+ jHja0t$Rt^RIHt^RIt^RIt^RIt^RIt^RIt^RI t ^RI H t ^RI H t RsR]R&R]!4R]!4R ]!4/tR ]R &R tR t!RR]4tRRltRRltRRltRRltRZRRlltRRltRRltRRltR R!ltR"R#ltR$R%lt R&R'lt!R(R)lt"R*R+lt#R,R-lt$R.R/lt%R0R1lt&R2R3lt'R4R5/R6R7llt(R8R9lt)R:R;lt*R[R<R=llt+R>R?lt,R@RAlt-RBRClt.RDRElt/RFRGlt0RHRIlt1RJRKlt2RLRMlt3RNROlt4RPRQlt5RRRSlt6RTRUlt7RVRWlt8RXRYlt9R#)\a Pre-swap gateway pause for ``hermes update`` on Linux and macOS. Windows pauses its gateways before the fetch (``update_cmd_windows``) because they lock venv files. On POSIX nothing is locked, but a gateway that keeps running while git rewrites the tree loads new-tree modules into an old process on its next lazy import. So every gateway running from this install is stopped at the commit point, immediately before the first checkout move (``update_cmd_commit.arm_commit_point``), and the same set is restarted once the dependencies are synced and the launchers published, before the product builds (``source_completion.complete_source_checkout``'s ``before_build``). The set rides the Windows pause contract: one token in ``update_pause_record`` (identity = pid + start time; supervisor unit, or argv + home for bare processes), written before the first stop, resumed by ``_resume_windows_gateways_after_update`` (which dispatches here on ``platform == "posix"``), adopted by the next ``hermes update`` and resumed by the next launch when this process dies. * every gateway is drained first: the control socket's ``pause-for-update`` runs the same ``request_restart`` drain as the post-update restart's SIGUSR1 (refuse new turns, let the in-flight ones finish within ``agent.restart_after_turn_timeout``, then exit), waited for within that restart's budget. The request is recorded before it is sent, and carries no planned-stop marker (the gateway's marker watcher would take the immediate stop path instead). * systemd ``hermes-gateway*`` units and launchd ``ai.hermes.gateway*`` jobs are then stopped and later started through their supervisor, so it cannot respawn them on old code mid-update; one that does not take the drain gets its supervisor's stop at once. * bare ``gateway run`` processes that do not answer get marker + SIGTERM, and are replayed from their recorded argv + home. * the updater never kills itself: a unit whose cgroup holds this process (``/update`` from chat: the update runs in the gateway unit's cgroup, which ``systemctl stop`` kills) is stopped only after this process moved into a transient scope of its own. A launchd job whose coalition holds it, or any gateway whose stop would take the updater down, is left running for the post-commit restart. ) annotationsN)suppress)Path dict | None_RUNunitslabelspidsdict _RESTARTEDg^@>@c]tRt^8tRtRtR#) PauseRefusedzJThe pause could not be completed; everything it stopped was started again.N)__name__ __module__ __qualname____firstlineno____doc____static_attributes__rF/home/ubuntu/.hermes/hermes-agent/hermes_cli/update_cmd_posix_pause.pyrr8sTrrcV^8dQhRR/#returnr r)formats"r __annotate__r<s//d/rc RRRRR/R.R.R./#)platformposix resume_neededFprofiles unmapped_pidsunmapped posix_unitsrrrr _empty_tokenr&<s' %RZ\ M2 //rc V^8dQhRRRR/#)rno_gateway_restartboolrrr)rs"rrrAsTkrc"\PR8Xg V'dR#^RIHp\ 4pVP 4wr4VeVP /VCRR/CW44pRVR&Vs V# \dp\RT 24Rp?Ts T#Rp?ii;i) aHArm this run's pause (no stop yet) and adopt any set a killed update left paused. ``None`` on Windows (its own pause) and with ``--no-gateway-restart`` (the user manages gateways, nothing is stopped). An adopted orphan set is recorded under this process now, so even an update that never moves the checkout restarts it at exit.win32Nupdate_pause_recordr!Tr ru9 ⚠ Could not adopt gateways an earlier update paused: ) sysr hermes_clir-r& adopt_orphans record_pause Exceptionprintr)r( pause_recordtokenadoptedclaimsexcs$ r arm_pauser9As ||w"4> NEQ&446   --.N.N.NPW`E 'E*  D L Q I#OPP D LQs3A(( B3B  BcV^8dQhRR/#)rrr)r)rs"rrrWsRRRrc\\;'d3\PR4;'d\PR44#)zQTrue while this run holds gateways stopped (no gateway can answer a chat prompt). posix_stoppedr!)r)rgetrrrgateways_pausedr>Ws/ PP/2PPtxx7P QQrc$V^8dQhRRRRRR/#)r scope_cmdlistunitstrr str | Noner)rs"rrr^s!D rc l^RIHp\\\P 4;_uu_4V!VRVRR.,^ R7pVP ^8XdEVPP4'd%VPP4uuRRR4#RRR4R# +'giR#;i) _systemctlshowz--property=ControlGroup--valuetimeoutN) hermes_cli.update_cmd_fleetrHrOSError subprocessTimeoutExpired returncodestdoutstrip)r@rBrHresults&& r _unit_cgrouprU^s6 ':44 5 5I7PR[(\\fhi    !fmm&9&9&;&;==&&( 6 5 5  6 5 sAB"4B"" B3 c V^8dQhRRRR/#)rpidz int | strrrDr)rs"rrrgs  Y J rc@\\4;_uu_4\RV R24PRR7P 4p\ RV4R4p\ RV4R4pT;'g T;'gRuuRRR4# +'giR#;i) zThe systemd cgroup path of *pid*: the unified (v2) line, else the v1 ``name=systemd`` hierarchy, which systemd names units in identically. ``None``: not Linux or unreadable, which callers must treat as unknown, never as outside a unit./proc/z/cgrouputf-8encodingc3"TF4qPR4'gKVR,P4xK6 R#5i)z0:::NNN) startswithrS.0lines& r _pid_cgroup..ms*UUTooe>T(R((Us>>c3"TF5pRV9gK VPR^4^,P4xK7 R#5i)z:name=systemd::N)splitrSr`s& rrcrdns8cJZ^bJb4tzz#q)!,2244s ?/?N)rrNr read_text splitlinesnext)rWlinesunifiedlegacys& r _pid_cgrouprogs '  vcU'*+55w5GRRTUUUWYZccegh((&((D     sA&B ?B  B c$V^8dQhRRRRRR/#)rcgrouprD unit_cgrouprr)r)rs"rrrss&UUJUZUDUrcV'd V'gR#W8H;'g(VPVPR4R,4#)F/)r_rstrip)rqrrs&&r_insidervss7   T TF$5$5k6H6H6MPS6S$TTrcV^8dQhRR/#rr list[dict]r)rs"rrrys:rcL^RIHpHp^RIHpHpHpHpV!4'g.#\\4;_uu_4V!4RRR4.p\4pV!4Fwrp .p V!V PV PRR7V Fp V PR4'gKV!W4p V ^8:gW9gV!W4'gKAVPV 4VPRRR VR V R V R \W4R \!V 4/4K K V# +'giL;i)zGActive ``hermes-gateway*`` units of this install, one per live MainPID.)_ensure_user_systemd_envsupports_systemd_services)_for_each_systemd_gateway_unit_systemd_gateway_unit_listings_systemd_unit_owned_by_update_unit_main_pidNcR#Nr)namer8s&&r#_discover_systemd..sr) process_uniton_unit_timeouthermes-gatewaykindsystemdscoperBrWrqhome)hermes_cli.gatewayr{r|rMr}r~rrrr2setrRappendr_addrU_home_of)r{r|r}r~rrr seen_pidsrr@listingnamesrrWs r_discover_systemdrysV % & & )   " E%I%C%E!'