swarm-bao: make the reader's identity declarable, not just the store's

`swarm-bao.nix` declared the store's half of the mTLS pair as options —
`serverCertFile`, `serverKeyFile`, `clientCaFile` — and left the reader's
half as a literal inside `glue-bao-tls.nix`, which only runs where
`deploy.bao.enable` is set. A hive that did not host the store therefore
could not read from it and could not be pointed at a certificate even
when one had been placed by hand.

Adds `clientCertFile`, `clientKeyFile` and `serverCaFile` beside their
three server siblings, `mkDefault`ed by the glue to the leaf it already
mints, and moves `glue-matrix-bao-token.nix` onto them. Its gate becomes
"this host holds an identity" rather than "the store is a neighbour",
and the unit ordering that names store-local units is now conditional --
`Requires=` on an absent unit fails the job.

`serverCaFile` is separate from `clientCaFile` on purpose: one is the
store choosing which readers to trust, the other a reader choosing which
store to trust. Self-signing collapses them to one file, which is a
property of that deployment and not of the pairing.

Closes #3855.

docs/setup.md

@@ -67,11 +67,14 @@ Whether anything more is needed depends on
 The store serves TLS, and on a hive that deploys it you need do nothing: a
 first-boot unit mints a CA of the store's own plus the two leaves it signs —
 the store's server certificate and this host's client certificate — and points
-`deploy.bao.serverCertFile`, `.serverKeyFile` and `.clientCaFile` at them.
+`deploy.bao.serverCertFile`, `.serverKeyFile` and `.clientCaFile` at the store's
+half, `.clientCertFile`, `.clientKeyFile` and `.serverCaFile` at the reader's.
 
 Those are `mkDefault`s, so naming your own paths wins. Do that when your
 certificates come from a real internal CA; the store has no opinion about
-which. ⚠️ Not the gateway's HTTPS certificates and not the hive CA — this is
+which. A hive that does **not** deploy the store names the reader's three
+itself: that leaf is issued out of band, and it is the one credential the store
+cannot hand you, being what opens it. ⚠️ Not the gateway's HTTPS certificates and not the hive CA — this is
 **mTLS between services and the store**, a separate trust domain, because a
 store that took its identity from an authority it will itself distribute could
 never come up before that authority.

docs/swarm/secrets.md

@@ -156,6 +156,10 @@ mints a CA that signs exactly two things, the store's server certificate and a
 reader's client certificate, and distributes nothing. A deployment with a real
 internal CA deletes that file and names its own paths in
 `deploy.bao.serverCertFile` / `clientCaFile`; the store itself has no opinion.
+A hive that reads from a store on **another** machine names the reader's half —
+`clientCertFile`, `clientKeyFile`, `serverCaFile` — and places that leaf by hand.
+It is the one credential that cannot come out of the store, being what opens it;
+everything else a hive needs does.
 
 ## The constraint that decides where the root lives
 

nix/host-modules/glue-bao-tls.nix

@@ -73,6 +73,14 @@ in
       serverCertFile = lib.mkDefault "${pkiDir}/server.pem";
       serverKeyFile = lib.mkDefault "${pkiDir}/server-key.pem";
       clientCaFile = lib.mkDefault "${pkiDir}/ca.pem";
+
+      # A reader on this host, which happens to be the host that mints. Only
+      # these three are what a reader elsewhere needs placed by hand; that they
+      # collapse to the same CA file here is a property of self-signing, not of
+      # the pairing.
+      clientCertFile = lib.mkDefault "${pkiDir}/client.pem";
+      clientKeyFile = lib.mkDefault "${pkiDir}/client-key.pem";
+      serverCaFile = lib.mkDefault "${pkiDir}/ca.pem";
     };
 
