Merge branch 'dev' into udp-peer

* touch keys

* add key sets to Transaction class

* add optional performance monitoring to Transaction class

* add replay network 65534 as nid where txns from other networks can be replayed

* replay network code

* experimental simple txq

* Revert "experimental simple txq"

This reverts commit 8dfa91617e.

* experimental large ledgers with no txq

src/ripple/app/consensus/RCLConsensus.h

@@ -129,6 +129,12 @@ class RCLConsensus
             return mode_;
         }
 
+        void
+        setProposing()
+        {
+            mode_ = ConsensusMode::proposing;
+        }
+
         /** Called before kicking off a new consensus round.
 
             @param prevLedger Ledger that will be prior ledger for next round
@@ -465,6 +471,12 @@ public:
         return adaptor_.mode();
     }
 
+    void
+    setProposing()
+    {
+        adaptor_.setProposing();
+    }
+
     ConsensusPhase
     phase() const
     {

src/ripple/app/ledger/impl/LedgerMaster.cpp

@@ -212,6 +212,7 @@ LedgerMaster::getCurrentLedgerIndex()
 LedgerIndex
 LedgerMaster::getValidLedgerIndex()
 {
+    std::cout << "getValidLedgerIndex: " << mValidLedgerSeq << "\n";
     return mValidLedgerSeq;
 }
 

src/ripple/app/misc/HashRouter.cpp

@@ -128,4 +128,20 @@ HashRouter::shouldRelay(uint256 const& key)
     return s.releasePeerSet();
 }
 
+void
+HashRouter::setTouchedKeys(uint256 const& id, std::set<uint256>&& k)
+{
+    std::unique_lock<std::shared_mutex> lock(touchedKeysMutex_);
+    touchedKeysMap_.insert_or_assign(id, std::move(k));
+}
+
+std::optional<std::reference_wrapper<const std::set<uint256>>>
+HashRouter::getTouchedKeys(uint256 const& id)
+{
+    std::shared_lock<std::shared_mutex> lock(touchedKeysMutex_);
+    if (auto it = touchedKeysMap_.find(id); it != touchedKeysMap_.end())
+        return std::cref(it->second);
+    return std::nullopt;
+}
+
 }  // namespace ripple

src/ripple/app/misc/HashRouter.h

@@ -27,6 +27,8 @@
 #include <ripple/beast/container/aged_unordered_map.h>
 
 #include <optional>
+#include <set>
+#include <shared_mutex>
 
 namespace ripple {
 
@@ -195,6 +197,12 @@ public:
     int
     getFlags(uint256 const& key);
 
+    void
+    setTouchedKeys(uint256 const& id, std::set<uint256>&& k);
+
+    std::optional<std::reference_wrapper<const std::set<uint256>>>
+    getTouchedKeys(uint256 const& id);
+
     /** Determines whether the hashed item should be relayed.
 
         Effects:
@@ -217,6 +225,9 @@ private:
 
     std::mutex mutable mutex_;
 
+    mutable std::shared_mutex touchedKeysMutex_;
+    std::map<uint256, std::set<uint256>> touchedKeysMap_;
+
     // Stores all suppressed hashes and their expiration time
     beast::aged_unordered_map<
         uint256,

src/ripple/app/misc/NetworkOPs.cpp

@@ -944,7 +944,13 @@ NetworkOPsImp::processHeartbeatTimer()
         // do we have sufficient peers? If not, we are disconnected.
         if (numPeers < minPeerCount_)
         {
-            if (mMode != OperatingMode::DISCONNECTED)
+            if (app_.config().NETWORK_ID == 65534)
+            {
+                // replay network is always considered to be connected
+                // ensuring that it actually is is up to the tester
+                setMode(OperatingMode::FULL);
+            }
+            else if (mMode != OperatingMode::DISCONNECTED)
             {
                 setMode(OperatingMode::DISCONNECTED);
                 JLOG(m_journal.warn())
@@ -1797,6 +1803,13 @@ NetworkOPsImp::beginConsensus(uint256 const& networkClosed)
 {
     assert(networkClosed.isNonZero());
 
+    if (app_.config().NETWORK_ID == 65534)
+    {
+        // replay network automatically goes to proposing
+        setMode(OperatingMode::FULL);
+        mConsensus.setProposing();
+    }
+
     auto closingInfo = m_ledgerMaster.getCurrentLedger()->info();
 
     JLOG(m_journal.info()) << "Consensus time for #" << closingInfo.seq

src/ripple/app/misc/NetworkOPs.h

@@ -110,7 +110,7 @@ public:
         std::shared_ptr<Transaction>& transaction,
         bool bUnlimited,
         bool bLocal,
-        FailHard failType) = 0;
+        FailHard failType = FailHard::no) = 0;
 
     //--------------------------------------------------------------------------
     //

src/ripple/app/misc/Transaction.h

@@ -29,6 +29,7 @@
 #include <ripple/protocol/TxMeta.h>
 #include <boost/optional.hpp>
 #include <optional>
+#include <set>
 #include <variant>
 
 namespace ripple {
@@ -406,6 +407,8 @@ private:
     std::shared_ptr<STTx const> mTransaction;
     Application& mApp;
     beast::Journal j_;
+
+    std::set<uint256> keysTouched;
 };
 
 }  // namespace ripple

src/ripple/app/misc/TxQ.h

@@ -30,6 +30,7 @@
 #include <boost/circular_buffer.hpp>
 #include <boost/intrusive/set.hpp>
 #include <optional>
+#include <set>
 #include <vector>
 
 namespace ripple {
@@ -105,13 +106,13 @@ public:
         FeeLevel64 minimumEscalationMultiplier = baseLevel * 500;
         /// Minimum number of transactions to allow into the ledger
         /// before escalation, regardless of the prior ledger's size.
-        std::uint32_t minimumTxnInLedger = 32;
+        std::uint32_t minimumTxnInLedger = 5000;
         /// Like @ref minimumTxnInLedger for standalone mode.
         /// Primarily so that tests don't need to worry about queuing.
         std::uint32_t minimumTxnInLedgerSA = 1000;
         /// Number of transactions per ledger that fee escalation "works
         /// towards".
-        std::uint32_t targetTxnInLedger = 1000;
+        std::uint32_t targetTxnInLedger = 10000;
         /** Optional maximum allowed value of transactions per ledger before
             fee escalation kicks in. By default, the maximum is an emergent
             property of network, validator, and consensus performance. This
@@ -741,6 +742,7 @@ private:
         FeeMetrics::Snapshot const& metricsSnapshot,
         std::lock_guard<std::mutex> const& lock) const;
 
+public:
     // Helper function for TxQ::apply.  If a transaction's fee is high enough,
     // attempt to directly apply that transaction to the ledger.
     std::optional<std::pair<TER, bool>>
@@ -751,6 +753,7 @@ private:
         ApplyFlags flags,
         beast::Journal j);
 
+private:
     // Helper function that removes a replaced entry in _byFee.
     std::optional<TxQAccount::TxMap::iterator>
     removeFromByFee(

src/ripple/app/misc/impl/Transaction.cpp

@@ -34,8 +34,270 @@
 #include <ripple/protocol/jss.h>
 #include <ripple/rpc/CTID.h>
 
+#include <algorithm>
+#include <atomic>
+#include <chrono>
+#include <iomanip>
+#include <thread>
+#include <unordered_map>
+#include <vector>
+
+#define ENABLE_PERFORMANCE_TRACKING 0
+
 namespace ripple {
 