&w~~ELL7M OD??#344 1Cax3+3PQZ3a3a MM#  LL&)WeVT5RU"L$A68TW=Z [ &F L#  s D D# cV^8dQhRR/#rxr)rs"rrrs:rc\PR8wd.#^RIHpHpHpHp^RIHpV!4p.pWS!\V4R7,FpV!V4eKV!V4wrVeV 'dV ^8:dK.V!4PV R24p VPRRRVR VR \V 4R \V 4R \\V 44/4K V#) zOLoaded ``ai.hermes.gateway*`` jobs of this install with a live process (macOS).darwin)_locate_launchd_gateway_serviceget_launchd_plist_path"launchd_gateway_labels_for_install!legacy_launchd_labels_for_install)launchd_label_foreign_home)excludez.plistrlaunchdlabeldomainplistrWr)r.rrrrrrhermes_cli.update_fleet_scoperr with_namerrCintr) rrrrrderivedjobsrrrWrs r_discover_launchdrs ||x I02G Dq &(22eWF3CD VY&'SVW\S]_dfijmfnXc#h/1 2S Krc V^8dQhRRRR/#rrWrrrDr)rs"rrrs#*rc^RIHp\\4;_uu_4V!V4;p'd\ V4uuRRR4#RRR4R# +'giR#;i)zKThe Hermes home *pid* serves (its control socket lives there), or ``None``.gateway_pid_homeN)rrrr2rC)rWrrs& rrrsJ> )  #C( (4 (t9   (     s A A c V^8dQhRRRR/#rr)rs"rrrsSZrc|^RIHpRV!V4;'g.9dR#^RIpVPV4P 4pTP R4'dR#\T4;p'dRT 2#TP RR4pTR 9dTPR 4'gR T 2#R# \ dR#i;i) znWhy *pid* answers to a supervisor other than a hermes unit/job (it would respawn it on old code), or ``None``.)_capture_gateway_argvz--external-supervisorNHERMES_S6_SUPERVISED_CHILDs6z systemd unit XPC_SERVICE_NAME0z application.z launchd job )rer) rrpsutilProcessenvironr2r=_respawning_servicer_)rWrrenvrBxpcs& r_supervisor_markersrs9#8#=#C#CD&nnS!))+ ww+,,"3''t'tf%% ''$c *C )CNN>$B$BcU##  s#B,, B;:B;c V^8dQhRRRR/#rr)rs"rrrsSZrc ^RIHpHp\V4;'gRpVP R^4R ,pVP R4'gR#RV9dRR.MR.p\ \\P4;_uu_4V!WT4pV^8d^W`\V439dMV!VR VR R .,^ R 7pVP;'gRP4R9dTMRuuRRR4#RRR4R# +'giR#;i)aThe systemd service that would respawn *pid* when it exits, or ``None``. systemd restarts a unit only when its main process exits, so the gateway must BE the MainPID or its direct child (a non-exec wrapper that exits with it). The inherited INVOCATION_ID is no evidence: a terminal tab's shell carries one (its .scope is never respawned), and so does anything deep under a CI runner's service.)rHrrertz.serviceNz/user@ systemctl--userrIz--property=RestartrJrK)reno) rMrHrrorsplitendswithrrNrOSubprocessError_ppidrRrS)rWrHrrqleafr@mainrestarts& rrrsG   # #F ==a  $D == $ $+3v+=h'K=I ':55 6 6i. !8eCj 11 fd,#&7T&J"KKG &'8g\`'a bD'#h$S)*3/1A#1Fd T/5N6(+gk;_:  NN/u4[\_[``ab c  UCdVSYNdehNijk =rc V^8dQhRRRR/#)rrWrrr)rs"rrrs  s s rc.\\\\4;_uu_4\ RV R24P RR7p\ VPR^4^,P4^,4uuRRR4# +'gi^#;i)rYz/statrZr[rN) rrN ValueError IndexErrorrrirrrh)rWstats& rrrsl ':z 2 2fSE'(22G2D4;;sA&q)//1!45 3 2 2 2 s AB B c V^8dQhRRRR/#)rrBr rr)r)rs"rrrs   $ rc8\4VPR4r!Ve V'gR#\W4'gR#^RIHp.\ P !4rTV^8dWPR,8wd\\V4VR,4'd\\4;_uu_4\RV R24P4PR R 4PR R 4pV\ P !48wdV!V4'd RRR4M&RRR4VPV4\V4pKVR ,R8XdRMRpRVRRRRRRR\ P !4 R2RRRR\\!V44.