     # Idempotent on ABSENCE, never on content. Re-issuing the CA invalidates

nix/host-modules/glue-matrix-bao-token.nix

@@ -16,11 +16,11 @@
 # overwrites it with the swarm's copy when the store has one. A store that is
 # empty or unreachable leaves a working hive with a local token.
 #
-# 📌 LIMIT, stated rather than hidden: this gates on the store running HERE.
+# 📌 This runs wherever a client identity is configured, NOT only where the
-# A hive reading from a store on another machine needs the same unit with that
+# store is. `deploy.bao.enable` would have been the co-location assumption
-# machine's address and a client leaf issued out of band — the mechanism is
+# itself; the reader needs a certificate, not a neighbour. On the store's own
-# identical, only `-address` and the cert's provenance differ. Deferred until
+# host ./glue-bao-tls.nix supplies one as a `mkDefault` and nothing changes;
-# there is a second hive to test it against, rather than shipped untested.
+# elsewhere an operator places the leaf and names it, and the same unit works.
 {
   pkgs,
   lib,
@@ -33,11 +33,12 @@ let
   baoCfg = hyperhiveCfg.swarm.bao;
   matrixCfg = hyperhiveCfg.swarm.matrix;
 
-  # Owned by ./glue-bao-tls.nix, which mints them. Named here rather than
+  baoDeploy = deployCfg.bao;
-  # shared through a `let`: a cross-module binding would make these two files
+
-  # one file with a gap in the middle, and the whole point of a glue module is
+  # What decides whether this unit exists at all. A reader is defined by holding
-  # that it can be deleted on its own.
+  # a certificate the store accepts, and that is true on the store's own host
-  pkiDir = "/var/lib/swarm-bao-pki";
+  # and on a hive three networks away for exactly the same reason.
+  haveClientIdentity = baoDeploy.clientCertFile != null && baoDeploy.clientKeyFile != null;
 
   # Where the token lives in the store. A path, not a convention to guess at:
   # whoever writes it and whoever reads it must agree, and the agreement
@@ -52,15 +53,19 @@ let
   matrixMachine = "hive-matrix";
 in
 {
-  config = lib.mkIf (hyperhiveCfg.enable && deployCfg.bao.enable && deployCfg.matrix.enable) {
+  config = lib.mkIf (hyperhiveCfg.enable && haveClientIdentity && deployCfg.matrix.enable) {
     systemd.services.swarm-bao-matrix-token = {
       description = "fetch the matrix registration token from the swarm secret store";
-      after = [
+      # Every one of these names a unit that exists only where the store runs.
+      # `Requires=` on an absent unit fails the job outright, so the ordering is
+      # conditional even though the read is not: off-host there is nothing local
+      # to wait for, and the timeout below is what bounds the attempt instead.
+      after = lib.optionals baoDeploy.enable [
         "swarm-bao-pki.service"
         "container@${baoCfg.machine}.service"
       ];
-      wants = [ "container@${baoCfg.machine}.service" ];
+      wants = lib.optionals baoDeploy.enable [ "container@${baoCfg.machine}.service" ];
-      requires = [ "swarm-bao-pki.service" ];
+      requires = lib.optionals baoDeploy.enable [ "swarm-bao-pki.service" ];
       before = [ "container@${matrixMachine}.service" ];
       wantedBy = [ "container@${matrixMachine}.service" ];
       path = [
@@ -77,9 +82,13 @@ in
       };
       environment = {
         BAO_ADDR = "https://${baoCfg.domain}:${toString baoCfg.port}";
-        BAO_CACERT = "${pkiDir}/ca.pem";
+        BAO_CLIENT_CERT = baoDeploy.clientCertFile;
-        BAO_CLIENT_CERT = "${pkiDir}/client.pem";
+        BAO_CLIENT_KEY = baoDeploy.clientKeyFile;
-        BAO_CLIENT_KEY = "${pkiDir}/client-key.pem";
+      }
+      # Absent means the system trust store, which is what a deployment with a
+      # real CA wants and what a self-signed one must not be left with.
+      // lib.optionalAttrs (baoDeploy.serverCaFile != null) {
+        BAO_CACERT = baoDeploy.serverCaFile;
       };
       script = ''
         set -euo pipefail

nix/host-modules/swarm-bao.nix

@@ -268,6 +268,54 @@ in
       '';
     };
 