+#if ENABLE_PERFORMANCE_TRACKING
+
+// Performance monitoring statistics
+namespace {
+// Design: Uses thread-local storage for most stats to avoid contention.
+// Only global concurrency tracking uses atomics, as it requires cross-thread
+// visibility. Statistics are aggregated using dirty reads for minimal
+// performance impact.
+
+// Thread-local statistics - no synchronization needed!
+struct ThreadLocalStats
+{
+    uint64_t executionCount = 0;
+    uint64_t totalTimeNanos = 0;
+    uint64_t totalKeys = 0;
+    uint32_t currentlyExecuting = 0;  // 0 or 1 for this thread
+    std::thread::id threadId = std::this_thread::get_id();
+
+    // For global registry
+    ThreadLocalStats* next = nullptr;
+
+    ThreadLocalStats();
+    ~ThreadLocalStats();
+};
+
+// Global registry of thread-local stats (only modified during thread
+// creation/destruction)
+struct GlobalRegistry
+{
+    std::atomic<ThreadLocalStats*> head{nullptr};
+    std::atomic<uint64_t> globalExecutions{0};
+    std::atomic<uint32_t> globalConcurrent{
+        0};  // Current global concurrent executions
+    std::atomic<uint32_t> maxGlobalConcurrent{0};  // Max observed
+
+    // For tracking concurrency samples
+    std::vector<uint32_t> concurrencySamples;
+    std::mutex sampleMutex;  // Only used during printing
+
+    std::chrono::steady_clock::time_point startTime =
+        std::chrono::steady_clock::now();
+    std::chrono::steady_clock::time_point lastPrintTime =
+        std::chrono::steady_clock::now();
+
+    static constexpr auto PRINT_INTERVAL = std::chrono::seconds(10);
+    static constexpr uint64_t PRINT_EVERY_N_CALLS = 1000;
+
+    void
+    registerThread(ThreadLocalStats* stats)
+    {
+        // Add to linked list atomically
+        ThreadLocalStats* oldHead = head.load();
+        do
+        {
+            stats->next = oldHead;
+        } while (!head.compare_exchange_weak(oldHead, stats));
+    }
+
+    void
+    unregisterThread(ThreadLocalStats* stats)
+    {
+        // In production, you'd want proper removal logic
+        // For this example, we'll just leave it in the list
+        // (threads typically live for the process lifetime anyway)
+    }
+
+    void
+    checkAndPrint(uint64_t localCount)
+    {
+        // Update approximate global count
+        uint64_t approxGlobal =
+            globalExecutions.fetch_add(localCount) + localCount;
+
+        auto now = std::chrono::steady_clock::now();
+        if (approxGlobal % PRINT_EVERY_N_CALLS < localCount ||
+            (now - lastPrintTime) >= PRINT_INTERVAL)
+        {
+            // Only one thread prints at a time
+            static std::atomic<bool> printing{false};
+            bool expected = false;
+            if (printing.compare_exchange_strong(expected, true))
+            {
+                // Double-check timing
+                now = std::chrono::steady_clock::now();
+                if ((now - lastPrintTime) >= PRINT_INTERVAL)
+                {
+                    printStats();
+                    lastPrintTime = now;
+                }
+                printing = false;
+            }
+        }
+    }
+
+    void
+    printStats()
+    {
+        // Dirty read of all thread-local stats
+        uint64_t totalExecs = 0;
+        uint64_t totalNanos = 0;
+        uint64_t totalKeyCount = 0;
+        uint32_t currentConcurrent = globalConcurrent.load();
+        uint32_t maxConcurrent = maxGlobalConcurrent.load();
+        std::unordered_map<
+            std::thread::id,
+            std::tuple<uint64_t, uint64_t, uint64_t>>
+            threadData;
+
+        // Walk the linked list of thread-local stats
+        ThreadLocalStats* current = head.load();
+        while (current)
+        {
+            // Dirty reads - no synchronization!
+            uint64_t execs = current->executionCount;
+            if (execs > 0)
+            {
+                uint64_t nanos = current->totalTimeNanos;
+                uint64_t keys = current->totalKeys;
+
+                totalExecs += execs;
+                totalNanos += nanos;
+                totalKeyCount += keys;
+
+                threadData[current->threadId] = {execs, nanos, keys};
+            }
+            current = current->next;
+        }
+
+        if (totalExecs == 0)
+            return;
+
+        double avgTimeUs =
+            static_cast<double>(totalNanos) / totalExecs / 1000.0;
+        double avgKeys = static_cast<double>(totalKeyCount) / totalExecs;
+        double totalTimeMs = static_cast<double>(totalNanos) / 1000000.0;
+
+        // Calculate wall clock time elapsed
+        auto now = std::chrono::steady_clock::now();
+        auto wallTimeMs = std::chrono::duration_cast<std::chrono::milliseconds>(
+                              now - startTime)
+                              .count();
+        double effectiveParallelism = wallTimeMs > 0
+            ? totalTimeMs / static_cast<double>(wallTimeMs)
+            : 0.0;
+
+        std::cout
+            << "\n=== Transaction::tryDirectApply Performance Stats ===\n";
+        std::cout << "Total executions: ~" << totalExecs << " (dirty read)\n";
+        std::cout << "Wall clock time: " << wallTimeMs << " ms\n";
+        std::cout << "Total CPU time: " << std::fixed << std::setprecision(2)
+                  << totalTimeMs << " ms\n";
+        std::cout << "Effective parallelism: " << std::fixed
+                  << std::setprecision(2) << effectiveParallelism << "x\n";
+        std::cout << "Average time: " << std::fixed << std::setprecision(2)
+                  << avgTimeUs << " μs\n";
+        std::cout << "Average keys touched: " << std::fixed
+                  << std::setprecision(2) << avgKeys << "\n";
+        std::cout << "Current concurrent executions: " << currentConcurrent
+                  << "\n";
+        std::cout << "Max concurrent observed: " << maxConcurrent << "\n";
+        std::cout << "Active threads: " << threadData.size() << "\n";
+
+        std::cout << "Thread distribution:\n";
+
+        // Sort threads by total time spent (descending)
+        std::vector<std::pair<
+            std::thread::id,
+            std::tuple<uint64_t, uint64_t, uint64_t>>>
+            sortedThreads(threadData.begin(), threadData.end());
+        std::sort(
+            sortedThreads.begin(),
+            sortedThreads.end(),
+            [](const auto& a, const auto& b) {
+                return std::get<1>(a.second) >
+                    std::get<1>(b.second);  // Sort by time
+            });
+
+        for (const auto& [tid, data] : sortedThreads)
+        {
+            auto [count, time, keys] = data;
+            double percentage =
+                (static_cast<double>(count) / totalExecs) * 100.0;
+            double avgThreadTimeUs = static_cast<double>(time) / count / 1000.0;
+            double totalThreadTimeMs = static_cast<double>(time) / 1000000.0;
+            double timePercentage =
+                (static_cast<double>(time) / totalNanos) * 100.0;
+            std::cout << "  Thread " << tid << ": " << count << " executions ("
+                      << std::fixed << std::setprecision(1) << percentage
+                      << "%), total " << std::setprecision(2)
+                      << totalThreadTimeMs << " ms (" << std::setprecision(1)
+                      << timePercentage << "% of time), avg "
+                      << std::setprecision(2) << avgThreadTimeUs << " μs\n";
+        }
+
+        std::cout << "Hardware concurrency: "
+                  << std::thread::hardware_concurrency() << "\n";
+        std::cout << "===================================================\n\n";
+        std::cout.flush();
+    }
+};
+
+static GlobalRegistry globalRegistry;
+
+// Thread-local instance
+thread_local ThreadLocalStats tlStats;
+
+// Constructor/destructor for thread registration
+ThreadLocalStats::ThreadLocalStats()
+{
+    globalRegistry.registerThread(this);
+}
+
+ThreadLocalStats::~ThreadLocalStats()
+{
+    globalRegistry.unregisterThread(this);
+}
+
+// RAII class to track concurrent executions (global)
+class ConcurrentExecutionTracker
+{
+    // Note: This introduces minimal atomic contention to track true global
+    // concurrency. The alternative would miss concurrent executions between
+    // print intervals.
+public:
+    ConcurrentExecutionTracker()
+    {
+        tlStats.currentlyExecuting = 1;
+
+        // Update global concurrent count
+        uint32_t current = globalRegistry.globalConcurrent.fetch_add(1) + 1;
+
+        // Update max if needed (only contends when setting new maximum)
+        uint32_t currentMax = globalRegistry.maxGlobalConcurrent.load();
+        while (current > currentMax &&
+               !globalRegistry.maxGlobalConcurrent.compare_exchange_weak(
+                   currentMax, current))
+        {
+            // Loop until we successfully update or current is no longer >
+            // currentMax
+        }
+    }
+
+    ~ConcurrentExecutionTracker()
+    {
+        tlStats.currentlyExecuting = 0;
+        globalRegistry.globalConcurrent.fetch_sub(1);
+    }
+};
+}  // namespace
+
+#endif  // ENABLE_PERFORMANCE_TRACKING
+
 Transaction::Transaction(
     std::shared_ptr<STTx const> const& stx,
     std::string& reason,
@@ -45,6 +307,38 @@ Transaction::Transaction(
     try
     {
         mTransactionID = mTransaction->getTransactionID();
+
+        OpenView sandbox(*app.openLedger().current());
+
+        sandbox.getAndResetKeysTouched();
+
+        ApplyFlags flags{0};
+
+#if ENABLE_PERFORMANCE_TRACKING
+        ConcurrentExecutionTracker concurrentTracker;
+        auto startTime = std::chrono::steady_clock::now();
+#endif
+
+        if (auto directApplied =
+                app.getTxQ().tryDirectApply(app, sandbox, stx, flags, j_))
+            keysTouched = sandbox.getAndResetKeysTouched();
+
+#if ENABLE_PERFORMANCE_TRACKING
+        auto endTime = std::chrono::steady_clock::now();
+        auto elapsedNanos =
+            std::chrono::duration_cast<std::chrono::nanoseconds>(
+                endTime - startTime)
+                .count();
+
+        tlStats.executionCount++;
+        tlStats.totalTimeNanos += elapsedNanos;
+        tlStats.totalKeys += keysTouched.size();
+
+        if (tlStats.executionCount % 100 == 0)
+        {
+            globalRegistry.checkAndPrint(100);
+        }
+#endif
     }
     catch (std::exception& e)
     {

src/ripple/app/misc/impl/TxQ.cpp

@@ -30,6 +30,7 @@
 #include <algorithm>
 #include <limits>
 #include <numeric>
+#include <set>
 
 namespace ripple {
 
@@ -739,10 +740,20 @@ TxQ::apply(
     STAmountSO stAmountSO{view.rules().enabled(fixSTAmountCanonicalize)};
     NumberSO stNumberSO{view.rules().enabled(fixUniversalNumber)};
 
+    auto const transactionID = tx->getTransactionID();
+
     // See if the transaction paid a high enough fee that it can go straight
     // into the ledger.
+
+    view.getAndResetKeysTouched();
     if (auto directApplied = tryDirectApply(app, view, tx, flags, j))
+    {
+        app.getHashRouter().setTouchedKeys(
+            transactionID, view.getAndResetKeysTouched());
         return *directApplied;
+    }
+
+    return {telCAN_NOT_QUEUE, false};
 
     // If we get past tryDirectApply() without returning then we expect
     // one of the following to occur:
@@ -758,6 +769,47 @@ TxQ::apply(
     if (!isTesSuccess(pfresult.ter))
         return {pfresult.ter, false};
 
+    bool const isReplayNetwork = (app.config().NETWORK_ID == 65534);
+
+    if (isReplayNetwork)
+    {
+        // in the replay network everything is always queued no matter what
+
+        std::lock_guard lock(mutex_);
+        auto const metricsSnapshot = feeMetrics_.getSnapshot();
+        auto const feeLevelPaid =
+            getRequiredFeeLevel(view, flags, metricsSnapshot, lock);
+
+        auto const account = (*tx)[sfAccount];
+        AccountMap::iterator accountIter = byAccount_.find(account);
+        bool const accountIsInQueue = accountIter != byAccount_.end();
+
+        if (!accountIsInQueue)
+        {
+            // Create a new TxQAccount object and add the byAccount lookup.
+            bool created;
+            std::tie(accountIter, created) =
+                byAccount_.emplace(account, TxQAccount(tx));
+            (void)created;
+            assert(created);
+        }
+
+        flags &= ~tapRETRY;
+
+        auto& candidate = accountIter->second.add(
+            {tx, transactionID, feeLevelPaid, flags, pfresult});
+
+        // Then index it into the byFee lookup.
+        byFee_.insert(candidate);
+        JLOG(j_.debug()) << "Added transaction " << candidate.txID
+                         << " with result " << transToken(pfresult.ter)
+                         << " from " << (accountIsInQueue ? "existing" : "new")
+                         << " account " << candidate.account << " to queue."
+                         << " Flags: " << flags;
+
+        return {terQUEUED, false};
+    }
+
     // If the account is not currently in the ledger, don't queue its tx.
     auto const account = (*tx)[sfAccount];
     Keylet const accountKey{keylet::account(account)};
@@ -768,6 +820,7 @@ TxQ::apply(
         if (tx->getTxnType() == ttIMPORT)
             return {telCAN_NOT_QUEUE_IMPORT, false};
 
+        std::cout << "TxQ.cpp:823 NO_ACCOUNT\n";
         return {terNO_ACCOUNT, false};
     }
 
@@ -841,7 +894,6 @@ TxQ::apply(
     // is allowed in the TxQ:
     //  1. If the account's queue is empty or
     //  2. If the blocker replaces the only entry in the account's queue.
-    auto const transactionID = tx->getTransactionID();
     if (pfresult.consequences.isBlocker())
     {
         if (acctTxCount > 1)
@@ -1148,11 +1200,11 @@ TxQ::apply(
                 (potentialTotalSpend == XRPAmount{0} &&
                  multiTxn->applyView.fees().base == 0));
             sleBump->setFieldAmount(sfBalance, balance - potentialTotalSpend);
-            // The transaction's sequence/ticket will be valid when the other
-            // transactions in the queue have been processed. If the tx has a
-            // sequence, set the account to match it. If it has a ticket, use
-            // the next queueable sequence, which is the closest approximation
-            // to the most successful case.
+            // The transaction's sequence/ticket will be valid when the
+            // other transactions in the queue have been processed. If the
+            // tx has a sequence, set the account to match it. If it has a
+            // ticket, use the next queueable sequence, which is the closest
+            // approximation to the most successful case.
             sleBump->at(sfSequence) = txSeqProx.isSeq()
                 ? txSeqProx.value()
                 : nextQueuableSeqImpl(sleAccount, lock).value();
@@ -1207,6 +1259,8 @@ TxQ::apply(
     {
         OpenView sandbox(open_ledger, &view, view.rules());
 
+        sandbox.getAndResetKeysTouched();
+
         auto result = tryClearAccountQueueUpThruTx(
             app,
             sandbox,
@@ -1219,6 +1273,10 @@ TxQ::apply(
             flags,
             metricsSnapshot,
             j);
+
+        app.getHashRouter().setTouchedKeys(
+            transactionID, sandbox.getAndResetKeysTouched());
+
         if (result.second)
         {
             sandbox.apply(view);
@@ -1657,11 +1715,16 @@ TxQ::accept(Application& app, OpenView& view)
             JLOG(j_.trace()) << "Applying queued transaction "
                              << candidateIter->txID << " to open ledger.";
 
+            view.getAndResetKeysTouched();
+
             auto const [txnResult, didApply] =
                 candidateIter->apply(app, view, j_);
 
             if (didApply)
             {
+                app.getHashRouter().setTouchedKeys(
+                    candidateIter->txID, view.getAndResetKeysTouched());
+
                 // Remove the candidate from the queue
                 JLOG(j_.debug())
                     << "Queued transaction " << candidateIter->txID
@@ -1868,13 +1931,15 @@ TxQ::tryDirectApply(
     const bool isFirstImport = !sleAccount &&
         view.rules().enabled(featureImport) && tx->getTxnType() == ttIMPORT;
 
