{'logical_id':'proof-b','enabled':True,'due_seconds':100}]\n142- first=legacy.choose_due_target(targets,state,now[0])\n143- if not isinstance(first,dict) or first.get('logical_id')!='proof-b': raise RuntimeError('SCHEDULER_PROOF_PEER_NOT_SELECTED')\n144: retry=int(a.get('retry_after_epoch',0)); preserved=int(a.get('last_success_epoch',0))\n145: if preserved!=1000 or a.get('last_success_generation')!='proof-a-good' or int(a.get('consecutive_failures',0))!=1 or retry!=101800:\n146- raise RuntimeError('SCHEDULER_PROOF_FAILURE_STATE')\n147- now[0]=retry+1\n148- second=legacy.choose_due_target(targets,legacy.load_state_all(),now[0])\n149- if not isinstance(second,dict) or second.get('logical_id')!='proof-a': raise RuntimeError('SCHEDULER_PROOF_RETRY_NOT_REELIGIBLE')\n150: legacy.write_state('proof-a','SUCCESS',generation='proof-a-recovered',**common)\n151- recovered=legacy.load_state_all()['proof-a']\n152: if int(recovered.get('consecutive_failures',-1))!=0 or int(recovered.get('retry_after_epoch',-1))!=0:\n153- raise RuntimeError('SCHEDULER_PROOF_SUCCESS_RESET')\n154: proof={'failed_target':'proof-a','peer_selected_during_backoff':first['logical_id'],'retry_after_epoch':retry,'last_success_preserved':preserved==1000,'retry_target_after_backoff':second['logical_id'],'success_resets_failures':True,'production_state_root_used':False}\n155- print('SCHEDULER_FAILURE_ISOLATION_PROOF='+json.dumps(proof,separators=(',',':'),sort_keys=True),flush=True)\n156- return 0\n157- finally:\n158- g['STATE_ROOT']=old_root; g['epoch']=old_epoch\n159- if argv==['scheduler-select-live']:\n--\n174- if len(argv)==2 and argv[0]=='verify-target':\n175- t=_target(wrapper,argv[1])\n176- if t.get('logical_id')=='guest-pve01-lxc-112':\n177- raise RuntimeError('VERIFY_SYNTHETIC_CT112_FORBIDDEN')\n178- try:\n179: out=wrapper.backup_target(t,True)\n180- except legacy.BackupError as e:\n181- print('VERIFY_TARGET_RESULT=FAIL',flush=True)\n182- print('VERIFY_TARGET_REASON='+str(e.code),flush=True)\n183- return 2\n184- print('VERIFY_TARGET_RESULT='+json.dumps(out,separators=(',',':'),sort_keys=True),flush=True)\nSOURCE_FILE=/opt/homelab-backup-v2/releases/backup-v2-overlay-504r1-ca6876fa9c3b/scheduler_adapter.py SHA256=a6f67159a64067773323f028126e3109e15cf17a8142cea5e6129970f1d4da96\n14- n=int(v)\n15- except Exception:\n16- return default\n17- return n if n>=0 else default\n18-\n19:def backoff_seconds(consecutive_failures:int)->int:\n20: n=max(1,_as_nonneg_int(consecutive_failures,1))\n21- return min(MAX_BACKOFF,MIN_BACKOFF*(2**min(n-1,4)))\n22-\n23-def _read_prior(legacy,logical_id):\n24- root=pathlib.Path(legacy.STATE_ROOT)/'state'\n25- p=root/(str(logical_id)+'.json')\n--\n60- if not isinstance(t,dict) or t.get('enabled') is not True: continue\n61- lid=t.get('logical_id'); seconds=_as_nonneg_int(t.get('due_seconds'),0)\n62- _need(isinstance(lid,str) and lid and seconds>0,'SCHEDULER_TARGET_SCHEMA')\n63- s=state.get(lid,{})\n64- if not isinstance(s,dict): s={}\n65: retry_after=_as_nonneg_int(s.get('retry_after_epoch'),0)\n66: if retry_after>now: continue\n67- last_success=_as_nonneg_int(s.get('last_success_epoch'),0)\n68- age=(now-last_success) if last_success and now>=last_success else (10**12 if last_success==0 else 0)\n69- if age