+    clientCertFile = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/var/lib/swarm-bao/client.pem";
+      description = ''
+        Certificate a **reader on this machine** presents to the store.
+
+        The counterpart to {option}`services.hyperhive.deploy.bao.clientCaFile`,
+        which is the store's side of the same handshake. Only the store's side
+        was declarable, so a hive that did not run the store had no way to be
+        pointed at a certificate even when one was placed for it.
+
+        On a hive that runs the store, a glue module supplies the leaf it minted,
+        as a `mkDefault`. Everywhere else this is the credential an operator
+        places by hand — the one secret that cannot come out of the store,
+        because it is what opens it.
+
+        A path, never a value.
+      '';
+    };
+
+    clientKeyFile = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/var/lib/swarm-bao/client-key.pem";
+      description = ''
+        Private key for {option}`services.hyperhive.deploy.bao.clientCertFile`.
+        Both or neither — a certificate with no key authenticates nothing.
+      '';
+    };
+
+    serverCaFile = lib.mkOption {
+      type = lib.types.nullOr lib.types.str;
+      default = null;
+      example = "/var/lib/swarm-bao/ca.pem";
+      description = ''
+        Authority a **reader on this machine** validates the store's certificate
+        against.
+
+        Not {option}`services.hyperhive.deploy.bao.clientCaFile` with the words
+        rearranged: that one is the store choosing which readers to trust, this
+        one is a reader choosing which store to trust. A deployment that
+        self-signs both ends points them at the same file and reads that as
+        confirmation they are interchangeable — they are not, and they diverge
+        the moment either end gets a real CA.
+      '';
+    };
+
     extraListenAddresses = lib.mkOption {
       type = lib.types.listOf lib.types.str;
       default = [ ];

nix/module-eval.nix

@@ -90,6 +90,18 @@ let
     deploy.bao.enable = true;
     deploy.matrix.enable = true;
   };
+  # A hive that reads from a store it does not run: no `deploy.bao.enable`, so
+  # nothing here mints a leaf and the operator names one placed by hand. The
+  # deployment this pairing exists to serve, and the one that was previously
+  # inexpressible — the gate asked whether the store was a neighbour.
+  baoRemoteReader = hive {
+    deploy.matrix.enable = true;
+    deploy.bao.clientCertFile = "/etc/pki/bao-client.pem";
+    deploy.bao.clientKeyFile = "/etc/pki/bao-client-key.pem";
+  };
+  # The same hive with the identity taken away, which separates "a homeserver
+  # is deployed" from "this host can authenticate to the store".
+  matrixNoBaoIdentity = hive { deploy.matrix.enable = true; };
 
   # A priority collision is a property of the *option*, not
   # of the merged value's interior — nix throws the moment the value is
@@ -205,6 +217,29 @@ let
       name = "a store with no homeserver beside it renders no token reader";
       ok = !(baoPkcs11.systemd.services ? swarm-bao-matrix-token);
     }
+    {
+      name = "a hive that names a client identity reads from a store it does not run";
+      ok = baoRemoteReader.systemd.services ? swarm-bao-matrix-token;
+    }
+    {
+      # Absence arm for the one above, and the reason the gate is the identity
+      # rather than the homeserver: without it, deploying matrix anywhere would
+      # render a reader that cannot authenticate.
+      name = "a homeserver with no way to authenticate to the store renders no token reader";
+      ok = !(matrixNoBaoIdentity.systemd.services ? swarm-bao-matrix-token);
+    }
+    {
+      # `Requires=` on a unit that does not exist fails the job, and nothing
+      # local mints certificates off-host — so this orders against nothing.
+      # Eval cannot see that failure; only the empty list here stands in for it.
+      name = "an off-host reader requires no unit the store's host would have provided";
+      ok = baoRemoteReader.systemd.services.swarm-bao-matrix-token.requires == [ ];
+    }
+    {
+      # Presence control for the case above: the list is conditional, not gone.
+      name = "a co-located reader still orders after the local pki unit";
+      ok = baoWithMatrix.systemd.services.swarm-bao-matrix-token.requires == [ "swarm-bao-pki.service" ];
+    }
   ];
 
   bad = builtins.filter (c: !c.ok) cases;