src/hydra-queue-runner/build-remote.cc

@@ -9,10 +9,13 @@
 #include "path.hh"
 #include "legacy-ssh-store.hh"
 #include "serve-protocol.hh"
+#include "serve-protocol-impl.hh"
 #include "state.hh"
 #include "current-process.hh"
 #include "processes.hh"
 #include "util.hh"
+#include "serve-protocol.hh"
+#include "serve-protocol-impl.hh"
 #include "ssh.hh"
 #include "finally.hh"
 #include "url.hh"
@@ -36,6 +39,38 @@ bool ::Machine::isLocalhost() const
 
 namespace nix::build_remote {
 
+static std::unique_ptr<SSHMaster::Connection> openConnection(
+    ::Machine::ptr machine, SSHMaster & master)
+{
+    Strings command = {"nix-store", "--serve", "--write"};
+    if (machine->isLocalhost()) {
+        command.push_back("--builders");
+        command.push_back("");
+    } else {
+        auto remoteStore = machine->storeUri.params.find("remote-store");
+        if (remoteStore != machine->storeUri.params.end()) {
+            command.push_back("--store");
+            command.push_back(shellEscape(remoteStore->second));
+        }
+    }
+
+    auto ret = master.startCommand(std::move(command), {
+        "-a", "-oBatchMode=yes", "-oConnectTimeout=60", "-oTCPKeepAlive=yes"
+    });
+
+    // XXX: determine the actual max value we can use from /proc.
+
+    // FIXME: Should this be upstreamed into `startCommand` in Nix?
+
+    int pipesize = 1024 * 1024;
+
+    fcntl(ret->in.get(), F_SETPIPE_SZ, &pipesize);
+    fcntl(ret->out.get(), F_SETPIPE_SZ, &pipesize);
+
+    return ret;
+}
+
+
 static void copyClosureTo(
     ::Machine::Connection & conn,
     Store & destStore,
@@ -52,8 +87,8 @@ static void copyClosureTo(
     // FIXME: substitute output pollutes our build log
     /* Get back the set of paths that are already valid on the remote
        host. */
-    auto present = conn.store->queryValidPaths(
+    auto present = conn.queryValidPaths(
-        closure, true, useSubstitutes);
+        destStore, true, closure, useSubstitutes);
 
     if (present.size() == closure.size()) return;
 
@@ -68,7 +103,12 @@ static void copyClosureTo(
     std::unique_lock<std::timed_mutex> sendLock(conn.machine->state->sendLock,
         std::chrono::seconds(600));
 
-    conn.store->addMultipleToStoreLegacy(destStore, missing);
+    conn.to << ServeProto::Command::ImportPaths;
+    destStore.exportPaths(missing, conn.to);
+    conn.to.flush();
+
+    if (readInt(conn.from) != 1)
+        throw Error("remote machine failed to import closure");
 }
 
 
@@ -188,7 +228,7 @@ static BuildResult performBuild(
     counter & nrStepsBuilding
 )
 {
-    auto kont = conn.store->buildDerivationAsync(drvPath, drv, options);
+    conn.putBuildDerivationRequest(localStore, drvPath, drv, options);
 
     BuildResult result;
 
@@ -197,10 +237,7 @@ static BuildResult performBuild(
     startTime = time(0);
     {
         MaintainCount<counter> mc(nrStepsBuilding);
-        result = kont();
+        result = ServeProto::Serialise<BuildResult>::read(localStore, conn);
-        // Without proper call-once functions, we need to manually
-        // delete after calling.
-        kont = {};
     }
     stopTime = time(0);
 
@@ -216,7 +253,7 @@ static BuildResult performBuild(
 
     // If the protocol was too old to give us `builtOutputs`, initialize
     // it manually by introspecting the derivation.
-    if (GET_PROTOCOL_MINOR(conn.store->getProtocol()) < 6)
+    if (GET_PROTOCOL_MINOR(conn.remoteVersion) < 6)
     {
         // If the remote is too old to handle CA derivations, we can’t get this
         // far anyways
@@ -261,10 +298,11 @@ static void copyPathFromRemote(
               lambda function only gets executed if someone tries to read
               from source2, we will send the command from here rather
               than outside the lambda. */
-        conn.store->narFromPath(info.path, [&](Source & source) {
+          conn.to << ServeProto::Command::DumpStorePath << localStore.printStorePath(info.path);
-            TeeSource tee{source, sink};
+          conn.to.flush();
-            extractNarData(tee, conn.store->printStorePath(info.path), narMembers);
+
-        });
+          TeeSource tee(conn.from, sink);
+          extractNarData(tee, localStore.printStorePath(info.path), narMembers);
       });
 
       destStore.addToStore(info, *source2, NoRepair, NoCheckSigs);
@@ -366,39 +404,30 @@ void State::buildRemote(ref<Store> destStore,
 
         updateStep(ssConnecting);
 
