Default Pimox workers back to DHCP networking

README.md

@@ -727,14 +727,13 @@ Worker indexes are stable. Index `1` maps to VMID `9010`, node name
 so on. `LAB_PIMOX_SKIP_WORKER_INDEXES=1` leaves the first slot unmanaged while
 allowing higher indexes to be automated.
 
-Worker networking is deterministic by default. After cloning or reusing a VM,
-the pipeline uses qemu-guest-agent to write a static `/etc/network/interfaces`
-entry before waiting for SSH, so a missing DHCP lease cannot stall the build.
-Defaults are `LAB_PIMOX_WORKER_NETWORK_MODE=static`,
+Worker networking uses DHCP by default to preserve the known-good Pimox bridge
+path. Set `LAB_PIMOX_WORKER_NETWORK_MODE=static` only when DHCP reservations are
+not available and you explicitly want Jeannie to configure guest static IPs with
+qemu-guest-agent. Static mode uses
 `LAB_PIMOX_WORKER_STATIC_IPS="192.168.100.66 192.168.100.67 192.168.100.76"`,
 `LAB_PIMOX_WORKER_GATEWAY=192.168.100.1`, and
-`LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`. Set
-`LAB_PIMOX_WORKER_NETWORK_MODE=dhcp` to restore the old DHCP-only behavior.
+`LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`.
 Worker creation fails faster than template creation: static guest-network
 configuration waits up to `LAB_PIMOX_GUEST_AGENT_CONFIG_TIMEOUT_SECONDS=120`,
 worker IP/SSH readiness waits up to `LAB_PIMOX_WORKER_WAIT_TIMEOUT_SECONDS=600`,

README.md.tmpl

@@ -727,14 +727,13 @@ Worker indexes are stable. Index `1` maps to VMID `9010`, node name
 so on. `LAB_PIMOX_SKIP_WORKER_INDEXES=1` leaves the first slot unmanaged while
 allowing higher indexes to be automated.
 
-Worker networking is deterministic by default. After cloning or reusing a VM,
-the pipeline uses qemu-guest-agent to write a static `/etc/network/interfaces`
-entry before waiting for SSH, so a missing DHCP lease cannot stall the build.
-Defaults are `LAB_PIMOX_WORKER_NETWORK_MODE=static`,
+Worker networking uses DHCP by default to preserve the known-good Pimox bridge
+path. Set `LAB_PIMOX_WORKER_NETWORK_MODE=static` only when DHCP reservations are
+not available and you explicitly want Jeannie to configure guest static IPs with
+qemu-guest-agent. Static mode uses
 `LAB_PIMOX_WORKER_STATIC_IPS="192.168.100.66 192.168.100.67 192.168.100.76"`,
 `LAB_PIMOX_WORKER_GATEWAY=192.168.100.1`, and
-`LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`. Set
-`LAB_PIMOX_WORKER_NETWORK_MODE=dhcp` to restore the old DHCP-only behavior.
+`LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`.
 Worker creation fails faster than template creation: static guest-network
 configuration waits up to `LAB_PIMOX_GUEST_AGENT_CONFIG_TIMEOUT_SECONDS=120`,
 worker IP/SSH readiness waits up to `LAB_PIMOX_WORKER_WAIT_TIMEOUT_SECONDS=600`,

bootstrap/provisioning/README.md

@@ -142,14 +142,12 @@ destroy and recreate existing worker VMs from the current template. The pipeline
 refuses `LAB_PIMOX_WORKER_STORAGE=local` so only the template VM lives on local
 storage.
 
-Worker networking defaults to deterministic static LAN addresses configured
-through qemu-guest-agent after clone/start and before SSH wait:
-`LAB_PIMOX_WORKER_NETWORK_MODE=static`,
+Worker networking defaults to DHCP to preserve the known-good Pimox bridge path:
+`LAB_PIMOX_WORKER_NETWORK_MODE=dhcp`. Static mode is still available when DHCP
+reservations are not available: set `LAB_PIMOX_WORKER_NETWORK_MODE=static`,
 `LAB_PIMOX_WORKER_STATIC_IPS="192.168.100.66 192.168.100.67 192.168.100.76"`,
 `LAB_PIMOX_WORKER_GATEWAY=192.168.100.1`, `LAB_PIMOX_WORKER_INTERFACE=enp0s18`,
-and `LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`. Use
-`LAB_PIMOX_WORKER_NETWORK_MODE=dhcp` only when DHCP reservations are known to be
-stable.
+and `LAB_PIMOX_WORKER_DNS_SERVERS="192.168.100.89 1.1.1.1"`.
 Worker clone failures are intentionally faster than template build failures:
 `LAB_PIMOX_GUEST_AGENT_CONFIG_TIMEOUT_SECONDS=120`,
 `LAB_PIMOX_WORKER_WAIT_TIMEOUT_SECONDS=600`, and

docs/jeannie.1.md

@@ -539,6 +539,10 @@ falls back to the first Lynis warning, then the first suggestion.
 : Desired Pimox worker count when not supplied by generated cluster topology
 state.
 
