+ jHjDxRt^RIHt^RIt^RIt^RIHt^RIt]P!] 4t Rt Rt Rt RtRtR tR tR tR tR tRtRtRtRtRtRtRtRtRtRtRRRR/RltRt R1Rlt!Rt"R1Rlt#!R R!]$4t%R"t&R#t'R$t(R%t)R&t*R't+R(t,R)t-R*t.R+t/R,t0R-t1R.t2R/t3R2R0lt4R#)3aSession-runtime lifecycle owned by the ordinary gateway bootstrap. One SessionAuthority per reserved profile home. The launch home always has one; under ``gateway.multiplex_profiles`` every served secondary gets its own, built under that profile's runtime scope against that profile's ``state.db``. ``runner.session_authority`` stays the launch profile's authority so single-profile behaviour is byte-identical. The served set is not frozen at boot: ``serve_profile_runtime`` / ``unserve_profile_runtime`` grow and shrink it for the hot-serve reconcile, and a secondary whose store cannot be opened is parked (logged, left out of ``served_profiles``) instead of aborting every other profile. ) annotationsN)Pathc^RIHpV!4P4p\VRR4;'gRV3.p\VPRR4'd|\VPRR4;'gRpVFTwrV\ V4P4pWr8wgK&YsUU u0uFwrV kK up p9gKBVP WW34KV V#uup pi) zY``(name, home)`` pairs this process reserved: launch home first, then served secondaries.get_hermes_home_primary_profile_nameNdefaultmultiplex_profilesF_runtime_profile_homes)hermes_constantsrresolvegetattrconfigrappend) runnerrlaunchhomesreservednamehome canonical_hs & 8/home/ubuntu/.hermes/hermes-agent/gateway/run_runtime.pyreserved_profile_homesrs0   & & (Ff5t<II 6 R SEv}}2E::6==*BDIOOR"JDT **,I"yu8Mutqu8M'M d./# L9Ns! C c "^RIHp^RIHpVPpVP R,pV!VRR7;_uu_4\ VPRVP4pVf\\W44h\P!\VP4P4GRjxL PV8wd\RV 24hVP!VRVR 7V!V\#V4WxVR 7GRjxL p RRR4TP)TX 4T #LpL# \$dTP'T4hi;i +'giLI;i5i) zYOne profile's authority against its own ``state.db``; raises when that store is unusable.)_profile_runtime_scope)initialize_session_authority instance_idF)hydrate_secrets_dbNzCsession authority database does not belong to the reserved profile )r) profile_idrdbregister) gateway.runrgateway.session_authorityrsession_authoritiessession_runtime_descriptorr _session_db RuntimeError_store_open_failureasyncio to_threadrdb_pathr parentaddstr BaseExceptionremovereplace) rrrr$rrregistryrr# authoritys &&&$ r_build_profile_authorityr7$s)2F))H33MBK e < < V''0B0B C :26@A A##D$4$<$<= = E E M!deidjkl l T4d + :3t9+W_aaI = T9%  >a  OOD !   = EA0E0D!15E'D%?D#D%E!E#D%%EE E EcaVR, o\\VRR4RR4\VRR43p\RV3RlV44R4pV'dS R V 2#S R 2#) aWhy *home*'s ``state.db`` has no handle: the scoped handle caches answer None for a failed open (and while they back off), so the park reason named an ownership mismatch over the real cause, a legacy store that crashed its migration on every open (#141467). The reason is the exception the failed open recorded; reopening here to learn it could create or migrate the store, and would skip the caches' backoff.zstate.db session_storeN_db_handle_cache_session_db_handle_cachec38"TFq'gK