fix(gateway): persist git-gate state across gateway restarts
tracker-policy-pr / check-pr (pull_request) Successful in 12s
lint / lint (push) Successful in 59s
test / unit (pull_request) Successful in 53s
test / integration-docker (pull_request) Successful in 1m7s
test / coverage (pull_request) Successful in 19s

Per-bottle git-gate state (bare repos under /git/<id>, deploy creds under
/git-gate/creds/<id>) was provisioned once at bottle launch and lived only
in the gateway's ephemeral storage. A gateway rebuild/restart wiped it and
nothing re-provisioned already-running bottles, so their agents 404'd on
fetch/push. Same class of bug as the CA (#510); the orchestrator restores
only egress tokens, not git-gate declarations.

Persist the state on both backends, mirroring the CA-persistence approach:

- firecracker: attach a second persistent data drive (/dev/vdc) to the
  gateway VM and bind-mount its git/ + creds/ subdirs onto /git and
  /git-gate/creds in the gateway guest init, before the data plane starts.
  Generalize the VM config to a stable-ordered data_drives tuple (CA=vdb,
  git=vdc; orchestrator registry stays vdb).
- docker: bind-mount host dirs (host_gateway_git_dir / creds_dir, under the
  never-pruned app-data root) onto /git and /git-gate/creds, with
  BOT_BOTTLE_DOCKER_GIT_MOUNT / _CREDS_MOUNT env overrides so CI isolates
  them to per-run volumes it cleans up.

Teardown already rm -rf's /git/<id> + creds, so the persistent store
self-cleans over the normal lifecycle.

Closes #512

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit is contained in:
2026-07-26 17:34:22 -04:00
parent 0ddaa95c14
commit dff811f190
13 changed files with 213 additions and 27 deletions
+24 -3
View File
@@ -63,6 +63,15 @@ _GUEST_GATEWAY_JWT_PATH = "/var/lib/bot-bottle/gateway-jwt"
# survives a gateway-VM rebuild; without it every rebuild mints a fresh CA that
# already-running bottles distrust, breaking the TLS handshake (issue #450).
_GATEWAY_CA_MOUNT = "/home/mitmproxy/.mitmproxy"
# The gateway VM's persistent git-gate volume staging mount (/dev/vdc). The
# gateway boots with this volume (see `FirecrackerGateway._ensure_git_volume`)
# and the init bind-mounts its `git/` + `creds/` subdirs onto the load-bearing
# `/git` and `/git-gate/creds` paths, so per-bottle bare repos + deploy creds
# survive a gateway-VM rebuild; without it a restart drops every already-running
# bottle's git-gate state and its agent 404s on fetch/push (issue #512).
_GATEWAY_GIT_MOUNT = "/var/lib/bot-bottle-gitgate"
_GATEWAY_GIT_REPO_ROOT = "/git"
_GATEWAY_GIT_CREDS_DIR = "/git-gate/creds"
# The two per-plane rootfs source images. The orchestrator VM boots a control
# plane + buildah rootfs (Dockerfile.orchestrator.fc, FROM orchestrator); the
@@ -198,11 +207,12 @@ def boot_vm(
run_dir: Path,
role: str,
mem_mib: int,
data_drive: Path | None = None,
data_drives: tuple[Path, ...] = (),
extra_boot_args: str = "",
) -> InfraVm:
"""Boot the `role` infra VM from its per-plane rootfs on `slot`'s link.
Records the PID."""
Records the PID. `data_drives` are attached as /dev/vdb, /dev/vdc, ... in
order, so callers must keep the order stable (see `firecracker_vm._config`)."""
if not netpool.tap_present(slot.iface):
die(f"infra link {slot.iface} not present.\n"
f" ./cli.py backend setup --backend=firecracker")
@@ -224,7 +234,7 @@ def boot_vm(
name=name, rootfs=rootfs, tap=slot.iface,
guest_ip=slot.guest_ip, host_ip=slot.host_ip, pubkey=pubkey,
run_dir=run_dir, mem_mib=mem_mib, detached=True,
data_drive=data_drive, extra_boot_args=boot_args,
data_drives=data_drives, extra_boot_args=boot_args,
)
_pid_file(run_dir).write_text(str(vm.process.pid))
return InfraVm(guest_ip=slot.guest_ip, private_key=private_key, vm=vm)
@@ -461,6 +471,17 @@ def _gateway_init() -> str:
# mitmproxy) starts.
mkdir -p {_GATEWAY_CA_MOUNT}
mount -t ext4 /dev/vdb {_GATEWAY_CA_MOUNT} 2>/dev/null || true
# Persistent git-gate volume (/dev/vdc): its git/ + creds/ subdirs are
# bind-mounted onto the load-bearing /git and /git-gate/creds so per-bottle bare
# repos + deploy creds survive a gateway-VM rebuild (issue #512). On first boot
# the volume is empty; the subdirs are created here. Must mount BEFORE the data
# plane (hence git-http) starts, and before any per-bottle provisioning writes.
mkdir -p {_GATEWAY_GIT_MOUNT}
mount -t ext4 /dev/vdc {_GATEWAY_GIT_MOUNT} 2>/dev/null || true
mkdir -p {_GATEWAY_GIT_MOUNT}/git {_GATEWAY_GIT_MOUNT}/creds
mkdir -p {_GATEWAY_GIT_REPO_ROOT} {_GATEWAY_GIT_CREDS_DIR}
mount --bind {_GATEWAY_GIT_MOUNT}/git {_GATEWAY_GIT_REPO_ROOT} 2>/dev/null || true
mount --bind {_GATEWAY_GIT_MOUNT}/creds {_GATEWAY_GIT_CREDS_DIR} 2>/dev/null || true
ORCH=$(sed -n 's/.*bb_orch=\\([^ ]*\\).*/\\1/p' /proc/cmdline)
GW_JWT=""
i=0