+    bool const isReplayNetwork = (app.config().NETWORK_ID == 65534);
+
     // Don't attempt to direct apply if the account is not in the ledger.
-    if (!sleAccount && !isFirstImport)
+    if (!sleAccount && !isFirstImport && !isReplayNetwork)
         return {};
 
     std::optional<SeqProxy> txSeqProx;
 
-    if (!isFirstImport)
+    if (!isFirstImport && !isReplayNetwork)
     {
         SeqProxy const acctSeqProx =
             SeqProxy::sequence((*sleAccount)[sfSequence]);
@@ -1886,18 +1951,19 @@ TxQ::tryDirectApply(
             return {};
     }
 
-    FeeLevel64 const requiredFeeLevel =
-        isFirstImport ? FeeLevel64{0} : [this, &view, flags]() {
-            std::lock_guard lock(mutex_);
-            return getRequiredFeeLevel(
-                view, flags, feeMetrics_.getSnapshot(), lock);
-        }();
+    FeeLevel64 const requiredFeeLevel = (isFirstImport || isReplayNetwork)
+        ? FeeLevel64{0}
+        : [this, &view, flags]() {
+              std::lock_guard lock(mutex_);
+              return getRequiredFeeLevel(
+                  view, flags, feeMetrics_.getSnapshot(), lock);
+          }();
 
     // If the transaction's fee is high enough we may be able to put the
     // transaction straight into the ledger.
     FeeLevel64 const feeLevelPaid = getFeeLevelPaid(view, *tx);
 
-    if (feeLevelPaid >= requiredFeeLevel)
+    if (feeLevelPaid >= requiredFeeLevel || isReplayNetwork)
     {
         // Attempt to apply the transaction directly.
         auto const transactionID = tx->getTransactionID();

src/ripple/app/tx/applySteps.h

@@ -203,7 +203,7 @@ public:
     beast::Journal const j;
 
     /// Intermediate transaction result
-    TER const ter;
+    TER ter;
     /// Success flag - whether the transaction is likely to
     /// claim a fee
     bool const likelyToClaimFee;

src/ripple/app/tx/impl/Change.cpp

@@ -23,6 +23,7 @@
 #include <ripple/app/main/Application.h>
 #include <ripple/app/misc/AmendmentTable.h>
 #include <ripple/app/misc/NetworkOPs.h>
+#include <ripple/app/misc/ReplayNetworkAccIDs.h>
 #include <ripple/app/tx/impl/Change.h>
 #include <ripple/app/tx/impl/SetSignerList.h>
 #include <ripple/app/tx/impl/XahauGenesis.h>
@@ -458,11 +459,26 @@ Change::activateXahauGenesis()
     bool const isTest =
         (ctx_.tx.getFlags() & tfTestSuite) && ctx_.app.config().standalone();
 
-    // RH NOTE: we'll only configure xahau governance structure on networks that
-    // begin with 2133... so production xahau: 21337 and its testnet 21338
-    // with 21330-21336 and 21339 also valid and reserved for dev nets etc.
-    // all other Network IDs will be conventionally configured.
-    if ((ctx_.app.config().NETWORK_ID / 10) != 2133 && !isTest)
+    // RH NOTE: we'll only configure xahau governance structure on certain
+    // network ids
+
+    const auto nid = ctx_.app.config().NETWORK_ID;
+
+    const bool isReplayNetwork = (nid == 65534);
+
+    if (nid >= 65520)
+    {
+        // networks 65520 - 65535 are are also configured as xahau gov
+    }
+    else if (isTest)
+    {
+        // test is configured like this too
+    }
+    else if (nid / 10 == 2133)
+    {
+        // networks 2133X are the valid xahau prod dev and testnets
+    }
+    else
         return;
 
     auto [ng_entries, l1_entries, l2_entries, gov_params] =
@@ -506,18 +522,29 @@ Change::activateXahauGenesis()
     sle->setFieldAmount(sfBalance, GenesisAmount);
 
     // Step 2: mint genesis distribution
-    auto mint = [&](std::string const& account, XRPAmount const& amount) {
-        auto accid_raw = parseBase58<AccountID>(account);
-        if (!accid_raw)
+    auto mint = [&](auto const& account, XRPAmount const& amount) {
+        AccountID accid;
+
+        if constexpr (std::is_same_v<
+                          std::decay_t<decltype(account)>,
+                          std::string>)
         {
-            JLOG(j_.warn())
-                << "featureXahauGenesis could not parse an r-address: "
-                << account;
-            return;
+            // String path - parse it
+            auto accid_raw = parseBase58<AccountID>(account);
+            if (!accid_raw)
+            {
+                JLOG(j_.warn())
+                    << "featureXahauGenesis could not parse an r-address: "
+                    << account;
+                return;
+            }
+            accid = *accid_raw;
+        }
+        else
+        {
+            // Direct AccountID
+            accid = account;
         }
-
-        auto accid = *accid_raw;
-
         auto const kl = keylet::account(accid);
 
         auto sle = sb.peek(kl);
@@ -556,6 +583,21 @@ Change::activateXahauGenesis()
     for (auto const& [account, amount] : l1_entries)
         mint(account, amount);
 
+    // on the replay network (nid=65534) private keys 0 through 19999
+    // are populated for stress testing purposes, or they would be
+    // if the rest of the codebase allowed this, so we're only using 5000
+    // for now.
+    if (isReplayNetwork)
+    {
+        for (int i = 0; i < 5000; ++i)
+        {
+            uint8_t const* entry = replayNetworkAccIDs[i];
+            AccountID const acc = AccountID::fromVoid(entry);
+            JLOG(j_.info()) << "Replay Network AccID: " << acc;
+            mint(acc, XRPAmount{100});
+        }
+    }
+
     // Step 3: blackhole genesis
     sle->setAccountID(sfRegularKey, noAccount());
     sle->setFieldU32(sfFlags, lsfDisableMaster);

src/ripple/app/tx/impl/Import.cpp

@@ -167,6 +167,9 @@ Import::preflight(PreflightContext const& ctx)
     if (!xpop)
         return temMALFORMED;
 
+    if (ctx.app.config().NETWORK_ID == 65534 /* replay network */)
+        return tesSUCCESS;
+
     // we will check if we recognise the vl key in preclaim because it may be
     // from on-ledger object
     std::optional<PublicKey> masterVLKey;
@@ -270,7 +273,9 @@ Import::preflight(PreflightContext const& ctx)
         return temMALFORMED;
     }
 
-    if (stpTrans->getFieldU32(sfOperationLimit) != ctx.app.config().NETWORK_ID)
+    const auto nid = ctx.app.config().NETWORK_ID;
+    if (stpTrans->getFieldU32(sfOperationLimit) != nid &&
+        nid != 65534 /* replay network */)
     {
         JLOG(ctx.j.warn()) << "Import: Wrong network ID for OperationLimit in "
                               "inner txn. outer txid: "
@@ -1307,8 +1312,8 @@ Import::doApply()
             view().rules().enabled(featureXahauGenesis)
                 ? view().info().parentCloseTime.time_since_epoch().count()
                 : view().rules().enabled(featureDeletableAccounts)
-                    ? view().seq()
-                    : 1};
+                ? view().seq()
+                : 1};
 
         sle = std::make_shared<SLE>(keylet::account(id));
         sle->setAccountID(sfAccount, id);

src/ripple/app/tx/impl/Transactor.cpp

@@ -67,11 +67,16 @@ preflight0(PreflightContext const& ctx)
         else
         {
             // new networks both require the field to be present and require it
-            // to match
-            if (!txNID)
-                return telREQUIRES_NETWORK_ID;
+            // to match, except for some special networks
 
-            if (*txNID != nodeNID)
+            if (nodeNID == 65534 /* replay network */)
+            {
+                // on the replay network any other network's transactions can be
+                // replayed last ledger sequence is also ignored on this network
+            }
+            else if (!txNID)
+                return telREQUIRES_NETWORK_ID;
+            else if (*txNID != nodeNID)
                 return telWRONG_NETWORK;
         }
     }
@@ -115,8 +120,11 @@ preflight1(PreflightContext const& ctx)
     auto const fee = ctx.tx.getFieldAmount(sfFee);
     if (!fee.native() || fee.negative() || !isLegalAmount(fee.xrp()))
     {
-        JLOG(ctx.j.debug()) << "preflight1: invalid fee";
-        return temBAD_FEE;
+        if (ctx.app.config().NETWORK_ID != 65534 /* replay network */)
+        {
+            JLOG(ctx.j.debug()) << "preflight1: invalid fee";
+            return temBAD_FEE;
+        }
     }
 
     // if a hook emitted this transaction we bypass signature checks
@@ -432,6 +440,10 @@ Transactor::minimumFee(
 TER
 Transactor::checkFee(PreclaimContext const& ctx, XRPAmount baseFee)
 {
+    // on the replay network fees are unimportant
+    if (ctx.app.config().NETWORK_ID == 65534 /* replay network */)
+        return tesSUCCESS;
+
     if (!ctx.tx[sfFee].native())
         return temBAD_FEE;
 
@@ -473,6 +485,7 @@ Transactor::checkFee(PreclaimContext const& ctx, XRPAmount baseFee)
                        "a fee and an existing account.";
             }
         }
+        std::cout << "transactor 485 NO_ACCOUNT\n";
         return terNO_ACCOUNT;
     }
 
@@ -544,6 +557,7 @@ Transactor::checkSeqProxy(
         JLOG(j.trace())
             << "applyTransaction: delay: source account does not exist "
             << toBase58(id);
+        std::cout << "transactor 557 NO_ACCOUNT\n";
         return terNO_ACCOUNT;
     }
 
@@ -630,6 +644,7 @@ Transactor::checkPriorTxAndLastLedger(PreclaimContext const& ctx)
         JLOG(ctx.j.trace())
             << "applyTransaction: delay: source account does not exist "
             << toBase58(id);
+        std::cout << "transactor 644 NO_ACCOUNT\n";
         return terNO_ACCOUNT;
     }
 
@@ -641,9 +656,18 @@ Transactor::checkPriorTxAndLastLedger(PreclaimContext const& ctx)
             return tefWRONG_PRIOR;
     }
 
+    uint32_t nodeNID = ctx.app.config().NETWORK_ID;
+
     if (ctx.tx.isFieldPresent(sfLastLedgerSequence) &&
         (ctx.view.seq() > ctx.tx.getFieldU32(sfLastLedgerSequence)))
-        return tefMAX_LEDGER;
+    {
+        if (ctx.app.config().NETWORK_ID == 65534)
+        {
+            // on the replay network lls is ignored to allow txns to be replayed
+        }
+        else
+            return tefMAX_LEDGER;
+    }
 
     if (ctx.view.txExists(ctx.tx.getTransactionID()))
         return tefALREADY;
@@ -778,12 +802,14 @@ Transactor::apply()
 
     // If the transactor requires a valid account and the transaction doesn't
     // list one, preflight will have already a flagged a failure.
-    auto const sle = view().peek(keylet::account(account_));
+    auto sle = view().peek(keylet::account(account_));
+
+    const bool isReplayNetwork = (ctx_.app.config().NETWORK_ID == 65534);
 