\#\V4ORNRNR VR!, R"2NR#Np\\\$P&4;_uu_4\$P(!VRRR R \$P*^RR$7RRR4\,P.!4R%,p \,P.!4V 8d:\4;p e\W4'gR#\,P0!R&4KSR# +'giELi;i +'giL;i)'a'Move this process (and its ancestors in the same unit, e.g. ``/update``'s bash wrapper) into a transient scope so stopping *unit* cannot kill it. True only when proven out: a membership that cannot be read (either side) keeps the unit running, since its stop would take this update down with it.rqNFT)looks_like_gateway_command_linerWrYz/cmdline rZreplaceruserrz--systembusctlcallzorg.freedesktop.systemd1z/org/freedesktop/systemd1z org.freedesktop.systemd1.ManagerStartTransientUnitzssa(sv)a(sa(sv))zhermes-update-z.scopefail2PIDsau Descriptionszhermes update (outside rBz while it is paused)rcapture_outputtextr\errorsstdinrLcheckg@g?)ror=rvrrosgetpidrrNr read_bytesrdecoderrrClenmaprOrrunDEVNULLtime monotonicsleep) rBownrrrr rWcmdlinebuscmddeadlinenows & r_escape_cgroupr s' #}dhhx&8 {+ 3 $ $>BIIK# 'c%[(W[5EtH~-V-V g  VC512==?GGtT[[\cenoGbiik!&Eg&N&N   CCj7mv-(:C S&"<>Y -/CEW BIIK= /fdCPSTXPYN b]`adfj]k b  b b"9fFZ[ b^a bC ':55 6 6s4dWU^#++Ru F 7~~#%H .. X %= C -gc6O6O 3 '   7 6s%A%I5 I5-J 5 J J c$V^8dQhRRRRRR/#)rrryrrz(tuple[list[dict], list[dict], list[str]]r)rs"rrr#s"***:*:b*rc^RIHp...rTpVFup\VR,4fVPRVR, R24K5\ V4'dVPV4KYVPRVR, R24Kw VFrpV!VR,4'g,\ P PR4VR ,8XdVPRVR , R 24KaVPV4Kt W4V3#) zDDrop the runtimes whose stop would kill this updater; name each one.)#_is_pid_ancestor_of_current_processru ↷ rBz^ keeps running until after the update (stopping a system unit needs root or passwordless sudo)zf keeps running until after the update (this update could not prove it runs outside that unit's cgroup)rWrrzO keeps running until after the update (this update runs inside its launchd job))rr  _manage_cmdrr rrr=)rrr  keep_units keep_jobsrrBjobs&& r _left_runningr#sF%'R7J tG} % - NNVDL>2CC D D ! !   d # NNVDL>2KK L /E ; ;::>>"45WE NNVCL>255 6   S !  ' ))rc V^8dQhRRRR/#)rrrCrz list | Noner)rs"rrr<ss{rc^RIHpHpVR8XdRR.MR.pVR.,pVR8Xd0V!V4'd"RR.VOpV!V.R O,4'dV#R #V#) rF) _needs_sudo_sudo_noninteractive_okrrrz--no-ask-passwordsystemsudoz-nN)rIz-pIdr)rMrr)rrrrrs& rr r <siP%*f_; !;-C $% %C [//#s#.s5[/[\\tfbff Jrc V^8dQhRRRR/#)rrBr rNoner)rs"rrrFseeTederc:^RIHp\VR,4pVf\RVR, 24hV!VRVR,.,\R7pVP ^8wd;\RVR, R VP ;'gR P4 24hR#) rFrGrNz!no privilege to stop system unit rBstoprKzsystemctl stop failed: re)rMrHr r_STOP_TIMEOUT_SrQstderrrS)rBrHrrTs& r _stop_unitr Fs6 d7m $C {>tF|nMNN vtF|44o NF A_T&\N)V]]EXEXVXD_D_DaCbcddrc V^8dQhRRRR/#)rrBr rrr)rs"rrrPs33d3s3rc,^RIHpHpHpHp\ VR,4pVf\ RVR, 24hV!VR,VR,4V!VRVR,.,^ R7V!VRVR,.,\R7pVR,R8XdR R .MR .pVP^8wgV!WpR,R R7'gD\ VR, R VP;'gR P4;'gR 24hV!WpR,4#)zStart *unit* through systemd; the new MainPID (raises when it does not come up). The unit gets the definition repair the restart it replaces ran first (``_restart_one_systemd_gateway_unit``).)$_repair_unit_without_fatal_exit_parkrHr_wait_for_service_activerz"no privilege to start system unit rBz reset-failedrKstartrrr.