diff --git a/test/integration/mesh.test.ts b/test/integration/mesh.test.ts new file mode 100644 index 0000000..b98cc78 --- /dev/null +++ b/test/integration/mesh.test.ts @@ -0,0 +1,200 @@ +/** + * A mesh, raised from nothing, joined by two machines, delivering a credential neither the mesh + * nor the broker can read. + * + * Everything before this proves a part. This proves the parts meet — which is the thing the + * project keeps saying cannot be checked any other way (novox/hq ADR 0001: every fault of + * 2026-08-22 was found in production because nothing could be stood up locally). + * + * It needs a host binary and the substrate bundle: + * + * MESH_LAB_HOST_BINARY=.../mesh-host + * MESH_LAB_BUNDLE=.../examples/substrate-first-node.lock + * + * The bundle's image references are rewritten to the ones this scenario's own registry serves. + * A digest belongs to whatever registry serves it, so a committed bundle names a registry that is + * not this one — rewriting is what makes it applicable rather than a placeholder to tidy away. + */ + +import { test, before, after } from "node:test"; +import assert from "node:assert/strict"; +import { existsSync, readFileSync } from "node:fs"; +import { loadScenario } from "../../src/declaration/parse.ts"; +import { raise } from "../../src/lifecycle/raise.ts"; +import { destroy, exec } from "../../src/lifecycle/operate.ts"; +import { hostBinaryPath, HOST_PATH } from "../../src/lifecycle/place.ts"; +import { labIsUsable, destroyAll } from "./harness.ts"; + +const capability = await labIsUsable(); +const binary = hostBinaryPath(); +const bundle = process.env["MESH_LAB_BUNDLE"] ?? ""; + +const skip = !capability.usable + ? `lab not usable: ${capability.why}` + : !binary || !existsSync(binary) + ? "MESH_LAB_HOST_BINARY is not set to a built mesh-host" + : !bundle || !existsSync(bundle) + ? "MESH_LAB_BUNDLE is not set to a substrate bundle (mesh-host examples/)" + : false; + +const SCENARIO = "two-nodes"; +let instanceId = ""; + +function quote(s: string): string { + return `'${s.replaceAll("'", `'\\''`)}'`; +} + +async function on(machine: string, command: string): Promise<{ out: string; ok: boolean }> { + const { stdout } = await exec(instanceId, machine, [ + "sh", "-c", `${command} 2>&1; echo "__exit=$?"`, + ]); + const marker = stdout.lastIndexOf("__exit="); + return { out: stdout.slice(0, marker), ok: Number(stdout.slice(marker + 7).trim()) === 0 }; +} + +async function must(machine: string, command: string): Promise { + const { out, ok } = await on(machine, command); + if (!ok) throw new Error(`${machine}: ${command}\n${out}`); + return out; +} + +/** The control plane, which runs in a container on the first node. */ +async function mesh(command: string): Promise { + return must("anchor", `docker exec mesh-control /mesh-control ${command}`); +} + +/** + * The bundle, with every image reference pointed at this scenario's registry. + * + * Matched by repository rather than by the whole reference, because the address and the digest + * both differ from whatever the committed bundle names — and a bundle that names the wrong + * registry is not wrong, it is built for a different target. + */ +function bundleFor(images: string[]): string { + let text = readFileSync(bundle, "utf8"); + for (const pinned of images) { + const repository = pinned.slice(pinned.indexOf("/") + 1, pinned.indexOf("@")); + const escaped = repository.replaceAll("/", "\\/").replaceAll(".", "\\."); + text = text.replaceAll( + new RegExp(`[A-Za-z0-9_.:-]+\\/${escaped}@sha256:[0-9a-f]+`, "g"), + pinned, + ); + } + return text; +} + +/** Take a token out of what `token issue` printed. It is the one base64url blob on its own line. */ +function tokenFrom(said: string): string { + const found = said.split("\n").map((l) => l.trim()).find((l) => l.length > 100 && !l.includes(" ")); + assert.ok(found, `no token in:\n${said}`); + return found; +} + +before(async () => { + if (skip) return; + const raised = await raise(loadScenario(`scenarios/${SCENARIO}.yml`), {}); + instanceId = raised.instanceId; + + // The first node raises everything from a file rather than from a bundle built into the binary, + // because the digests are this registry's and are not known until it is up. + await must("anchor", `cat > /tmp/substrate.lock <<'MESHBUNDLE'\n${bundleFor(raised.images)}\nMESHBUNDLE`); + await must("anchor", `${HOST_PATH} apply /tmp/substrate.lock`); +}, { timeout: 1_800_000 }); + +after(async () => { + if (instanceId) await destroy(instanceId); + await destroyAll(`${SCENARIO}-`); +}, { timeout: 600_000 }); + +test("a bare machine becomes a mesh", { skip, timeout: 600_000 }, async () => { + const running = await must("anchor", `docker ps --format '{{.Names}}'`); + for (const container of ["mesh-store", "mesh-broker", "mesh-control"]) { + assert.match(running, new RegExp(container), `${container} is not running`); + } + // Answering, not