     // sle must exist except for transactions
     // that allow zero account. (and ttIMPORT)
     assert(
-        sle != nullptr || account_ == beast::zero ||
+        sle != nullptr || account_ == beast::zero || isReplayNetwork ||
         view().rules().enabled(featureImport) &&
             ctx_.tx.getTxnType() == ttIMPORT &&
             !ctx_.tx.isFieldPresent(sfIssuer));
@@ -806,6 +832,39 @@ Transactor::apply()
 
         view().update(sle);
     }
+    else if (isReplayNetwork)
+    {
+        // create missing acc for replay network
+        // Create the account.
+        std::uint32_t const seqno{
+            view().rules().enabled(featureXahauGenesis)
+                ? view().info().parentCloseTime.time_since_epoch().count()
+                : view().rules().enabled(featureDeletableAccounts)
+                ? view().seq()
+                : 1};
+
+        sle = std::make_shared<SLE>(keylet::account(account_));
+        sle->setAccountID(sfAccount, account_);
+
+        sle->setFieldU32(sfSequence, seqno);
+        sle->setFieldU32(sfOwnerCount, 0);
+
+        if (view().exists(keylet::fees()) &&
+            view().rules().enabled(featureXahauGenesis))
+        {
+            auto sleFees = view().peek(keylet::fees());
+            uint64_t accIdx = sleFees->isFieldPresent(sfAccountCount)
+                ? sleFees->getFieldU64(sfAccountCount)
+                : 0;
+            sle->setFieldU64(sfAccountIndex, accIdx);
+            sleFees->setFieldU64(sfAccountCount, accIdx + 1);
+            view().update(sleFees);
+        }
+
+        // we'll fix this up at the end
+        sle->setFieldAmount(sfBalance, STAmount{XRPAmount{100}});
+        view().insert(sle);
+    }
 
     return doApply();
 }
@@ -828,7 +887,7 @@ Transactor::checkSign(PreclaimContext const& ctx)
 
     // wildcard network gets a free pass on all signatures
     if (ctx.tx.isFieldPresent(sfNetworkID) &&
-        ctx.tx.getFieldU32(sfNetworkID) == 65535)
+        ctx.tx.getFieldU32(sfNetworkID) >= 65534)
         return tesSUCCESS;
 
     // pass ttIMPORTs, their signatures are checked at the preflight against the
@@ -862,7 +921,19 @@ Transactor::checkSingleSign(PreclaimContext const& ctx)
     auto const sleAccount = ctx.view.read(keylet::account(idAccount));
 
     if (!sleAccount)
-        return terNO_ACCOUNT;
+    {
+        std::cout << "transactor 922 NO_ACCOUNT\n";
+
+        if (ctx.app.config().NETWORK_ID == 65534)
+        {
+            std::cout << "(success)\n";
+            // replay network allows transactions to create missing accounts
+            // implicitly and in this event we will just pass the txn
+            return tesSUCCESS;
+        }
+        else
+            return terNO_ACCOUNT;
+    }
 
     bool const isMasterDisabled = sleAccount->isFlag(lsfDisableMaster);
 
@@ -1928,7 +1999,9 @@ Transactor::operator()()
     {
         // Check invariants: if `tecINVARIANT_FAILED` is not returned, we can
         // proceed to apply the tx
-        result = ctx_.checkInvariants(result, fee);
+
+        if (ctx_.app.config().NETWORK_ID != 65534)
+            result = ctx_.checkInvariants(result, fee);
 
         if (result == tecINVARIANT_FAILED)
         {

src/ripple/app/tx/impl/apply.cpp

@@ -17,6 +17,7 @@
 */
 //==============================================================================
 
+#include <ripple/app/main/Application.h>
 #include <ripple/app/misc/HashRouter.h>
 #include <ripple/app/tx/apply.h>
 #include <ripple/app/tx/applySteps.h>
@@ -149,6 +150,12 @@ apply(
 
     auto pfresult = preflight(app, view.rules(), tx, flags, j);
     auto pcresult = preclaim(pfresult, app, view);
+    if (app.config().NETWORK_ID == 65534)
+    {
+        // replay network
+        if (pcresult.ter == terNO_ACCOUNT)
+            pcresult.ter = tesSUCCESS;
+    }
     return doApply(pcresult, app, view);
 }
 

src/ripple/app/tx/impl/applySteps.cpp

@@ -199,14 +199,19 @@ invoke_preclaim(PreclaimContext const& ctx)
     // list one, preflight will have already a flagged a failure.
     auto const id = ctx.tx.getAccountID(sfAccount);
 
+    bool const isReplayNetwork = (ctx.app.config().NETWORK_ID == 65534);
+
     if (id != beast::zero)
     {
-        TER result = T::checkSeqProxy(ctx.view, ctx.tx, ctx.j);
+        TER result = isReplayNetwork
+            ? tesSUCCESS
+            : T::checkSeqProxy(ctx.view, ctx.tx, ctx.j);
 
         if (!isTesSuccess(result))
             return result;
 
-        result = T::checkPriorTxAndLastLedger(ctx);
+        if (!isReplayNetwork)
+            result = T::checkPriorTxAndLastLedger(ctx);
 
         if (!isTesSuccess(result))
             return result;

src/ripple/basics/impl/contract.cpp

@@ -49,7 +49,7 @@ LogicError(std::string const& s) noexcept
 {
     JLOG(debugLog().fatal()) << s;
     std::cerr << "Logic error: " << s << std::endl;
-    detail::accessViolation();
+    //detail::accessViolation();
 }
 
 }  // namespace ripple

src/ripple/core/Config.h

@@ -174,6 +174,7 @@ public:
 
     // Network parameters
     uint32_t NETWORK_ID = 0;
+    uint16_t UDP_HIGHWAY_PORT = 0;  // this will be the first peer port
 
     // DEPRECATED - Fee units for a reference transction.
     // Only provided for backwards compatibility in a couple of places

src/ripple/ledger/OpenView.h

@@ -99,6 +99,12 @@ private:
     bool open_ = true;
 
 public:
+    std::set<uint256>
+    getAndResetKeysTouched()
+    {
+        return items_.getAndResetKeysTouched();
+    }
+
     OpenView() = delete;
     OpenView&
     operator=(OpenView&&) = delete;

src/ripple/ledger/detail/RawStateTable.h

@@ -35,6 +35,9 @@ namespace detail {
 // Helper class that buffers raw modifications
 class RawStateTable
 {
+private:
+    mutable std::set<uint256> keysTouched_;
+
 public:
     using key_type = ReadView::key_type;
     // Initial size for the monotonic_buffer_resource used for allocations
@@ -98,6 +101,20 @@ public:
     std::unique_ptr<ReadView::sles_type::iter_base>
     slesUpperBound(ReadView const& base, uint256 const& key) const;
 
+    // each time a key is read or written it will be placed in the keysTouched_
+    // set.
+    std::set<uint256>
+    getAndResetKeysTouched()
+    {
+        std::set<uint256> out;
+        out.swap(keysTouched_);
+        //        std::cout << "--------------\n";
+        //        for (auto const& k : out)
+        //            std::cout << "getAndResetKeysTouched: " << to_string(k) <<
+        //            "\n";
+        return out;
+    }
+
 private:
     enum class Action {
         erase,

src/ripple/ledger/impl/OpenView.cpp

@@ -263,6 +263,9 @@ OpenView::rawTxInsert(
     std::shared_ptr<Serializer const> const& txn,
     std::shared_ptr<Serializer const> const& metaData)
 {
+    if (txExists(key))
+        return;
+
     auto const result = txs_.emplace(
         std::piecewise_construct,
         std::forward_as_tuple(key),

src/ripple/ledger/impl/RawStateTable.cpp

@@ -173,6 +173,8 @@ RawStateTable::apply(RawView& to) const
                 to.rawReplace(item.sle);
                 break;
         }
+
+        keysTouched_.emplace(elem.first);
     }
 }
 
@@ -180,6 +182,9 @@ bool
 RawStateTable::exists(ReadView const& base, Keylet const& k) const
 {
     assert(k.key.isNonZero());
+
+    keysTouched_.insert(k.key);
+
     auto const iter = items_.find(k.key);
     if (iter == items_.end())
         return base.exists(k);
@@ -227,12 +232,18 @@ RawStateTable::succ(
     // what we got from the parent.
     if (last && next >= last)
         return std::nullopt;
+
+    if (next.has_value())
+        keysTouched_.insert(*next);
+
     return next;
 }
 
 void
 RawStateTable::erase(std::shared_ptr<SLE> const& sle)
 {
+    keysTouched_.insert(sle->key());
+
     // The base invariant is checked during apply
     auto const result = items_.emplace(
         std::piecewise_construct,
@@ -259,6 +270,7 @@ RawStateTable::erase(std::shared_ptr<SLE> const& sle)
 void
 RawStateTable::insert(std::shared_ptr<SLE> const& sle)
 {
+    keysTouched_.insert(sle->key());
     auto const result = items_.emplace(
         std::piecewise_construct,
         std::forward_as_tuple(sle->key()),
@@ -284,6 +296,7 @@ RawStateTable::insert(std::shared_ptr<SLE> const& sle)
 void
 RawStateTable::replace(std::shared_ptr<SLE> const& sle)
 {
+    keysTouched_.insert(sle->key());
     auto const result = items_.emplace(
         std::piecewise_construct,
         std::forward_as_tuple(sle->key()),
@@ -306,6 +319,7 @@ RawStateTable::replace(std::shared_ptr<SLE> const& sle)
 std::shared_ptr<SLE const>
 RawStateTable::read(ReadView const& base, Keylet const& k) const
 {
+    keysTouched_.insert(k.key);
     auto const iter = items_.find(k.key);
     if (iter == items_.end())
         return base.read(k);

src/ripple/overlay/Overlay.h

@@ -244,6 +244,16 @@ public:
      */
     virtual Json::Value
     txMetrics() const = 0;
+
+    /** Process incoming Xahau UDP Super Highway message */
+    virtual void
+    processXUSH(
+        std::string const& message,
+        boost::asio::ip::tcp::endpoint const& remoteEndpoint) = 0;
+
+    /** Send the txn to UDP Super Highway peers **/
+    virtual void
+    publishTxXUSH(Slice const& tx, uint256 const& txid) = 0;
 };
 
 }  // namespace ripple

src/ripple/overlay/impl/OverlayImpl.cpp

@@ -24,6 +24,7 @@
 #include <ripple/app/misc/ValidatorSite.h>
 #include <ripple/app/rdb/RelationalDatabase.h>
 #include <ripple/app/rdb/Wallet.h>
+#include <ripple/app/tx/apply.h>
 #include <ripple/basics/base64.h>
 #include <ripple/basics/make_SSLContext.h>
 #include <ripple/basics/random.h>
@@ -101,6 +102,7 @@ OverlayImpl::Timer::on_timer(error_code ec)
         return;
     }
 
+    std::cout << "on_timer\n";
     overlay_.m_peerFinder->once_per_second();
     overlay_.sendEndpoints();
     overlay_.autoConnect();
@@ -141,7 +143,8 @@ OverlayImpl::OverlayImpl(
           app.config().section(SECTION_RELATIONAL_DB).empty() ||
               !boost::iequals(
                   get(app.config().section(SECTION_RELATIONAL_DB), "backend"),
-                  "rwdb")))
+                  "rwdb"),
+          app))
     , m_resolver(resolver)
     , next_id_(1)
     , timer_count_(0)
@@ -1515,6 +1518,257 @@ OverlayImpl::deleteIdlePeers()
     slots_.deleteIdlePeers();
 }
 