@z did not start: rez not active) rMr#rHrr$r  RuntimeErrorrrQrrS)rBr#rHrr$rrTr@s& r _start_unitr(Ps d7m $C {?V ~NOO(ftG}Esnd6l33R@ wV 55 OF+/=F+Bh' I A%=if_c%d%dd6l^+;V]]=P=Pbrrr2)rWrr>rs&& r_aliver?s\B3s8$$% : G G/C9dBZGG sA AAc$V^8dQhRRRRRR/#)rrLfloatbornrrrr)rs"rrrs!e;(rc `\P!4V,p\V4pV'dz\P!4V8d`VUu0uF1p\YR;'g/P V44'gK/VkK3 ppV'gKi\P !R4KV#uupi)皙?)rrrr?r=r)r rLrBrliverWs&&& rr1r1sv~~')H t9D 4>>#h.#JtvcJJB3C3CC3H'ItJ 4 JJsO KKsB+!B+?B+c V^8dQhRRRR/#)rentryr rr)r)rs"rrrsMMM$Mrc^RIHpVPR4'gR#V!\VR,44p\ T;'d+TPR4;'gTPR44# \dR#i;i)auAsk the gateway over its control socket to finish its in-flight turns and exit; True on its ACK. That is the ``request_restart`` drain the post-update restart uses (SIGUSR1); SIGTERM, ``systemctl stop`` and ``launchctl bootout`` enter ``stop()`` at once and cut a turn off. No planned-stop marker on this path: the gateway's marker watcher takes that same immediate signal-stop path, and could fire before the request lands. The caller records the request (``mark_stop_sent``) BEFORE sending it, so a request that reaches the gateway after this update died still leaves the evidence that keeps the drained gateway's restart debt.)pause_gateway_for_updaterFpausingalready_stopping)gateway.control_socketrIr=rr2r))rGrIacks& r _ask_to_drainrNsv@ 99V  &tE&M':; KK+JJsww7I/J LL sA66 BBc V^8dQhRRRR/#rrGr rrr)rs"rrrs 2 2 2$ 2rc^RIHpVPR4R8Xd\V4R#VPR4R8Xd\ V4R#V!\ VR,4VR,4\ \\4;_uu_4\P!VR,\P4RRR4R# +'giR#;i)z\The stop for a gateway that did not take the drain: its supervisor's, else marker + SIGTERM.r,rrrrrWN) r0r-r=r r2rrProcessLookupErrorPermissionErrorrkillsignalSIGTERM)rGr-s& r _stop_at_oncerWsO yyI%5 6 i '%)$uV}*=uU|L (/ : : GGE%L&.. 1; : : :s -C C cV^8dQhRR/#)rrrAr)rs"rrrs##u#rc^RIHpV!4#)z_The post-update restart's wait for a drained gateway (after-turn wait + stop drain + headroom)._gateway_drain_budget)rMr[rZs r _drain_budgetr\sA ""rc$V^8dQhRRRRRR/#)rr5r entriesryrrr)rs"rrrs&%d%d$%d%d%drc |^RIHp^RIHp\ 4.rTVFtpVP WR,4\ V4'dVPV4K?\V4VPR4'dKcVPV4Kv V'gR#\RVR R\V4 R 24VUu.uF5qs!VPR 4'd\VR ,4MRVR 7NK7 pp\P!4V,p V'd\P!4V 8dVUu.uF/p\VR,VPR 44'dK-VNK1 upFpVP!V4\#V4K! V'dTMRF p V !4K \P$!R 4KVFpVPR4'd\V4K'\'\(\*4;_uu_4\,P.!VR,\1\2R\2P444RRR4K VUu.uFqwPR4'dKVNK p p\7V Uu.uF qwR,NK upRV Uu/uFqwR,VPR 4bK up4;p 'd@\9RRP;\=\>\AV 444,R,4hR#uupiuupi +'giEKb;iuupiuupiuupi)aDrain every gateway (bare ones SIGTERMed when they do not answer), then wait for all of them within one drain budget. A supervised gateway that exited through the drain is stopped through its supervisor at once: its exit 75 asks for a restart, which systemd honours after ``RestartSec=5`` and launchd immediately, still on the old tree. Past the budget a supervised survivor gets its supervisor's stop and a bare one SIGKILL, as the restart path escalates.r,)drain_progress_reporterrWrNu ⏳ Waiting up to z.0fzs for in-flight turns on z to finish before the update...r)budget_srrDSIGKILLg$@zgateway PID(s) , z did not stopr)!r/r-"hermes_cli.update_cmd_drain_reportr`r\mark_stop_sentrNrrWr=r3_namesrrrr?remove_disarmrrrRrSrrTgetattrrUrVr1rjoinrrCsorted) r5r^r4r`budgetwaitingrGeticksrtickrlefts && r_stop_gatewaysrrsa ?J#orG##E<8    NN5 ! % 99V$$u%   ,EfWoEVVu vwkr skrfg $f T!F)_4Z` akrE s~~&(H dnn&1!