VxK R#5iNr ).0es& r &_store_open_failure..FsXRVW!!Rs  c3P<"TFqfKVPS4xK R#5ir=) last_error)r>cpaths& rr@rAFs R&Q0all400&s&&z did not open: z5 is unavailable (its last open failed; retry pending))rnext)rrcacheserrorrEs&& @rr+r+=sv * Dgfot<>PRVWf8$?AF XR&RXZ\ ]E ug..VH IIch\PRWV4\W4\WRV 24R#)zA secondary whose store is unusable is parked: logged, unreserved, left out of ``served_profiles`` and named under ``parked_profiles`` in the runtime descriptor. Boot continues for every other profile (a runtime hot-add parks the same way).zK[MULTIPLEX] Profile '%s' not served: its session store is unusable (%s): %szsession store unusable: N)loggerrIrelease_profile_home park_profile)rrrexcs&&&&r_park_reserved_profilerPLs2 LL^S"&!9#?@rJcZ\V4p\V4R,W1&\V4R#)aPublish *name* under ``parked_profiles`` (name -> reason) in the runtime descriptor and the root's ``gateway_state.json``: a client of that profile then gets a terminal ``profile_parked`` verdict from ``ensure`` instead of waiting its whole deadline for a service that never comes.:NiNN)parked_profile_mapr1_record_parked_profiles)rrreasonparkeds&&& rrNrNVs&  'Fv;t$FLF#rJc^\V4pVPVR4e\V4R#R#r=)rRpoprS)rrrUs&& runpark_profilerX_s*  'F zz$)'*rJc\VRR4;'g/p\V4;'dYPR4;'g/9#)zTrue when *name* is parked: not served now, but a rescan serves it again, so its ingress is temporarily unavailable (retry it), never an invalid route (refuse it for good).r(Nparked_profiles)rboolget)rr descriptors&& rprofile_parkedr^esB!=tDJJJ : K K$>>2C#D#J#JKKrJcv^RIHp^RIHpV!V!44p\ W4'd^RIHpV!V4hR#)zAn API request under a parked profile's ``/p//`` prefix finds no authority: refuse it retryably (``runtime_draining``, a 503 with ``Retry-After``), not with the final ``profile_mismatch`` of a profile this gateway does not serve.)served_profile_namer)ParkedProfileRefusalN)gateway.session_authoritiesr`r rr^gateway.parked_route_refusalra)rr`rrras& rrefuse_parked_scoperdls6@0 0 1Df##E"4(($rJcX\VRR4pVf /;qnVPR/4#)r(NrZ)rr( setdefault)rr]s& rrRrRxs6!=tDJ9;; 6  !2B 77rJc^RIHpV!\V4R7R# \d\P RRR7R#i;i))write_runtime_status)rZz could not record parked_profilesTexc_infoN)gateway.statusridict ExceptionrLdebug)rUris& rrSrSs9H7T&\: H 7$ GHs"AAc "^RIHp^RIHp^RIHp\ V4pVF*wrVVPV4'dK\RV 24h \P!4PpRVR^RRR .R ./pWn V!V^,^,\VPR R 4R 7p Wn\!V4F!wp wr\#W Wj^8HR7GRjxL K# V P(P*VR&V P-4VR &\/\1V44V!WyP344Vn\7V4R#Lp \$d p T ^8Xdh\'Y Yl4Rp ? KRp ? ii;i5i)rh) TicketStoreprocess_ownership)SessionAuthoritiesz7session authority requires reserved profile ownership: rruntime_protocolstatestarting capabilitiesserved_profilesr F) multiplexedr$Nauthority_epoch)gateway.runtime_bootstraprqgateway.runtime_ownershiprsrbrtrownsr*uuiduuid4hexr(rrr' enumerater7rnrPrepochryrSrR profile_idssession_ticket_storerelease_unknown_turn_markers) rrqrsrtr_namerrr]r5indexrrOs & rinitialize_gateway_runtimers`5;> "6 *E  %%d++!XY]X^_` `**,""K{$6^R2J )3%!