merely up. A container that is running is not a control plane that replies — + // a distinction this project has already paid for once. + assert.ok((await mesh("status")).length > 0); +}); + +test("both machines join it, and the token is all they need", { skip, timeout: 900_000 }, async () => { + for (const [machine, node] of [["anchor", "anchor"], ["laptop", "laptop"]] as const) { + await mesh(`node add ${node}`); + const token = tokenFrom(await mesh(`token issue --node ${node}`)); + // No --name. The token says what the mesh calls the machine, which is the fault this walk + // found the first time it was run. + const said = await must(machine, `${HOST_PATH} enrol --token ${quote(token)}`); + assert.match(said, new RegExp(`enrolled as ${node}`), said); + assert.match(said, /sealing key/, "no sealing key was generated"); + } + + const recorded = await must("anchor", + `docker exec mesh-store psql -U postgres -d inventory -qAt ` + + `-c "select name from node where sealing_key is not null order by name"`); + assert.equal(recorded.trim().split("\n").map((l) => l.trim()).sort().join(","), "anchor,laptop", + "the mesh did not record a sealing key for both machines"); +}); + +test("a credential reaches both ends and the mesh holds neither", { skip, timeout: 900_000 }, async () => { + // The whole argument, on real machines: the two ends must hold the SAME password, and it must + // appear nowhere the mesh or the broker could read it. + await must("anchor", `printf %s '{"module":"postgres","version":"1",` + + `"provides":[{"name":"database","scope":"mesh"}],"serves":{"database":{"port":5432}},` + + `"grants":{"database":"/var/lib/mesh-host/grants"},` + + `"receives":{"database":"/var/lib/mesh-host/grants/mesh.json"},"resources":[]}' > /tmp/pg.json`); + await must("anchor", `printf %s '{"module":"meshboard","version":"1",` + + `"requires":["database"],"contributes":{"database":{"name":"meshboard"}},` + + `"binds":{"database":"/etc/meshboard/database.json"},` + + `"secrets":{"database":"/etc/meshboard/database.password"},"resources":[]}' > /tmp/app.json`); + await must("anchor", `docker cp /tmp/pg.json mesh-control:/pg.json`); + await must("anchor", `docker cp /tmp/app.json mesh-control:/app.json`); + await mesh("module add /pg.json"); + await mesh("module add /app.json"); + + await mesh("overlay place anchor --hub --endpoint 192.0.2.10:51820 --site lab"); + await mesh("overlay place laptop --site lab"); + for (const node of ["anchor", "laptop"]) await mesh(`assign ${node} networking`); + await mesh("assign anchor postgres"); + await mesh("assign laptop meshboard"); + + for (const machine of ["anchor", "laptop"]) { + await must(machine, `nohup ${HOST_PATH} run > /var/log/mesh-host.log 2>&1 & sleep 3`); + } + await mesh("push"); + await new Promise((r) => setTimeout(r, 8000)); + + const onConsumer = (await must("laptop", `cat /etc/meshboard/database.password`)).trim(); + const onProvider = (await must("anchor", `cat /var/lib/mesh-host/grants/laptop.secret`)).trim(); + assert.ok(onConsumer.length >= 40, `the consumer's credential is ${onConsumer.length} characters`); + assert.equal(onConsumer, onProvider, + "the two ends hold different passwords, so nothing could ever authenticate"); + + // Only the machine it is for may read it. + assert.match(await must("laptop", `stat -c %a /etc/meshboard/database.password`), /^600/); + + // And it is nowhere it could have been read on the way. The declaration crossed the broker; the + // database is the control plane's; the state is what the node reported back. + for (const [machine, where] of [ + ["laptop", "/var/lib/mesh-host/declared.json"], + ["laptop", "/var/lib/mesh-host/state.json"], + ["anchor", "/var/lib/mesh-host/declared.json"], + ] as const) { + const { out } = await on(machine, `grep -c ${quote(onConsumer)} ${where}`); + assert.equal(out.trim(), "0", `the password is in ${where} on ${machine}`); + } + const inTheMesh = await must("anchor", + `docker exec mesh-store psql -U postgres -d inventory -qAt ` + + `-c "select count(*) from secret where for_consumer like '%${onConsumer}%' ` + + `or for_provider like '%${onConsumer}%'"`); + assert.equal(inTheMesh.trim(), "0", "the control plane's database holds the password in the clear"); +}); + +test("the consumer is also told where its database is", { skip, timeout: 300_000 }, async () => { + // A password with no address is not a connection. This is the readable half, which stays + // readable on purpose — it is the secret half that could not be composed, not this one. + const told = JSON.parse(await must("laptop", `cat /etc/meshboard/database.json`)); + assert.equal(told.from, "anchor"); + assert.equal(told.at, "anchor.internal"); + assert.equal(told.serves.port, 5432); + // And the name it resolves to was written by the mesh as well, on this machine. + assert.match(await must("laptop", `grep anchor.internal /etc/hosts`), /10\.42\.0\.\d+/); +});