(NAqxt0MaaNE NN5 ! ENO%E",D F- 3 99V   % ,o>>e gfi&PQ?>   4w!eeFmAAwD 4T2TU88T2DZ^:_Z^UVU8QUU4[;PZ^:_``t`,tyyS&,9O/PPSbbcca! tO?>> 42:_s6<;L;*L*L .s?c[bVWUU6]QQaeeGnQQ,qxj8QQ[bsA A A )rj)r^s&rrfrfs 99c[bc ccrc$V^8dQhRRRRRR/#)rrrCrrrz int | Noner)rs"rrrs& A A3 A Aj Arc^RIHpV!\V44pT'd\ T4T9d \ T4#R# \dR#i;i)aThe gateway serving *home*, by the canonical identity reader (PID file + runtime lock, then the runtime status record), never a process scan: a replayed ``gateway run`` matches the process matcher for the moment before it finds the home already served and exits.)live_gateway_pid_for_homeN)rr}rr2r)rrr}rWs&& r_gateway_on_homer~sL9'T 3s3xw63s8@D@ s> A  A c V^8dQhRRRR/#)rrrCrr r)rs"rrrs P Pc Pd Prc^RIHpHp^RIHpHp\ V4P4pWQ!4P48wd'WR!4P48XdV!RRRV/R7#V!VRR7#)aThe environment a replayed gateway starts with, as the restart watcher it replaces chose it (``_spawn_gateway_restart_watcher``): a named profile's gateway, replayed by an update launched under that same profile, inherits the launcher's environment (an exported bot token may be its only credential). The multiplexing root, and any home other than the launcher's, get that home's own secrets and none of the launcher's (``served_profile_child_env``).)get_default_hermes_rootget_routing_process_hermes_home)build_subprocess_envserved_profile_child_envF HERMES_HOME) scrub_secretsinherit_profile_homeextraT) target_homeinherit_credentials)hermes_constantsrrtools.environments.localrrrresolve)rrrrrtargets& r _replay_envrsi ZW $Z   !F (*2244CbCdClClCn9n#%e\iko[pqq #$ OOrc V^8dQhRRRR/#)rrGr rrr)rs"rrrspp$p3prc \VR,\VR,404;peV#\VR,4R, pVPRRR7\ VR,4pRFpVP VR4K \ VR, R4;_uu_4p\P!\VR ,4V\PV\PRRR 7pRRR4\P!4\,p\P!4V8d\VR,\VR,404;peV#XP4e.\!R VR, R VP" R VR, R24h\P$!R4K\!R VR, R\\4 R24h +'giL;i)zReplay a bare gateway from its recorded argv under its recorded home; the new PID once it is serving. A gateway already serving that home (an earlier attempt's) is adopted instead.rrWNlogsT)parentsexist_okzgateway-update-resume.logabr)rrrRrstart_new_session close_fdszreplayed gateway for z exited z (see rg?z did not come up within r) INVOCATION_IDJOURNAL_STREAMHERMES_UPDATE_PAUSED_HERMES_GATEWAY)r~rrmkdirrpopopenrOPopenrArSTDOUTrr_READY_TIMEOUT_Spollr'rQr) rGexistingrrleakedlogprocrrWs & r_relaunch_barers%U6]Su5F4GHHU f  'DJJtdJ+ eFm $C` a d00$ 7 73U6] 3JDVDV_b'1'8'8D\`b 8~~"22H .. X %#E&MCe 4E3FG GC TJ 99; "!6uV}oXdooM^_''+.I'I&J! MN N 3 .uV}o=UVYZjVkUllmn oo 8 7s A G G" c V^8dQhRRRR/#)rr5r rrCr)rs"rrr sTcrcVPR4;'g.Uu.uF,qPR4;'gVPR4NK. ppY PR4;'g.Uu.uF'qPR4'gKRVR, 