%(1+76==Rfhm;noH!)(/| <*6QR S S S 0%-OO$9$9J !$,$<$<$>J !.v67"-k;O;O;Q"RF ( T <z "6 ; ; "C;%C;>*C;6C;; D cVPpVP4VPR&VP4VPnR#)ryN)r'ryr(rr)rr5s& r_publish_served_setrs>))H;C;S;S;UF%%&78.6.B.B.DF+rJcXa^RIHp\S4P4oVP S.4\ VP RR4pVe]\;QJdV3RlV4F 'dK RM RM!V3RlV44'd.VOVS3N5VP nR#R#R#)uGrow the process reservation by one profile (hot-serve): take its ``gateway.lock`` and add it to the frozen boot set, so every reader of the reservation — and the next restart's all-or-nothing reserve — sees it. ``OwnershipConflict`` when another gateway owns the home.rrr Nc3`<"TF#wr\V4P4S8gxK% R#5ir=rr )r>_nrrs& rr@'reserve_profile_home..s##T8%"DGOO$5$=8s+.FT) r~rsrr reserverrallr )rrrrsrs&&f rreserve_profile_homers< :   Dtf%v}}&>EH#T8#T#T8#T T T/H/HD$rrs& rr@)_withdraw_profile_home..s/5M'e4a>+A+A+Ct+KEExs'9 9)rrtupler )rrrs&f rrrsYv}}&>EH/4u5M'5Muu5M'5M0M ,rJc>VPP4Uu.uFqPfKVPNK ppVP\ VRR44^RIHpVPV!V44\\PV44#uupi)zISnapshot every async task family owned directly by one session authority._bot_receipt_tasks)mutation_tasksr ) sessionsvaluestaskextendrgateway.session_runtime_workersrlistrmfromkeys)r6livetasksrs& r_authority_tasksrsu#,#5#5#<#<#> X#>4))YTYY#>E X LL$8"=>> LL *+  e$ %% Ys BBgF@g$@c^RIHpV!V4#)zSUnmapped claims, managed workers and owner mutations across the served authorities.)uncounted_runtime_work)rr)rrs& rmanaged_turn_countrsF !& ))rJ in_processFrTc^RIHp^p\VR/4p\VP 44FwpVP P 4'g0\VRR4e#VPP 4'dKTW&nVPRR/4V^, pKy V\WV4, pV'gV#^RIHp^RIHp \VP$R /4p \VP&P)44FwrV P*P,P/R 4p V eW9gV \VR /49dKHVP0P/V R4wrW8wdV P/V P24pVeWJdWP4V &M V!VR R R7V^, pK V# Td,p\PRTP4Rp?ELVRp?ii;i)aCooperative Stop for every turn *authority* is executing: the control a user's Stop sends to a managed worker (acknowledged, or terminated within STOP_ACK_SECONDS, and the admission settles through its own fenced path). ``in_process`` also hard-interrupts the running agent of each executing in-process turn, or latches the generation ``adopt_agent`` consumes before it exists. ``reason`` (a gateway restart/shutdown) is recorded on each worker it signals, so the turn settles with that cause; a worker already stopped (the user's Stop came first) keeps none. A worker still being spawned gets the same Stop and cause through latches (``_stop_spawning_managed``).RuntimeStoreError_managed_workersstopNtypez4managed Stop not queued (%s); the latch escalates it)request_hard_interrupt)_AGENT_PENDING_SENTINEL_running_agentsexecution_generation_managed_spawningzProfile stoppingzgateway shutdown) tool_reasonNN)hermes_state_runtimerrrrclosedis_setr stop_causecontrolrLrorT_stop_spawning_managedagent.interrupt_compatrr%rrritems event_stream executionr\adoptedroute pending_stops)r6rrTr signalledworkersworkerrOrrrunningsidr generationadopted_generationagents&$$ rstop_authority_turnsrs7Ii!3R8Gw~~'( ==   ! !gffd&C&OTZT_T_TfTfThTh " ] NNFF+ , Q )' FCCI =3i&&(92>G),,2245 &&00445KL  3')M`bd:e3e $-$5$5$9$9#|$L!  +KK +E =E<+5 # #C ( "5* '' eiin 2Y(=tDF~9;;6%.FKc 1 U NrJc@a\V3Rl\V444#)zP``stop_authority_turns`` across every served authority (whole-runtime shutdown).c3><"TFp\VSR7xK R#5i))rTN)r)r>r6rTs& rr@%stop_managed_turns..:sdOc)#If==Ocs)sumr)rrTs&frstop_managed_turnsr8s d|\bOcd ddrJc^RIHpV!W4#)zYWhole-runtime shutdown: interrupt every counted side agent (``session_runtime_workers``).)interrupt_side_agents)rr)rrT interrupts&& rrr=sR V $$rJcr"^RIHp^RIHpHpVf\ MTpRVn\VPRR4e\VP4T!V4^RI H pV!V4\VRR4pVfRM5\P!\P!VPVR74pV!W4GRjxL RpVRJ;'g VGRjxL R Jp V 'g!\ P#R VP$4V'd V 'gR #\'V4p V Fp V P)4K V 'd\P*!V R R/GRjxL T!V4RVn^R IHp VP2P54VP6P94F2p V P;4'dKV P=V !R44K4 VP6P54R#EL> \d'\ P#R TP$4R pELji;iEL]L5i)aStop profile-local services and turns before its ownership is released. Claims are refused from here on; every executing turn gets a cooperative Stop, and its session task is joined (an in-process turn's task ends only after its executor thread returned and the admission settled). The hosted-room stop runs alongside that join under the same deadline: it never interrupts accepted turns and its room threads wait for the member turns they observe, so those turns must be signalled first or the room stop cannot finish. False when a turn or room thread misses *timeout* (default TURN_SETTLE_SECONDS): it may still call tools and write history, so the caller must keep the profile's reservation and store handles.) unbind_owner)join_authority_workstop_authority_workNTrrhosted_room_servicetimeoutz3Profile %s workers did not stop; ownership retainedFz8Profile %s hosted rooms did not stop; ownership retainedreturn_exceptionsrruntime_draining)gateway.session_cronrrrrTURN_SETTLE_SECONDSretiringrrrgateway.session_llm_oneshotrr, ensure_futurer- TimeoutErrorrLrIr"rcancelgatherretiredrrnative_waitersclearwaitersrdone set_exception)r6rrrr stop_oneshotsserviceroomsworkers_stopped rooms_stoppedrestrrwaiters&& r_retire_profile_authorityr Cs2X%,_!'GIy!7>JI,,- "A)i!6=GOD)>)>w?P?PQXQ]Q]gn?o)pE !)555TM==5[%=M  OQZQeQef   I &D  nnd;d;;;I7 ""$##**,{{}}  !23E!F G- 9 6   JIL`L`a &1 =G;>G> H7H7H2 H7$'H7 H7-H7H7H5A#H77H7;G>>-H/+H7.H//H75H7c]tRtRtRtRtR#)ProfileOwnershipRetainedixzZA retired attempt's writer is still live: the profile keeps its reservation and authority.r N)__name__ __module__ __qualname____firstlineno____doc____static_attributes__r rJrrrxsdrJrc"^RIHp^RIHpHp^RIHp^RIHp\P!\V4P4GRjxL pVPpVPV4p V eP\V RR4'd=\!V ^R7GRjxL 'g\#R V R 24hVP%V4MV eV #\'WVRR 7GRjxL p V!V 4;_uu_4\)V 4\+V 4GRjxL V!V R R 7V!V 4GRjxL \-W 4RRR4\VRR4eV!W 4GRjxL \1T4T!T4T #EL"LLLlLT +'giLO;iL7 \.d.