+void
+OverlayImpl::processXUSH(
+    std::string const& message,
+    boost::asio::ip::tcp::endpoint const& remoteEndpoint)
+{
+    std::cout << "processXUSH\n";
+    // Fragment tracking: txid -> {endpoint, timestamp, total_size,
+    // fragments_received, data_map}
+    struct FragmentInfo
+    {
+        boost::asio::ip::tcp::endpoint sender;
+        uint32_t timestamp;
+        uint32_t total_size;
+        uint32_t num_fragments;
+        std::map<uint32_t, std::string> fragments;
+    };
+    static std::map<uint256, FragmentInfo> fragment_map;
+    static std::map<boost::asio::ip::tcp::endpoint, uint32_t> bad_sender_score;
+    static std::mt19937 rng{std::random_device{}()};
+
+    const uint8_t* data = reinterpret_cast<const uint8_t*>(message.data());
+    uint32_t now = std::time(nullptr);
+
+    // Opportunistic cleanup - check up to 10 random entries
+    if (!fragment_map.empty())
+    {
+        int checks_to_perform =
+            std::min(10, static_cast<int>(fragment_map.size()));
+
+        for (int i = 0; i < checks_to_perform; i++)
+        {
+            auto it = fragment_map.begin();
+            std::advance(
+                it,
+                std::uniform_int_distribution<>(
+                    0, fragment_map.size() - 1)(rng));
+
+            if (now - it->second.timestamp > 30)
+            {  // 30 second timeout
+                bad_sender_score[it->second.sender]++;
+                fragment_map.erase(it);
+            }
+        }
+    }
+
+    // XUSHPEER packet
+    if (message.size() >= 10 && std::memcmp(data, "XUSHPEER", 8) == 0)
+    {
+        std::cout << "\tXUSHPEER packet\n";
+        uint8_t ipv4_count = data[8];
+        uint8_t ipv6_count = data[9];
+        size_t expected_size = 10 + ipv4_count * 8 + ipv6_count * 20;
+
+        if (message.size() != expected_size)
+        {
+            bad_sender_score[remoteEndpoint]++;
+            return;
+        }
+
+        size_t offset = 10;
+        // Parse IPv4 addresses
+        std::vector<beast::IP::Endpoint> endpoints;
+        endpoints.reserve((uint32_t)ipv4_count + (uint32_t)ipv6_count);
+
+        for (int i = 0; i < ipv4_count; i++)
+        {
+            boost::asio::ip::address_v4::bytes_type addr_bytes;
+            std::memcpy(addr_bytes.data(), data + offset, 4);
+
+            beast::IP::Address addr{boost::asio::ip::address_v4(addr_bytes)};
+
+            // Read port
+            uint32_t port_32 =
+                ntohl(*reinterpret_cast<const uint32_t*>(data + offset + 4));
+            beast::IP::Port port = static_cast<beast::IP::Port>(port_32);
+            offset += 8;
+
+            // Create endpoint
+            beast::IP::Endpoint endpoint(addr, port);
+
+            endpoints.push_back(endpoint);
+        }
+
+        // Parse IPv6 addresses
+        for (int i = 0; i < ipv6_count; i++)
+        {
+            boost::asio::ip::address_v6::bytes_type addr_bytes;
+            std::memcpy(addr_bytes.data(), data + offset, 16);
+
+            // Use extra parentheses or brace initialization
+            beast::IP::Address addr((boost::asio::ip::address_v6(addr_bytes)));
+            // Or: beast::IP::Address
+            // addr{boost::asio::ip::address_v6(addr_bytes)};
+
+            // Read port
+            uint32_t port_32 =
+                ntohl(*reinterpret_cast<const uint32_t*>(data + offset + 16));
+            beast::IP::Port port = static_cast<beast::IP::Port>(port_32);
+            offset += 20;
+
+            // Create endpoint
+            beast::IP::Endpoint endpoint(addr, port);
+
+            endpoints.push_back(endpoint);
+        }
+
+        m_peerFinder->add_highway_peers(endpoints);
+    }
+    // XUSHTXNF packet (fragmented transaction)
+    else if (message.size() >= 52 && std::memcmp(data, "XUSHTXNF", 8) == 0)
+    {
+        std::cout << "\tXUSHTXNF packet\n";
+        uint256 txid{
+            uint256::fromVoid(reinterpret_cast<const char*>(data + 8))};
+        uint32_t total_size =
+            ntohl(*reinterpret_cast<const uint32_t*>(data + 40));
+        uint32_t num_fragments =
+            ntohl(*reinterpret_cast<const uint32_t*>(data + 44));
+        uint32_t fragment_num =
+            ntohl(*reinterpret_cast<const uint32_t*>(data + 48));
+
+        if (fragment_num >= num_fragments || total_size > 1048576 * 2)
+            return;  // 2MB limit
+
+        // Mute bad senders progressively
+        if (bad_sender_score[remoteEndpoint] > 10)
+        {
+            if (std::uniform_int_distribution<>(
+                    0, bad_sender_score[remoteEndpoint])(rng) > 10)
+                return;
+        }
+
+        auto& info = fragment_map[txid];
+        if (info.fragments.empty())
+        {
+            info.sender = remoteEndpoint;
+            info.timestamp = now;
+            info.total_size = total_size;
+            info.num_fragments = num_fragments;
+        }
+
+        int flags = app_.getHashRouter().getFlags(txid);
+
+        if (flags & SF_BAD)
+        {
+            bad_sender_score[remoteEndpoint]++;
+            fragment_map.erase(txid);
+            return;
+        }
+
+        // Store fragment
+        info.fragments[fragment_num] = std::string(
+            reinterpret_cast<const char*>(data + 52), message.size() - 52);
+
+        // Check if complete
+        if (info.fragments.size() == info.num_fragments)
+        {
+            std::string complete_tx;
+            complete_tx.reserve(info.total_size);
+            for (uint32_t i = 0; i < info.num_fragments; i++)
+            {
+                complete_tx += info.fragments[i];
+            }
+
+            if (complete_tx.size() == info.total_size)
+            {
+                // Process complete transaction
+
+                Slice txSlice(complete_tx.data(), complete_tx.size());
+                SerialIter sit(txSlice);
+
+                try
+                {
+                    auto stx = std::make_shared<STTx const>(sit);
+                    uint256 computedTxid = stx->getTransactionID();
+
+                    std::cout << "XUSH txn complete " << strHex(computedTxid)
+                              << "\n";
+
+                    // if txn is corrupt (wrong txid) or an emitted txn, or
+                    // can't make it into a ledger bill the sender and drop
+                    if (txid != computedTxid ||
+                        stx->isFieldPresent(sfEmitDetails) ||
+                        (stx->isFieldPresent(sfLastLedgerSequence) &&
+                         (stx->getFieldU32(sfLastLedgerSequence) <
+                          app_.getLedgerMaster().getValidLedgerIndex())))
+                    {
+                        bad_sender_score[remoteEndpoint]++;
+                        fragment_map.erase(txid);
+                        return;
+                    }
+
+                    // Check the signature
+                    if (auto [valid, validReason] = checkValidity(
+                            app_.getHashRouter(),
+                            *stx,
+                            app_.getLedgerMaster().getValidatedRules(),
+                            app_.config());
+                        valid != Validity::Valid)
+                    {
+                        if (!validReason.empty())
+                        {
+                            JLOG(journal_.trace())
+                                << "Exception checking transaction: "
+                                << validReason;
+                        }
+
+                        app_.getHashRouter().setFlags(
+                            stx->getTransactionID(), SF_BAD);
+                        bad_sender_score[remoteEndpoint]++;
+                        fragment_map.erase(txid);
+                        return;
+                    }
+
+                    // execution to here means the txn passed basic checks
+                    // machine gun it to peers over the highway
+
+                    m_peerFinder->machine_gun_highway_peers(txSlice, txid);
+
+                    // add it to our own node for processing
+                    std::string reason;
+                    auto tpTrans =
+                        std::make_shared<Transaction>(stx, reason, app_);
+                    if (tpTrans->getStatus() != NEW)
+                        return;
+
+                    app_.getOPs().processTransaction(tpTrans, false, false);
+
+                    return;
+                }
+                catch (std::exception const& ex)
+                {
+                    JLOG(journal_.warn())
+                        << "Transaction invalid: " << strHex(txSlice)
+                        << ". Exception: " << ex.what();
+                }
+            }
+
+            // successful reconstruction would have returned before here
+            bad_sender_score[remoteEndpoint]++;
+            fragment_map.erase(txid);
+        }
+    }
+}
+
+void
+OverlayImpl::publishTxXUSH(Slice const& tx, uint256 const& txid)
+{
+    m_peerFinder->machine_gun_highway_peers(tx, txid);
+}
+
 //------------------------------------------------------------------------------
 
 Overlay::Setup

src/ripple/overlay/impl/OverlayImpl.h

@@ -454,6 +454,14 @@ public:
         txMetrics_.addMetrics(args...);
     }
 
+    void
+    processXUSH(
+        std::string const& message,
+        boost::asio::ip::tcp::endpoint const& remoteEndpoint) override;
+
+    void
+    publishTxXUSH(Slice const& tx, uint256 const& txid) override;
+
 private:
     void
     squelch(

src/ripple/peerfinder/PeerfinderManager.h

@@ -20,11 +20,13 @@
 #ifndef RIPPLE_PEERFINDER_MANAGER_H_INCLUDED
 #define RIPPLE_PEERFINDER_MANAGER_H_INCLUDED
 
+#include <ripple/app/main/Application.h>
 #include <ripple/beast/clock/abstract_clock.h>
 #include <ripple/beast/utility/PropertyStream.h>
 #include <ripple/core/Config.h>
 #include <ripple/peerfinder/Slot.h>
 #include <boost/asio/ip/tcp.hpp>
+#include <ranges>
 
 namespace ripple {
 namespace PeerFinder {
@@ -141,6 +143,12 @@ protected:
     Manager() noexcept;
 
 public:
+    virtual void
+    add_highway_peers(std::vector<beast::IP::Endpoint> addresses) = 0;
+
+    virtual void
+    machine_gun_highway_peers(Slice const& tx, uint256 const& txid) = 0;
+
     /** Destroy the object.
         Any pending source fetch operations are aborted.
         There may be some listener calls made before the

src/ripple/peerfinder/impl/Logic.h

@@ -1035,6 +1035,7 @@ public:
     int
     addBootcacheAddresses(IPAddresses const& list)
     {
+        // RHUPTO: add_highway_peers(
         int count(0);
         std::lock_guard _(lock_);
         for (auto addr : list)

src/ripple/peerfinder/impl/PeerfinderManager.cpp

@@ -17,6 +17,7 @@
 */
 //==============================================================================
 
+#include <ripple/app/main/Application.h>
 #include <ripple/core/ConfigSections.h>
 #include <ripple/peerfinder/PeerfinderManager.h>
 #include <ripple/peerfinder/impl/Checker.h>
@@ -35,6 +36,16 @@ namespace PeerFinder {
 
 class ManagerImp : public Manager
 {
+protected:
+    Application& app_;
+
+    std::map<
+        beast::IP::Endpoint /* udp endpoint */,
+        uint32_t /* unixtime last seen */>
+        m_udp_highway_peers;
+
+    std::mutex m_udp_highway_mutex;
+
 public:
     boost::asio::io_service& io_service_;
     std::optional<boost::asio::io_service::work> work_;
@@ -45,7 +56,133 @@ public:
     Logic<decltype(checker_)> m_logic;
     BasicConfig const& m_config;
 