2NK) up, pRPV4#uupiuupi)r%rBrr$rrxrWrc)r=rj)r5urs& r _describer s6;ii 6N6T6TRT6T U6TUU6] , ,aeeGn ,6TE U :1F1L1L"1L ^1LAPUPUV\P]' QuXJ'1L ^^E 99U  V ^sB6B6/B; B;cV^8dQhRR/#)rrrDr)rs"rrr(s ( (z (rcV\pVfR#RV9d\V4VR&VR,#)aStop this install's gateways before the first checkout move; once per run. Never refuses the update: the pause is protection, not a gate. When it cannot complete (discovery or the durable record fails, a stop fails), the gateways run through the swap and the post-commit restart refreshes them, which is how POSIX updated before the pause existed. Always ``None``; the caller's refusal branch stays for the contract it shares with Windows.Nposix_pause_verdict)r_pause)r5s rpause_at_commit_pointr(s4 E }E)'-e}#$ & ''rc V^8dQhRRRR/#)rr5r rrDr)rs"rrr7s::$:::rc^RIHp\4\4r2\ W#4wr#p\ VUu0uF qUR,kK upVUu0uF qfR,kK up,4wrxYH,Fp \T 4K Y#,p T 'g T'gR#^RIH p ^RI H p \P!T4pY,FpT !TR,4TR&K .TPR .4OT OTR &.TPR .4OTOTR &.TPR .4OR T4OTR &/TPR /4CY,Uu/uF3p\TR,4TP!TR,4R,bK5 upCTR &RTR&TP#TR.4TP%Y T,Uu.uF pTR,NK upY,Uu/uF<pTPR4'gKTR,T !\'TR,44bK> upR7RTR&\1Y T,4TP9T4\R\;T4 R24R#uupiuupi \ dp \RT R24Rp ? R#Rp ? ii;iuupiuupiuupi \ dp TPR4TPR48wdF\)\ 4;_uu_4TP+T4RRR4M +'giM;iTP-4TP/T4\RT R24Rp ? R#Rp ? ii;i \ dp \RT R24\3T4M' \ dp\RT R24Rp?MRp?ii;i\4P74FpTP-4K TP9T4Rp ? R#Rp ? ii;i)rFr,rWu+ ⚠ Could not list the running gateways (z7); they keep running and are restarted after the updateNr)_planned_stop_marker_pathrr%r$r#c32"TF qR,xK R#5i)rWNrrys& rrc_pause..Ls@XSWa5SWs identitiesTr!r)markerspause_idu. ⚠ Could not record the gateways to pause (r<u% ⚠ Could not pause every gateway (z:); restarting them, the update continues without the pauseu ⚠ z/; they restart once the dependencies are syncedu ⏸ Paused for the update: z- (restarted once the dependencies are synced))r/r-rrrrr2r3rrr0rcopydeepcopyr=rCidentityr1mark_stop_requestedrr dischargeclearupdaterrresume_paused_setr valuessyncr)r5r4rrrrjrmorer8rb supervisedrrbeforerGrnagain restarteds& rrr7s>')+<+>t,U9W#u$=u!uXXu$=SW@XSWa5SW@X$XY  d J d5G ]]5 !F"",U5\:d #GUYY}b9GJGE-;%))J3;d;E*Yuyy"=Y@XSW@XYE/pUYY|R8p\f\mn\mWXc!E(m\-B-B1U8-LT-RR\mnpE,!E/!!%r2(( 4&78&7AeHH&78S]SdvSdahihmhmnthuIQuX8aiIISdv ) x"E/ u4/0e ))E*:);;h ij k%>@X  ;C5@wxy$o 9v   99Z FJJz$: :)$$&&u-%$$$  V >seCz{|    5cU:tuv S e $ S F5'!PQ R R S#**,I OO -%  s*I3I)I3 I. I39J)J*>J  J*J% :$J% J*,M) I33 J>JJ J** M5*:!E&M?(KLV}   T "   12 6 = =d C0&-I:&,,",yy   $$U+G v  " "7 +   |E%L>wiH I -"*09: 1hh :E/E* 3DIIi4H3IJKE RZH[[\\"E/"E// % OOCH %   e $ $ % $ OOCH %   U # # $;s7CH AIJ I+I  I J+JJcV^8dQhRR/#rr)rs"rrrs4rc\#)zXWhat the early resume restarted on the new code in this process (for the fleet restart).)r rrralready_restartedrs r)selfr):__conditional_annotations__r __future__rrrrUrOr.r contextlibrpathlibrr__annotations__rr rrr'rr&r9r>rUrorvrrrrrrrr rr r r(r2r;r?r1rNrWr\rrrhrfr~rrrrrrr)rs@rrs9B#  kSUHceVSUC DCU<U/ ,R U :.,*0  F*2e3$ hE h NHM$ 2# %dP'd A Pp0 (:~(#Vr