\!T 4GRjxL 'dTP%T4hi;i5i)usHot-serve one reserved profile's runtime: build its authority, recover its durable state and publish it in the descriptor/ticket store — the steps boot performs per secondary. Raises when the profile's store is unusable; the caller parks it (and releases the reservation). Its queued native rows are recovered once its adapters connect (``recover_served_native_rows``).r)_ensure_hosted_servicestart_ready_hosted_servicesrecover_local_sessionsrecover_webhook_finalizationsNrFrzan earlier serve of z is still stoppingr{Tschedulesession_control_server)rbrgateway.session_hosted_servicerrgateway.session_local_recoveryr!gateway.platforms.webhook_ingressrr,r-rr r'rrr rr3r7_drop_retired_local_adapterrecover_bot_mailbox_recover_published_api_turnsr2r) rrrrrrrrr5existingr6s &&& rserve_profile_runtimer(|s 8bEO""4:#5#56 6D))H  &H*e D D/xCCC*-A$GY+Z[ [  .vTERRI   # # ' 2%i0 0 0 "9t </ : : : ( ; $ 63T : F(; ; ; ' 9 7D S 1 : $ # < *95 5 5 OOD ! sAG E+ AGE. G8GE0GF +E6E2E6E4 E6/"F F F G.G0G2E64E66 F  F G$F'% G0GGc^RIHp^RIHpV!V4p\ VP VP 4RR4pVe+W@Jd$\ VRR4'dW1P R#R#R#R#)aTA LOCAL adapter still bound to this profile's previous authority, once that authority fully retired, gives way to *authority*'s own: restore and create otherwise refuse every local session of the served profile as ``runtime_draining``. A retained writer never gets here (the serve is refused while the previous authority is still retiring).Platform)local_adapter_mapr6NrF)gateway.configr+r"r,rr\LOCAL)r6r+r,adapterspreviouss& rr$r$s` (@ +Hx||HNN3[$GH 9ghPY[`>a>a ^^ $?b 9rJc^RIHp\VRR4;'g/PVP4pVe^RIHpV!W14R#R#)zBoot recovers queued API turns when the launch API adapter is published (``recover_api_turns``); a profile served after that runs the same pass on its own store.r*r/N)recover_authority_api_turns)r-r+rr\ API_SERVERgateway.session_api_turnr2)rr6r+apir2s&& rr&r&sC( 6:t , 2 2 7 78K8K LC H#C3rJc"\VRR4fR#^RIHpV!V4Fp\W4GRjxL K R#L \d'\ P RTPRR7KOi;i5i)aBoot's native-row recovery once hot-served profiles' adapters connected (boot runs it after every adapter connects), in every serving ledger: a ``bot_profile`` route runs a bot's messages on another profile, so the new bots can carry rows queued in any ledger. A session already draining keeps its one drain. A store that cannot answer leaves its rows queued for the next start and never stops the other ledgers or the reconcile.r'N)serving_authoritiesz%Native session recovery failed for %sTrj)rrbr7recover_profile_native_sessionsrnrLrIr")rr7r6s& rrecover_served_native_rowsr9sm v,d3;?(0  g1&D D D1 D g LL@)BVBVaeL f gs5#A5A?AA5A-A2.A51A22A5c0"\P!\V4P4GRjxL pVPpVP V4pVfR#\ V4GRjxL pV'gR#VPV4\V4V#LbL.5i)zRetire one profile's authority (deleted while running) and shrink the published set. False when one of its turns outlived the Stop deadline (keep the reservation); True otherwise, including for a profile this process never served.NTF) r,r-rr r'rr r3r)rrr5r6rs&& runserve_profile_runtimer;s~""4:#5#56 6D))H!!