+`LAB_PIMOX_WORKER_NETWORK_MODE`
+: Worker guest networking mode. Defaults to `dhcp`. Set to `static` only when
+Jeannie should configure guest static IPs through qemu-guest-agent.
+
 `LAB_PIMOX_WORKER_REQUIRE_LAN_PROBE`
 : Require the worker guest to ping the LAN gateway and probe host after static
 IP assignment. Defaults to `true`.

jeannie

@@ -1051,6 +1051,26 @@ echo 'Pimox VM ${vmid} guest-agent network-get-interfaces:'
 sudo '${qm_bin}' guest cmd '${vmid}' network-get-interfaces" >&2 || true
 }
 
+pimox_guest_exec_exitcode() {
+    local guest_exec_json="$1"
+
+    GUEST_EXEC_JSON="${guest_exec_json}" python3 - <<'PY'
+import json
+import os
+import sys
+
+try:
+    document = json.loads(os.environ.get("GUEST_EXEC_JSON", ""))
+except Exception:
+    sys.exit(2)
+
+try:
+    print(int(document.get("exitcode", 1)))
+except Exception:
+    sys.exit(2)
+PY
+}
+
 wait_for_pimox_guest_ssh() {
     local host="$1"
     local user="$2"
@@ -1214,6 +1234,7 @@ configure_pimox_worker_guest_network() {
     local script
     local encoded_script
     local network_output
+    local network_exitcode
     local worker_address
     local lan_probe_host="${LAB_PIMOX_WORKER_LAN_PROBE_HOST:-${LAB_DEBIAN_LAN_IP:-}}"
     local require_lan_probe="${LAB_PIMOX_WORKER_REQUIRE_LAN_PROBE:-true}"
@@ -1308,7 +1329,14 @@ EOF
     while ((SECONDS < deadline)); do
         if network_output="$(pimox_ssh "${pimox_host}" "${pimox_user}" "${pimox_key}" "sudo '${qm_bin}' guest exec '${vmid}' -- bash -lc \"printf '%s' '${encoded_script}' | base64 -d | sudo bash\"" 2>&1)"; then
             printf '%s\n' "${network_output}" | sed 's/^/    guest-network: /'
-            return 0
+            if network_exitcode="$(pimox_guest_exec_exitcode "${network_output}")" && [[ "${network_exitcode}" == "0" ]]; then
+                return 0
+            fi
+            echo "Static guest network validation failed inside Pimox VM ${vmid}." >&2
+            if [[ -n "${network_exitcode:-}" ]]; then
+                echo "Guest exit code: ${network_exitcode}" >&2
+            fi
+            return 1
         fi
         sleep 5
     done
@@ -1397,7 +1425,7 @@ ensure_pimox_worker_node() {
     local worker_storage="${19}"
     local worker_replace_existing="${20}"
     local worker_cpu_affinity="${21}"
-    local worker_network_mode="${LAB_PIMOX_WORKER_NETWORK_MODE:-static}"
+    local worker_network_mode="${LAB_PIMOX_WORKER_NETWORK_MODE:-dhcp}"
     local worker_static_ips="${LAB_PIMOX_WORKER_STATIC_IPS:-192.168.100.66 192.168.100.67 192.168.100.76}"
     local worker_gateway="${LAB_PIMOX_WORKER_GATEWAY:-192.168.100.1}"
     local worker_dns_servers="${LAB_PIMOX_WORKER_DNS_SERVERS:-192.168.100.89 1.1.1.1}"

tests/jeannie-unit

@@ -330,6 +330,12 @@ test_pimox_worker_static_ip_selection() {
     fi
 }
 
+test_pimox_guest_exec_exitcode_parser() {
+    assert_eq "pimox guest exec parses success exitcode" "0" "$(pimox_guest_exec_exitcode '{"exitcode":0,"exited":1}')"
+    assert_eq "pimox guest exec parses failed exitcode" "1" "$(pimox_guest_exec_exitcode '{"exitcode":1,"exited":1}')"
+    assert_command_failure_contains "pimox guest exec rejects invalid JSON" "" pimox_guest_exec_exitcode 'not json'
+}
+
 test_pimox_worker_count_detects_generated_topology
 test_configured_worker_count_cannot_hide_existing_workers
 test_configured_worker_count_can_expand_target_set
@@ -343,6 +349,7 @@ test_inventory_validator_accepts_current_inventory
 test_inventory_validator_rejects_ip_drift
 test_heal_cooldown_escalates_to_diagnostics
 test_pimox_worker_static_ip_selection
+test_pimox_guest_exec_exitcode_parser
 
 if ((FAILURES > 0)); then
     printf '\n%s unit test(s) failed.\n' "${FAILURES}" >&2