-    //--------------------------------------------------------------------------
+    void
+    add_highway_peers(std::vector<beast::IP::Endpoint> addresses) override
+    {
+        std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+        uint32_t t = static_cast<uint32_t>(std::time(nullptr));
+        for (auto const& a : addresses)
+            m_udp_highway_peers.emplace(a, t);
+    }
+
+    /* send a transaction over datagram to a large random subset of highway
+     * peers */
+    void
+    machine_gun_highway_peers(Slice const& tx, uint256 const& txid) override
+    {
+        constexpr size_t kMaxDatagram = 65535;
+        constexpr size_t kUDPHeader = 8;
+        constexpr size_t kIPv4Header = 20;
+        constexpr size_t kIPv6Header = 40;
+        constexpr size_t kHeaderSize = 52;
+
+        // Create sockets once
+        static thread_local int udp4_sock = [this]() {
+            int s = ::socket(AF_INET, SOCK_DGRAM | SOCK_NONBLOCK, 0);
+            sockaddr_in addr{};
+            addr.sin_family = AF_INET;
+            addr.sin_port = htons(app_.config().UDP_HIGHWAY_PORT);
+            addr.sin_addr.s_addr = INADDR_ANY;
+            ::bind(s, reinterpret_cast<sockaddr*>(&addr), sizeof(addr));
+            return s;
+        }();
+
+        static thread_local int udp6_sock = [this]() {
+            int s = ::socket(AF_INET6, SOCK_DGRAM | SOCK_NONBLOCK, 0);
+            sockaddr_in6 addr{};
+            addr.sin6_family = AF_INET6;
+            addr.sin6_port = htons(app_.config().UDP_HIGHWAY_PORT);
+            addr.sin6_addr = in6addr_any;
+            ::bind(s, reinterpret_cast<sockaddr*>(&addr), sizeof(addr));
+            return s;
+        }();
+
+        std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+        if (m_udp_highway_peers.empty())
+            return;
+
+        // Select ~50% of peers randomly
+        static thread_local std::mt19937 rng{std::random_device{}()};
+        std::vector<beast::IP::Endpoint> targets;
+
+        auto sample_peers = [&](auto& map, auto& targets, auto& rng) {
+            auto n = map.size();
+            if (n == 0)
+                return;
+            auto k = n / 2 + 1;
+            std::vector<beast::IP::Endpoint> keys;
+            keys.reserve(n);
+            for (auto const& p : map)
+                keys.push_back(p.first);
+            std::shuffle(keys.begin(), keys.end(), rng);
+            targets.assign(keys.begin(), keys.begin() + std::min(k, n));
+        };
+
+        sample_peers(m_udp_highway_peers, targets, rng);
+
+        // Determine max payload size based on endpoint types
+        size_t max_data_v4 =
+            kMaxDatagram - kIPv4Header - kUDPHeader - kHeaderSize;
+        size_t max_data_v6 =
+            kMaxDatagram - kIPv6Header - kUDPHeader - kHeaderSize;
+
+        bool has_v6 = std::any_of(targets.begin(), targets.end(), [](auto& ep) {
+            return ep.address().is_v6();
+        });
+        size_t max_data_size = has_v6 ? max_data_v6 : max_data_v4;
+
+        uint32_t num_fragments =
+            (tx.size() + max_data_size - 1) / max_data_size;
+
+        for (uint32_t i = 0; i < num_fragments; ++i)
+        {
+            size_t offset = i * max_data_size;
+            size_t chunk_size = std::min(max_data_size, tx.size() - offset);
+
+            std::vector<uint8_t> packet(kHeaderSize + chunk_size);
+            std::memcpy(packet.data(), "XUSHTXNF", 8);
+            std::memcpy(packet.data() + 8, txid.data(), 32);
+            *reinterpret_cast<uint32_t*>(packet.data() + 40) = htonl(tx.size());
+            *reinterpret_cast<uint32_t*>(packet.data() + 44) =
+                htonl(num_fragments);
+            *reinterpret_cast<uint32_t*>(packet.data() + 48) = htonl(i);
+            std::memcpy(packet.data() + 52, tx.data() + offset, chunk_size);
+
+            for (auto& endpoint : targets)
+            {
+                if (endpoint.address().is_v4())
+                {
+                    sockaddr_in addr{};
+                    addr.sin_family = AF_INET;
+                    addr.sin_port = htons(endpoint.port());
+                    addr.sin_addr.s_addr =
+                        htonl(endpoint.address().to_v4().to_uint());
+                    ::sendto(
+                        udp4_sock,
+                        packet.data(),
+                        packet.size(),
+                        0,
+                        reinterpret_cast<sockaddr*>(&addr),
+                        sizeof(addr));
+                }
+                else
+                {
+                    sockaddr_in6 addr{};
+                    addr.sin6_family = AF_INET6;
+                    addr.sin6_port = htons(endpoint.port());
+                    auto bytes = endpoint.address().to_v6().to_bytes();
+                    std::memcpy(&addr.sin6_addr, bytes.data(), 16);
+                    ::sendto(
+                        udp6_sock,
+                        packet.data(),
+                        packet.size(),
+                        0,
+                        reinterpret_cast<sockaddr*>(&addr),
+                        sizeof(addr));
+                }
+            }
+        }
+    }
 
     ManagerImp(
         boost::asio::io_service& io_service,
@@ -53,8 +190,10 @@ public:
         beast::Journal journal,
         BasicConfig const& config,
         beast::insight::Collector::ptr const& collector,
-        bool useSqLiteStore)
+        bool useSqLiteStore,
+        Application& app)
         : Manager()
+        , app_(app)
         , io_service_(io_service)
         , work_(std::in_place, std::ref(io_service_))
         , m_clock(clock)
@@ -108,6 +247,14 @@ public:
         std::string const& name,
         std::vector<beast::IP::Endpoint> const& addresses) override
     {
+        // add fixed peers to superhighway
+        {
+            std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+            uint32_t t = static_cast<uint32_t>(std::time(nullptr));
+            for (auto const& a : addresses)
+                m_udp_highway_peers.emplace(a, t);
+        }
+
         m_logic.addFixedPeer(name, addresses);
     }
 
@@ -145,6 +292,14 @@ public:
     on_endpoints(std::shared_ptr<Slot> const& slot, Endpoints const& endpoints)
         override
     {
+        // add endpoints to superhighway
+        {
+            std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+            uint32_t t = static_cast<uint32_t>(std::time(nullptr));
+            for (auto const& a : endpoints)
+                m_udp_highway_peers.emplace(a.address, t);
+        }
+
         SlotImp::ptr impl(std::dynamic_pointer_cast<SlotImp>(slot));
         m_logic.on_endpoints(impl, endpoints);
     }
@@ -168,6 +323,15 @@ public:
         boost::asio::ip::tcp::endpoint const& remote_address,
         std::vector<boost::asio::ip::tcp::endpoint> const& eps) override
     {
+        // add redirects to superhighway
+        {
+            std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+            uint32_t t = static_cast<uint32_t>(std::time(nullptr));
+            for (auto const& a : eps)
+                m_udp_highway_peers.emplace(
+                    beast::IPAddressConversion::from_asio(a), t);
+        }
+
         m_logic.onRedirects(eps.begin(), eps.end(), remote_address);
     }
 
@@ -209,6 +373,127 @@ public:
     once_per_second() override
     {
         m_logic.once_per_second();
+
+        // clean superhighway in an amortized fashion
+        static std::mt19937 rng(std::random_device{}());
+        {
+            std::lock_guard<std::mutex> lock(m_udp_highway_mutex);
+
+            uint32_t t = static_cast<uint32_t>(std::time(nullptr));
+
+            std::cout << "m_udp_highway_peers size="
+                      << m_udp_highway_peers.size() << "\n";
+            for (auto const& [ep, ls] : m_udp_highway_peers)
+            {
+                std::cout << "ep: " << ep << " ls: " << ls << "\n";
+            }
+
+            for (int i = 0; i < std::min(3, (int)m_udp_highway_peers.size());
+                 ++i)
+            {
+                auto it = std::next(
+                    m_udp_highway_peers.begin(),
+                    std::uniform_int_distribution<>(
+                        0, m_udp_highway_peers.size() - 1)(rng));
+
+                /* highway peers we haven't seen anything from for 500 seconds
+                 * are removed */
+                if (t - it->second > 500)
+                    m_udp_highway_peers.erase(it);
+            }
+        }
+
+        // we will also randomly choose some peers and send our peer list
+        {
+            static boost::asio::io_context io_ctx;
+            static boost::asio::ip::udp::socket sock(
+                io_ctx, boost::asio::ip::udp::v4());
+
+            if (m_udp_highway_peers.empty())
+                return;
+
+            // Lambda to randomly sample N items from map
+            auto sample = [&](size_t n) {
+                std::vector<std::pair<beast::IP::Endpoint, uint32_t>> vec(
+                    m_udp_highway_peers.begin(), m_udp_highway_peers.end());
+                std::shuffle(vec.begin(), vec.end(), rng);
+                vec.resize(std::min(n, vec.size()));
+                return vec;
+            };
+
+            // Select up to 100 peers to encode
+            auto peers_to_encode = sample(100);
+
+            // Lambda to encode peers into binary packet
+            // Format: [8 bytes: "XUSHPEER"][1 byte: IPv4 count][1 byte: IPv6
+            // count][IPv4s][IPv6s]
+            auto encode = [](const auto& peers) {
+                std::vector<uint8_t> packet;
+                packet.reserve(
+                    10 +
+                    peers.size() * 24);  // 8 magic + 2 header + max peer data
+
+                // Separate IPv4 and IPv6
+                std::vector<const beast::IP::Endpoint*> ipv4s, ipv6s;
+                for (const auto& p : peers)
+                {
+                    if (p.first.address().is_v4())
+                        ipv4s.push_back(&p.first);
+                    else
+                        ipv6s.push_back(&p.first);
+                }
+
+                // Magic code: XUSHPEER
+                const char* magic = "XUSHPEER";
+                packet.insert(packet.end(), magic, magic + 8);
+
+                // Header: [IPv4 count][IPv6 count]
+                packet.push_back(static_cast<uint8_t>(ipv4s.size()));
+                packet.push_back(static_cast<uint8_t>(ipv6s.size()));
+
+                // Pack IPv4s (4 bytes IP + 4 bytes port)
+                for (auto ep : ipv4s)
+                {
+                    auto v4 = ep->address().to_v4().to_bytes();
+                    packet.insert(packet.end(), v4.begin(), v4.end());
+                    uint32_t port = htonl(ep->port());
+                    packet.insert(
+                        packet.end(), (uint8_t*)&port, (uint8_t*)&port + 4);
+                }
+
+                // Pack IPv6s (16 bytes IP + 4 bytes port)
+                for (auto ep : ipv6s)
+                {
+                    auto v6 = ep->address().to_v6().to_bytes();
+                    packet.insert(packet.end(), v6.begin(), v6.end());
+                    uint32_t port = htonl(ep->port());
+                    packet.insert(
+                        packet.end(), (uint8_t*)&port, (uint8_t*)&port + 4);
+                }
+
+                return packet;
+            };
+
+            auto packet = encode(peers_to_encode);
+
+            // Select 20 peers to send to (re-roll, overlap is fine)
+            auto targets = sample(20);
+
+            // Send packet to each target
+            for (const auto& [endpoint, _] : targets)
+            {
+                try
+                {
+                    boost::asio::ip::udp::endpoint udp_ep(
+                        endpoint.address(), endpoint.port());
+                    sock.send_to(boost::asio::buffer(packet), udp_ep);
+                }
+                catch (...)
+                {
+                    // Silent fail, continue to next peer
+                }
+            }
+        }
     }
 
     std::vector<std::pair<std::shared_ptr<Slot>, std::vector<Endpoint>>>
@@ -282,10 +567,11 @@ make_Manager(
     beast::Journal journal,
     BasicConfig const& config,
     beast::insight::Collector::ptr const& collector,
-    bool useSqLiteStore)
+    bool useSqLiteStore,
+    Application& app)
 {
     return std::make_unique<ManagerImp>(
-        io_service, clock, journal, config, collector, useSqLiteStore);
+        io_service, clock, journal, config, collector, useSqLiteStore, app);
 }
 
 }  // namespace PeerFinder

src/ripple/peerfinder/make_Manager.h

@@ -35,7 +35,8 @@ make_Manager(
     beast::Journal journal,
     BasicConfig const& config,
     beast::insight::Collector::ptr const& collector,
-    bool useSqliteStore);
+    bool useSqliteStore,
+    Application& app);
 