$'I-i88G  OOD N 7 9s'-BB5B%B& B2!BBc^RIHpV!V4#)rh)all_authorities)rbr=)rr=s& rrrs; 6 ""rJc"^RIHp^RIHpV!V4GRjxL R#L Td7p\P RTP PT4Rp?R#Rp?ii;i5i)aOwner-startup Bot Chat recovery, skipped when this token cannot use the mailbox dir. An elevated install/update can leave ``runtime/bot_live_delivery`` owned by Administrators with no entry for the user (#139485, #138872): the gateway cannot even open its ``.lock``. Bot Chat deliveries pause (``_locked`` already logged the dir, its owner and the remedy); every other client keeps working instead of the whole gateway refusing to start. Any other failure still aborts startup.)recover_bot_deliveries)MailboxUnusableNz$Bot Chat recovery skipped for %s: %s)gateway.session_botr?tools.bot_live_deliveryr@rLwarningr#r.)r6r?r@rOs& rr%r%sQ;7Z$Y/// Z=y||?S?SUXYYZs6 A$ $"$A$$A!+AA$A!!A$c"^RIHpV!V4GRjxL VnVPPVPR&\ V4Fp\ W4GRjxL K R#LSL 5i)rh)start_gateway_apiN api_origin)gateway.run_apirE session_apirFr(r_recover_profile_bot_mailbox)rrEr6s& rstart_gateway_runtime_apirJs[1088F6<6H6H6S6SF%%l3!&) *6===*9 >s"A,A(AA,A* A,*A,c "^RIHpV!V4;_uu_4\V4GRjxL RRR4R#L +'giR#;i \dp\ TRR4pTfRMTP T4pTfh\ PRYSRR7\P!\TP4P4GRjxL p\T4GRjxL 'g$\Y4\YRT RT 24Rp?R#TP!T4\#T4\%Y4\YR T R T 24Rp?R#Rp?ii;i5i) akStartup Bot Chat recovery of one served profile. A secondary whose durable leftovers cannot be recovered is parked like one whose store cannot be opened (retired, unreserved, named under ``parked_profiles``): under multiplex it would otherwise take every other profile down with it, on every restart (#140018). The launch profile's failure still aborts the start.rNr'z@[MULTIPLEX] Profile '%s' not served: startup recovery failed: %sTrjzQstartup recovery failed; ownership retained (a control-socket rescan, `hermes -p z1 gateway run`, serves it once its writer stops): z=startup recovery failed (a control-socket rescan, `hermes -p z! gateway run`, serves it again): )rbrr%rnr profile_namerLrIr,r-rr"r r rrNr3rrM)rr6rrOr5rrs&& rrIrIsD 8_  # #%i0 0 0$ # 0$ # # _6#8$?'tX-B-B9-M <  WY]"  $&&tI,@,@'A'I'IJJJ.y999 #6 0 *559F:klokp(r s F#V*V&1156WX[W\$^ _ _%_sEA:8:AE: A A E A EA7ECE$C'% E0E E9E EEEcx"^RIHpHpV!V4pV'g/#^RIHpV!V4GRjxL ^RIHp^RIHp/pVFXpV!V4;_uu_4V!VRR7V!V4GRjxL RRR4TP\W4GRjxL 4KZ V#LrL6 +'giL<;iL%5i)zRecover against the published routing index and currently connected adapters. Stored envelopes are input, not authority to create a route or reconnect a transport. The authority preflights every queued sender before any claim. )rr7)ensure_hosted_serviceNrrTr) rbrr7r!rNr"rr#rupdater8) rrr7 authoritiesrNrrresultsr6s & rrecover_gateway_native_sessionsrRs M%f-K  D  '''EOG  # # "9t </ : : :$ sl22J ##%%V___@P@P@PIJJG3KLJ'rJc"\P!VP44pVPPp\P !W0\P R7GRjxL wr4W#9ddVP'gRVPPR.R7VP4'dRMVP4p\R4VhVGRjxL VP4'gVP4\P!VRR7GRjxL R#LLNL TP4'gTP4\P!TRR7GRjxL i;i5i)zIA vanished interactive listener is fatal, not a healthy headless runtime.) return_whenNfailedr`z(gateway session API stopped unexpectedlyT)r)r, create_taskwait_for_shutdownrHrwaitFIRST_COMPLETEDrbr(rO cancelled exceptionr*rrr)rshutdownlistenerrrrIs& rwait_gateway_runtimeroHs""6#;#;#=>H!!&&H ? h%9wG^G^__  F$4$4$4  - - 4 48RT 4 U$..00Dh6H6H6JEIJPU U}} OO nnX>>>`  ?}} OO nnX>>>sl;E<*D0(D*)D02D0:#D0D,D0"AE<#D.$E<*D0,D0.E<0AE92E53E99E<c"\VRR4pVfR#RVnVPR.R7^RIHpVP R4V8XdVR^RIHpV!V4GRjxL VPP4FpVPR4K \VR /4P4F+pVP4FpVPR4K K- R#L5i) zFWithdraw admission before any await; close sockets before DB teardown.r(NTdrainingr`)RESTART_WAIT_REASON drain_reason)stop_hosted_service_profile_adapters) rrbrOgateway.drain_controlrrr\r!rtr/rset_message_handler)rr]rrrtadapterr/s& rdrain_gateway_runtimeryYs!=tDJFJR89~~n%)<< ~ &B f %%%??))+##D),F$7<CCE(G  ' ' -)F &sAC% C#!BC%c t"\V4p\W4^RIHpVF pV!V4K VUUu.uFp\ V4FqfNK K pppVf\ M$\ R\\ V\, 44pV'dV^8d \P!WxR7GRjxL M\4\V43wrV F+pVP4'dKVP4K- V 'd!\PR\!V 4V4^RIHp VFpV !V^4GRjxL K ^R IHp \V4F pV !V4K ^RIHp\)VRR4pVeVP;4\)VRR4pVeV!V4GRjxL R#R#uuppiEL L~ \&d&\PR\)TRR 44Ki;i \.d<^RIp T P2!\44PR \)TRR 4R R 7Ki;iL5i)asKeep authority tasks alive until their last durable settlement write, within a bound. A managed turn still running here gets its Stop first (its worker acknowledges or is terminated, and the admission settles interrupted or ``unknown``); a task that still misses TURN_SETTLE_SECONDS (clamped to ``budget``, what the stop's watchdog leash has left, minus STOP_TAIL_RESERVE_SECONDS) is logged and left running for recovery. The stop sequence continues: the executor quiesce skips the DB close while a writer is live, and process exit releases the ownership lock, so raising here would only skip teardown and disarm the watchdog.rNgrzC%d authority task(s) did not settle within %.0fs; left for recovery)rzOProfile %s still has live writers at shutdown; its admissions stay for recoveryr"?)end_idle_acp_sessionsz(ACP sessions of %s not ended at shutdownTrj)stop_gateway_apirrH)rrrrrrmaxminSTOP_TAIL_RESERVE_SECONDSr,risetrkrlrLrClenrrrrgateway.session_acp_lifecycler|rnlogging getLoggerrrGr}revoke)rbudgetrTrPrr6rrrirrYrr|rr}rhandles&&& rsettle_gateway_runtimernsv&Kv&A i !#. W;i;KI;V4T;VT;E W"(. c#sCVCILeCe@g7hD CG!8gll5???RURWY\]bYcQd D>>##   NN`w< /C  B%i3 3 3!D!&)  f !) ,*1 F2D 9E   V]D 1F v&&&E X@ 4 B NNl"9lC@ B B f    h ' / /0Z07 H8G-AH8&H6'H88F::,G*&H8)G**H8-AH3/H82H33H8r=r)5r __future__rr,rpathlibrrrrrLrr7r+rPrNrXr^rdrRrSrrrrrMrrrrrrrrrr r*rr(r$r&r9r;rr%rJrIrRr8rcroryrr rJrrs% #   8 $ 2 JA$( L )8H)BX(E I$M&!* (%((V(e % 2je|e%P %4 g "# Z >_>. (?".*1'rJ