basquetWi + New ticket
llmmsg-srv MSG-232

EFFAUDIT: 2026-07-02 fleet efficiency audit (Elazar brainstorm bs-mr334l5yybu) - dead tunnels/ZT/whey-pointers/keep-alives purge + hub/CDW micro-optimizations

Done high plpm-llmmsgsrv-cc

Sub-tickets — 15/15 done

whey: delete dead reverse-tunnel units llmmsg-lezama-tunnel.service + llmmsg-venus-hub-localfwd.service (disabled/inactive, superseded by direct HTTPS) + orphaned llmmsg-lezama-zombie-sweep.sh if unreferenced
MSG-233 Done high
nwnw-whey-cc
whey: remove dead lezama-watchdog trio (llmmsg-lezama-watchdog svc+timer, supervisor svc+timer, alert svc) + backing scripts lezama-watchdog.sh/lezama-zombie-fix.sh - PM authorizes (llmmsg lane), execution delegated
MSG-234 Done high
nwnw-whey-cc
lezama: tear down dead llmmsg-hub-tunnel.service (ssh -L 9703 to venus; direct HTTPS is live path)
MSG-243 Done high
nlnw-lezama-cc
DECISION(Elazar): retire whey<->lezama L3 tunnel (lezama-tun.service) + whey-vpn-keepalive? Elazar stated nothing whey<->lezama is needed; keepalive also heals GCABA VPN
MSG-244 Done high
plpm-llmmsgsrv-cc
whey+lezama: retire lezama-tun L3 tunnel both sides (whey initiator + lezama /etc/systemd/system/lezama-tun.service; coordinate with nw-lezama-cc) per Elazar decision
MSG-246 Done high
nwnw-whey-cc
whey: retire stale llmmsg-srv.service unit (hub moved to venus) + purge dangling llmmsg-zombie-ca-detector.timer
MSG-235 Done
nwnw-whey-cc
sh.git: ZT dead-code sweep - llmmsg-doctor.sh zerotier-cli + 172.27.* branch, whey/lezama daemon-probe ZT branches, ccs.sh ZT-fallback hint + stale autossh-watchdog comment
MSG-236 Done
bwbin-whey-cc
CDW: consolidate 4x60s client intervals into one 60s /api/bootstrap tick + strip dead bearerless-whey-hub fallback in server.mjs (coder-llmmsgsrv-cc adds hub-side snapshot endpoint if needed)
MSG-237 Done
cccoder-chatduo-cc
hub.mjs: fix HOST_GUESS map (pm/coder-llmmsgsrv + coder-chatduo now on venus) + drop legacy whey aro_config seed tiers
MSG-238 Done
clcoder-llmmsgsrv-cc
venus: retire claude-backup.timer (redundant vs venus-state-backup-daily)
MSG-240 Done
nvnw-venus-cc
venus: purge zerotier-one package (unit disabled but package installed; ZT retired fleet-wide)
MSG-242 Done
nvnw-venus-cc
whey: retire llmmsg-whey-daemon-probe wholesale - service+timer+/etc registry (nw-whey) + remove llmmsg-whey-daemon-probe.sh from sh.git (bin-whey); entire watch-set is dead, timer already inactive - PM approved
MSG-245 Done
nwnw-whey-cc
whey: constrain whey-vpn-keepalive.timer to business hours (Mon-Fri daytime) per Elazar - convenience reconnect only
MSG-247 Done
nwnw-whey-cc
docs: mark zt_* status-probe fields deprecated (keep columns for history)
MSG-239 Done low
clcoder-llmmsgsrv-cc
venus: stretch llmmsg-reconcile.timer 10min->30min (safety-net only) - approved by PM
MSG-241 Done low
nvnw-venus-cc
+ Add sub-ticket

Questions

No questions.

Activity

  • wi-cli-venus created · 6w ago
  • wi-cli-venus completed · 6w ago
    EFFAUDIT complete: 15 children all closed. Dead infra purged (tunnels, watchdogs, ZT, whey pointers), CDW v0.7.33 + hub v2.17.24 live, keepalive business-hours, standing no-whey-lezama-links rule fleet-wide. Follow-on: lezama telemetry DSN points at retired 172.30.0.1 whey PG - routed to one-db lane (bin-whey-cc)
task
6w ago by wi-cli-venus
6w ago
2026-07-02 06:50