 }  // namespace PeerFinder
 }  // namespace ripple

src/ripple/protocol/impl/STTx.cpp

@@ -302,7 +302,7 @@ STTx::checkSingleSign(RequireFullyCanonicalSig requireCanonicalSig) const
 
     // wildcard network gets a free pass on all signatures
     bool const isWildcardNetwork =
-        isFieldPresent(sfNetworkID) && getFieldU32(sfNetworkID) == 65535;
+        isFieldPresent(sfNetworkID) && getFieldU32(sfNetworkID) >= 65534;
 
     bool validSig = false;
     try

src/ripple/protocol/jss.h

@@ -669,18 +669,19 @@ JSS(strict);                // in: AccountCurrencies, AccountInfo
 JSS(sub_index);             // in: LedgerEntry
 JSS(subcommand);            // in: PathFind
 JSS(success);               // rpc
-JSS(supported);             // out: AmendmentTableImpl
-JSS(system_time_offset);    // out: NetworkOPs
-JSS(tag);                   // out: Peers
-JSS(taker);                 // in: Subscribe, BookOffers
-JSS(taker_gets);            // in: Subscribe, Unsubscribe, BookOffers
-JSS(taker_gets_funded);     // out: NetworkOPs
-JSS(taker_pays);            // in: Subscribe, Unsubscribe, BookOffers
-JSS(taker_pays_funded);     // out: NetworkOPs
-JSS(threshold);             // in: Blacklist
-JSS(ticket);                // in: AccountObjects
-JSS(ticket_count);          // out: AccountInfo
-JSS(ticket_seq);            // in: LedgerEntry
+JSS(success_count);
+JSS(supported);           // out: AmendmentTableImpl
+JSS(system_time_offset);  // out: NetworkOPs
+JSS(tag);                 // out: Peers
+JSS(taker);               // in: Subscribe, BookOffers
+JSS(taker_gets);          // in: Subscribe, Unsubscribe, BookOffers
+JSS(taker_gets_funded);   // out: NetworkOPs
+JSS(taker_pays);          // in: Subscribe, Unsubscribe, BookOffers
+JSS(taker_pays_funded);   // out: NetworkOPs
+JSS(threshold);           // in: Blacklist
+JSS(ticket);              // in: AccountObjects
+JSS(ticket_count);        // out: AccountInfo
+JSS(ticket_seq);          // in: LedgerEntry
 JSS(time);
 JSS(timeouts);                // out: InboundLedger
 JSS(track);                   // out: PeerImp
@@ -704,11 +705,13 @@ JSS(trusted);                 // out: UnlList
 JSS(trusted_validator_keys);  // out: ValidatorList
 JSS(tx);                      // out: STTx, AccountTx*
 JSS(txroot);
-JSS(tx_blob);               // in/out: Submit,
-                            // in: TransactionSign, AccountTx*
-JSS(tx_hash);               // in: TransactionEntry
-JSS(tx_json);               // in/out: TransactionSign
-                            // out: TransactionEntry
+JSS(tx_blob);  // in/out: Submit,
+JSS(tx_blobs);
+// in: TransactionSign, AccountTx*
+JSS(tx_hash);  // in: TransactionEntry
+JSS(tx_json);  // in/out: TransactionSign
+               // out: TransactionEntry
+JSS(tx_results);
 JSS(tx_signing_hash);       // out: TransactionSign
 JSS(tx_unsigned);           // out: TransactionSign
 JSS(txn_count);             // out: NetworkOPs

src/ripple/rpc/handlers/Submit.cpp

@@ -23,12 +23,16 @@
 #include <ripple/app/misc/TxQ.h>
 #include <ripple/app/tx/apply.h>
 #include <ripple/net/RPCErr.h>
+#include <ripple/overlay/Overlay.h>
 #include <ripple/protocol/ErrorCodes.h>
 #include <ripple/resource/Fees.h>
 #include <ripple/rpc/Context.h>
 #include <ripple/rpc/GRPCHandlers.h>
 #include <ripple/rpc/impl/RPCHelpers.h>
 #include <ripple/rpc/impl/TransactionSign.h>
+#include <future>
+#include <thread>
+#include <vector>
 
 namespace ripple {
 
@@ -82,15 +86,225 @@ doInject(RPC::JsonContext& context)
     return jvResult;
 }
 
+// Helper function to process a single transaction blob
+static Json::Value
+processSingleTransaction(
+    RPC::JsonContext& context,
+    const std::string& txBlob,
+    const NetworkOPs::FailHard& failType)
+{
+    Json::Value result;
+
+    auto ret = strUnHex(txBlob);
+    if (!ret || !ret->size())
+    {
+        result[jss::error] = "invalidTransaction";
+        result[jss::error_exception] = "Invalid hex encoding";
+        return result;
+    }
+
+    SerialIter sitTrans(makeSlice(*ret));
+    std::shared_ptr<STTx const> stpTrans;
+
+    try
+    {
+        stpTrans = std::make_shared<STTx const>(std::ref(sitTrans));
+
+        // RH TODO: basic signature check here to prevent abuse
+        // send over the UDP superhighway if it parsed correctly
+        context.app.overlay().publishTxXUSH(
+            makeSlice(*ret), stpTrans->getTransactionID());
+    }
+    catch (std::exception& e)
+    {
+        result[jss::error] = "invalidTransaction";
+        result[jss::error_exception] = e.what();
+        return result;
+    }
+
+    // Validity check
+    {
+        if (!context.app.checkSigs())
+            forceValidity(
+                context.app.getHashRouter(),
+                stpTrans->getTransactionID(),
+                Validity::SigGoodOnly);
+        auto [validity, reason] = checkValidity(
+            context.app.getHashRouter(),
+            *stpTrans,
+            context.ledgerMaster.getCurrentLedger()->rules(),
+            context.app.config());
+        if (validity != Validity::Valid)
+        {
+            result[jss::error] = "invalidTransaction";
+            result[jss::error_exception] = "fails local checks: " + reason;
+            return result;
+        }
+    }
+
+    std::string reason;
+    auto tpTrans = std::make_shared<Transaction>(stpTrans, reason, context.app);
+    if (tpTrans->getStatus() != NEW)
+    {
+        result[jss::error] = "invalidTransaction";
+        result[jss::error_exception] = "fails local checks: " + reason;
+        return result;
+    }
+
+    try
+    {
+        context.netOps.processTransaction(
+            tpTrans, isUnlimited(context.role), true, failType);
+    }
+    catch (std::exception& e)
+    {
+        result[jss::error] = "internalSubmit";
+        result[jss::error_exception] = e.what();
+        return result;
+    }
+
+    try
+    {
+        result[jss::tx_json] = tpTrans->getJson(JsonOptions::none);
+        result[jss::tx_blob] =
+            strHex(tpTrans->getSTransaction()->getSerializer().peekData());
+
+        if (temUNCERTAIN != tpTrans->getResult())
+        {
+            std::string sToken;
+            std::string sHuman;
+
+            transResultInfo(tpTrans->getResult(), sToken, sHuman);
+
+            result[jss::engine_result] = sToken;
+            result[jss::engine_result_code] = tpTrans->getResult();
+            result[jss::engine_result_message] = sHuman;
+
+            auto const submitResult = tpTrans->getSubmitResult();
+
+            result[jss::accepted] = submitResult.any();
+            result[jss::applied] = submitResult.applied;
+            result[jss::broadcast] = submitResult.broadcast;
+            result[jss::queued] = submitResult.queued;
+            result[jss::kept] = submitResult.kept;
+
+            if (auto currentLedgerState = tpTrans->getCurrentLedgerState())
+            {
+                result[jss::account_sequence_next] =
+                    safe_cast<Json::Value::UInt>(
+                        currentLedgerState->accountSeqNext);
+                result[jss::account_sequence_available] =
+                    safe_cast<Json::Value::UInt>(
+                        currentLedgerState->accountSeqAvail);
+                result[jss::open_ledger_cost] =
+                    to_string(currentLedgerState->minFeeRequired);
+                result[jss::validated_ledger_index] =
+                    safe_cast<Json::Value::UInt>(
+                        currentLedgerState->validatedLedger);
+            }
+        }
+
+        return result;
+    }
+    catch (std::exception& e)
+    {
+        result[jss::error] = "internalJson";
+        result[jss::error_exception] = e.what();
+        return result;
+    }
+}
+
 // {
 //   tx_json: <object>,
 //   secret: <secret>
 // }
+// OR for batch submission:
+// {
+//   "tx_blobs": [<blob1>, <blob2>, ...],
+// }
 Json::Value
 doSubmit(RPC::JsonContext& context)
 {
     context.loadType = Resource::feeMediumBurdenRPC;
 
+    // Check for batch submission
+    if (context.params.isMember("tx_blobs"))
+    {
+        if (!context.params["tx_blobs"].isArray())
+            return rpcError(rpcINVALID_PARAMS);
+
+        const auto& txBlobs = context.params["tx_blobs"];
+        const auto blobCount = txBlobs.size();
+
+        if (blobCount == 0)
+            return rpcError(rpcINVALID_PARAMS);
+
+        // Limit batch size to prevent resource exhaustion
+        constexpr size_t maxBatchSize = 100;
+        if (blobCount > maxBatchSize)
+        {
+            Json::Value error;
+            error[jss::error] = "batchSizeExceeded";
+            error["error_message"] =
+                "Batch size exceeds maximum of " + std::to_string(maxBatchSize);
+            return error;
+        }
+
+        auto const failType = getFailHard(context);
+
+        // Process transactions in parallel
+        std::vector<std::future<Json::Value>> futures;
+        futures.reserve(blobCount);
+
+        // Launch async tasks for each transaction
+        for (size_t i = 0; i < blobCount; ++i)
+        {
+            if (!txBlobs[i].isString())
+            {
+                // Create error result for invalid blob
+                std::promise<Json::Value> errorPromise;
+                Json::Value errorResult;
+                errorResult[jss::error] = "invalidTransaction";
+                errorResult[jss::error_exception] =
+                    "tx_blobs element must be string";
+                errorPromise.set_value(std::move(errorResult));
+                futures.push_back(errorPromise.get_future());
+                continue;
+            }
+
+            const std::string txBlobStr = txBlobs[i].asString();
+            futures.push_back(std::async(
+                std::launch::async, [&context, txBlobStr, failType]() {
+                    return processSingleTransaction(
+                        context, txBlobStr, failType);
+                }));
+        }
+
+        // Collect results
+        Json::Value jvResult;
+        Json::Value& results = jvResult["tx_results"] = Json::arrayValue;
+
+        for (auto& future : futures)
+        {
+            results.append(future.get());
+        }
+
+        jvResult["batch_count"] = static_cast<Json::UInt>(blobCount);
+
+        // Count successful submissions
+        Json::UInt successCount = 0;
+        for (const auto& result : results)
+        {
+            std::cout << result << "\n";
+            if (!result.isMember(jss::error))
+                ++successCount;
+        }
+        jvResult["success_count"] = successCount;
+
+        return jvResult;
+    }
+
+    // Single transaction submission (original code path)
     if (!context.params.isMember(jss::tx_blob))
     {
         auto const failType = getFailHard(context);
@@ -116,124 +330,10 @@ doSubmit(RPC::JsonContext& context)
         return ret;
     }
 