-        // FIXME: rewrite to use Store.
+        auto storeRef = machine->completeStoreReference();
-        ::Machine::Connection conn {
+
-            .machine = machine,
+        auto * pSpecified = std::get_if<StoreReference::Specified>(&storeRef.variant);
-            .store = [&]{
-                auto * pSpecified = std::get_if<StoreReference::Specified>(&machine->storeUri.variant);
         if (!pSpecified || pSpecified->scheme != "ssh") {
             throw Error("Currently, only (legacy-)ssh stores are supported!");
         }
 
-                auto remoteStore = machine->openStore().dynamic_pointer_cast<LegacySSHStore>();
+        LegacySSHStoreConfig storeConfig {
-                assert(remoteStore);
+            pSpecified->scheme,
-
+            pSpecified->authority,
-                remoteStore->connPipeSize = 1024 * 1024;
+            storeRef.params
-
-                if (machine->isLocalhost()) {
-                    auto rp_new = remoteStore->remoteProgram.get();
-                    rp_new.push_back("--builders");
-                    rp_new.push_back("");
-                    const_cast<nix::Setting<Strings> &>(remoteStore->remoteProgram).assign(rp_new);
-                }
-                remoteStore->extraSshArgs = {
-                    "-a", "-oBatchMode=yes", "-oConnectTimeout=60", "-oTCPKeepAlive=yes"
         };
-                const_cast<nix::Setting<int> &>(remoteStore->logFD).assign(logFD.get());
 
-                return nix::ref{remoteStore};
+        auto master = storeConfig.createSSHMaster(
-            }(),
+            false, // no SSH master yet
-        };
+            logFD.get());
+
+        // FIXME: rewrite to use Store.
+        auto child = build_remote::openConnection(machine, master);
 
         {
             auto activeStepState(activeStep->state_.lock());
             if (activeStepState->cancelled) throw Error("step cancelled");
-            activeStepState->pid = conn.store->getConnectionPid();
+            activeStepState->pid = child->sshPid;
         }
 
         Finally clearPid([&]() {
@@ -413,12 +442,35 @@ void State::buildRemote(ref<Store> destStore,
                process. Meh. */
         });
 
+        ::Machine::Connection conn {
+            {
+                .to = child->in.get(),
+                .from = child->out.get(),
+                /* Handshake. */
+                .remoteVersion = 0xdadbeef, // FIXME avoid dummy initialize
+            },
+            /*.machine =*/ machine,
+        };
+
         Finally updateStats([&]() {
-            auto stats = conn.store->getConnectionStats();
+            bytesReceived += conn.from.read;
-            bytesReceived += stats.bytesReceived;
+            bytesSent += conn.to.written;
-            bytesSent += stats.bytesSent;
         });
 
+        constexpr ServeProto::Version our_version = 0x206;
+
+        try {
+            conn.remoteVersion = decltype(conn)::handshake(
+                conn.to,
+                conn.from,
+                our_version,
+                machine->storeUri.render());
+        } catch (EndOfFile & e) {
+            child->sshPid.wait();
+            std::string s = chomp(readFile(result.logFile));
+            throw Error("cannot connect to ‘%1%’: %2%", machine->storeUri.render(), s);
+        }
+
         {
             auto info(machine->state->connectInfo.lock());
             info->consecutiveFailures = 0;
@@ -487,7 +539,7 @@ void State::buildRemote(ref<Store> destStore,
 
             auto now1 = std::chrono::steady_clock::now();
 
-            auto infos = conn.store->queryPathInfosUncached(outputs);
+            auto infos = conn.queryPathInfos(*localStore, outputs);
 
             size_t totalNarSize = 0;
             for (auto & [_, info] : infos) totalNarSize += info.narSize;
@@ -522,11 +574,9 @@ void State::buildRemote(ref<Store> destStore,
             }
         }
 
-        /* Shut down the connection done by RAII.
+        /* Shut down the connection. */
-
+        child->in = -1;
-           Only difference is kill() instead of wait() (i.e. send signal
+        child->sshPid.wait();
-           then wait())
-         */
 
     } catch (Error & e) {
         /* Disable this machine until a certain period of time has

src/hydra-queue-runner/state.hh

@@ -20,7 +20,9 @@
 #include "store-api.hh"
 #include "sync.hh"
 #include "nar-extractor.hh"
-#include "legacy-ssh-store.hh"
+#include "serve-protocol.hh"
+#include "serve-protocol-impl.hh"
+#include "serve-protocol-connection.hh"
 #include "machines.hh"
 
 
@@ -290,11 +292,9 @@ struct Machine : nix::Machine
     bool isLocalhost() const;
 
     // A connection to a machine
-    struct Connection {
+    struct Connection : nix::ServeProto::BasicClientConnection {
         // Backpointer to the machine
         ptr machine;
-        // Opened store
-        nix::ref<nix::LegacySSHStore> store;
     };
 };