-    Json::Value jvResult;
-
-    auto ret = strUnHex(context.params[jss::tx_blob].asString());
-
-    if (!ret || !ret->size())
-        return rpcError(rpcINVALID_PARAMS);
-
-    SerialIter sitTrans(makeSlice(*ret));
-
-    std::shared_ptr<STTx const> stpTrans;
-
-    try
-    {
-        stpTrans = std::make_shared<STTx const>(std::ref(sitTrans));
-    }
-    catch (std::exception& e)
-    {
-        jvResult[jss::error] = "invalidTransaction";
-        jvResult[jss::error_exception] = e.what();
-
-        return jvResult;
-    }
-
-    {
-        if (!context.app.checkSigs())
-            forceValidity(
-                context.app.getHashRouter(),
-                stpTrans->getTransactionID(),
-                Validity::SigGoodOnly);
-        auto [validity, reason] = checkValidity(
-            context.app.getHashRouter(),
-            *stpTrans,
-            context.ledgerMaster.getCurrentLedger()->rules(),
-            context.app.config());
-        if (validity != Validity::Valid)
-        {
-            jvResult[jss::error] = "invalidTransaction";
-            jvResult[jss::error_exception] = "fails local checks: " + reason;
-
-            return jvResult;
-        }
-    }
-
-    std::string reason;
-    auto tpTrans = std::make_shared<Transaction>(stpTrans, reason, context.app);
-    if (tpTrans->getStatus() != NEW)
-    {
-        jvResult[jss::error] = "invalidTransaction";
-        jvResult[jss::error_exception] = "fails local checks: " + reason;
-
-        return jvResult;
-    }
-
-    try
-    {
-        auto const failType = getFailHard(context);
-
-        context.netOps.processTransaction(
-            tpTrans, isUnlimited(context.role), true, failType);
-    }
-    catch (std::exception& e)
-    {
-        jvResult[jss::error] = "internalSubmit";
-        jvResult[jss::error_exception] = e.what();
-
-        return jvResult;
-    }
-
-    try
-    {
-        jvResult[jss::tx_json] = tpTrans->getJson(JsonOptions::none);
-        jvResult[jss::tx_blob] =
-            strHex(tpTrans->getSTransaction()->getSerializer().peekData());
-
-        if (temUNCERTAIN != tpTrans->getResult())
-        {
-            std::string sToken;
-            std::string sHuman;
-
-            transResultInfo(tpTrans->getResult(), sToken, sHuman);
-
-            jvResult[jss::engine_result] = sToken;
-            jvResult[jss::engine_result_code] = tpTrans->getResult();
-            jvResult[jss::engine_result_message] = sHuman;
-
-            auto const submitResult = tpTrans->getSubmitResult();
-
-            jvResult[jss::accepted] = submitResult.any();
-            jvResult[jss::applied] = submitResult.applied;
-            jvResult[jss::broadcast] = submitResult.broadcast;
-            jvResult[jss::queued] = submitResult.queued;
-            jvResult[jss::kept] = submitResult.kept;
-
-            if (auto currentLedgerState = tpTrans->getCurrentLedgerState())
-            {
-                jvResult[jss::account_sequence_next] =
-                    safe_cast<Json::Value::UInt>(
-                        currentLedgerState->accountSeqNext);
-                jvResult[jss::account_sequence_available] =
-                    safe_cast<Json::Value::UInt>(
-                        currentLedgerState->accountSeqAvail);
-                jvResult[jss::open_ledger_cost] =
-                    to_string(currentLedgerState->minFeeRequired);
-                jvResult[jss::validated_ledger_index] =
-                    safe_cast<Json::Value::UInt>(
-                        currentLedgerState->validatedLedger);
-            }
-        }
-
-        return jvResult;
-    }
-    catch (std::exception& e)
-    {
-        jvResult[jss::error] = "internalJson";
-        jvResult[jss::error_exception] = e.what();
-
-        return jvResult;
-    }
+    // Process single tx_blob
+    auto const failType = getFailHard(context);
+    return processSingleTransaction(
+        context, context.params[jss::tx_blob].asString(), failType);
 }
 
 }  // namespace ripple

src/ripple/rpc/impl/ServerHandlerImp.cpp

@@ -112,7 +112,7 @@ ServerHandlerImp::ServerHandlerImp(
     , m_resourceManager(resourceManager)
     , m_journal(app_.journal("Server"))
     , m_networkOPs(networkOPs)
-    , m_server(make_Server(*this, io_service, app_.journal("Server")))
+    , m_server(make_Server(*this, io_service, app_.journal("Server"), app))
     , m_jobQueue(jobQueue)
 {
     auto const& group(cm.group("rpc"));
@@ -365,8 +365,23 @@ void
 ServerHandlerImp::onUDPMessage(
     std::string const& message,
     boost::asio::ip::tcp::endpoint const& remoteEndpoint,
+    Port const& p,
     std::function<void(std::string const&)> sendResponse)
 {
+    uint8_t static is_peer[65536] = {};
+    auto const port = p.port;
+
+    if (is_peer[port] == 0 /* not yet known */)
+    {
+        is_peer[port] = p.has_peer() ? 1 : 2;
+        std::cout << "set port " << port << " to " << ('0' + is_peer[port])
+                  << "\n";
+    }
+
+    // udp messages arriving on peer protocol ports are sent back to overlay
+    if (is_peer[port] == 1)
+        return app_.overlay().processXUSH(message, remoteEndpoint);
+
     Json::Value jv;
     if (message.size() > RPC::Tuning::maxRequestSize ||
         !Json::Reader{}.parse(message, jv) || !jv.isObject())
@@ -447,9 +462,9 @@ logDuration(
     beast::Journal& journal)
 {
     using namespace std::chrono_literals;
-    auto const level = (duration >= 10s)
-        ? journal.error()
-        : (duration >= 1s) ? journal.warn() : journal.debug();
+    auto const level = (duration >= 10s) ? journal.error()
+        : (duration >= 1s)               ? journal.warn()
+                                         : journal.debug();
 
     JLOG(level) << "RPC request processing duration = "
                 << std::chrono::duration_cast<std::chrono::microseconds>(

src/ripple/rpc/impl/ServerHandlerImp.h

@@ -169,6 +169,7 @@ public:
     onUDPMessage(
         std::string const& message,
         boost::asio::ip::tcp::endpoint const& remoteEndpoint,
+        Port const& port,
         std::function<void(std::string const&)> sendResponse);
 
     void

src/ripple/server/Port.h

@@ -93,6 +93,12 @@ struct Port
         return protocol.count("udp") > 0;
     }
 
+    bool
+    has_peer() const
+    {
+        return protocol.count("peer") > 0;
+    }
+
     // Maximum UDP packet size (default 64KB)
     std::size_t udp_packet_size = 65536;
 };

src/ripple/server/Server.h

@@ -34,9 +34,11 @@ std::unique_ptr<Server>
 make_Server(
     Handler& handler,
     boost::asio::io_service& io_service,
-    beast::Journal journal)
+    beast::Journal journal,
+    Application& app)
 {
-    return std::make_unique<ServerImpl<Handler>>(handler, io_service, journal);
+    return std::make_unique<ServerImpl<Handler>>(
+        handler, io_service, journal, app);
 }
 
 }  // namespace ripple

src/ripple/server/impl/ServerImpl.h

@@ -76,6 +76,9 @@ public:
 template <class Handler>
 class ServerImpl : public Server
 {
+public:
+    Application& app_;
+
 private:
     using clock_type = std::chrono::system_clock;
 
@@ -99,7 +102,8 @@ public:
     ServerImpl(
         Handler& handler,
         boost::asio::io_service& io_service,
-        beast::Journal journal);
+        beast::Journal journal,
+        Application& app);
 
     ~ServerImpl();
 
@@ -139,8 +143,10 @@ template <class Handler>
 ServerImpl<Handler>::ServerImpl(
     Handler& handler,
     boost::asio::io_service& io_service,
-    beast::Journal journal)
-    : handler_(handler)
+    beast::Journal journal,
+    Application& app)
+    : app_(app)
+    , handler_(handler)
     , j_(journal)
     , io_service_(io_service)
     , strand_(io_service_)
@@ -192,6 +198,20 @@ ServerImpl<Handler>::ports(std::vector<Port> const& ports)
                 eps.push_back(sp->get_endpoint());
                 sp->run();
             }
+
+            if (port.has_peer())
+            {
+                if (app_.config().UDP_HIGHWAY_PORT == 0)
+                    app_.config().UDP_HIGHWAY_PORT = port.port;
+
+                // peer ports run dual tcp/udp stack
+                if (auto sp = ios_.emplace<UDPDoor<Handler>>(
+                        handler_, io_service_, ports_.back(), j_))
+                {
+                    eps.push_back(sp->get_endpoint());
+                    sp->run();
+                }
+            }
         }
     }
     return eps;

src/ripple/server/impl/UDPDoor.h

@@ -97,6 +97,25 @@ public:
             return;
         }
 
+        std::cout << "UDP Door created on port " << port.port << "\n";
+
+        // Port reuse means we can support dual udp/tcp on the same port
+        // used for peer upgrades to lightweight udp protocol
+        /*
+        RHNOTE: in boost 1.70 apparently SO_REUSEPORT is included with reuse_address, 
+                there's no actual option without modifying the native socket handle
+                despite the obvious need for one. 
+        */
+        /*
+        socket_.set_option(boost::asio::socket_base::reuse_port(true), ec);
+        if (ec)
+        {
+            JLOG(j_.debug())
+                << "UDP set reuse_port failed: " << ec.message();
+            // Not fatal - some platforms don't support it
+        }
+        */
+
         socket_.bind(udp_endpoint, ec);
         if (ec)
         {
@@ -104,7 +123,7 @@ public:
             return;
         }
 
-        JLOG(j_.info()) << "UDP-RPC listening on " << udp_endpoint;
+        JLOG(j_.info()) << "UDP listening on " << udp_endpoint;
     }
 
     endpoint_type
@@ -133,6 +152,8 @@ private:
     void
     do_receive()
     {
+        std::cout << "UDP Door receive on " << port_.port << "\n";
+
         if (!socket_.is_open())
             return;
 
@@ -169,6 +190,7 @@ private:
         handler_.onUDPMessage(
             std::string(recv_buffer_.data(), bytes_transferred),
             tcp_endpoint,
+            port_,
             [this, tcp_endpoint](std::string const& response) {
                 do_send(response, tcp_endpoint);
             });

src/test/server/Server_test.cpp

@@ -148,6 +148,7 @@ public:
         onUDPMessage(
             std::string const& message,
             boost::asio::ip::tcp::endpoint const& remoteEndpoint,
+            Port const& port,
             std::function<void(std::string const&)> sendResponse)
         {
         }
@@ -300,7 +301,9 @@ public:
         sink.threshold(beast::severities::Severity::kAll);
         beast::Journal journal{sink};
         TestHandler handler;
-        auto s = make_Server(handler, thread.get_io_service(), journal);
+        jtx::Env env{*this};
+        auto s =
+            make_Server(handler, thread.get_io_service(), journal, env.app());
         std::vector<Port> serverPort(1);
         serverPort.back().ip =
             beast::IP::Address::from_string(getEnvLocalhostAddr()),
@@ -361,6 +364,7 @@ public:
             onUDPMessage(
                 std::string const& message,
                 boost::asio::ip::tcp::endpoint const& remoteEndpoint,
+                Port const& port,
                 std::function<void(std::string const&)> sendResponse)
             {
             }
@@ -380,10 +384,12 @@ public:
         SuiteJournal journal("Server_test", *this);
 
         NullHandler h;
+        jtx::Env env{*this};
         for (int i = 0; i < 1000; ++i)
         {
             TestThread thread;
-            auto s = make_Server(h, thread.get_io_service(), journal);
+            auto s =
+                make_Server(h, thread.get_io_service(), journal, env.app());
             std::vector<Port> serverPort(1);
             serverPort.back().ip =
                 beast::IP::Address::from_string(getEnvLocalhostAddr()),