Commit a0632403 authored by jan.koester's avatar jan.koester
Browse files

test

parent b949e037

debian/changelog

+9 −0
Original line number Diff line number Diff line 
mediadb (20260419+28) unstable; urgency=critical



  * Fix cluster startup deadlock: initial_sync_ok_ was only set on empty 

    clusters but never after a successful index fetch, causing all nodes 

    with real data to remain permanently locked in "syncing" state (5/5 

    incomplete). Now properly set after load_index_from_buffer succeeds.



 -- Jan Koester <jan.koester@tuxist.de>  Sun, 19 Apr 2026 16:15:00 +0200



mediadb (20260419+27) unstable; urgency=critical



  * Fix catastrophic cluster data loss ("stores disappeared") caused by nodes

src/backend.cpp

+13 −14
Original line number Diff line number Diff line
@@ -2436,18 +2436,15 @@ void ClusterMediaBackend::start_sync() { 
    // to become fully available even after waitForPeers returns.

    for (int attempt = 1; attempt <= 5; ++attempt) {

        sync_from_cluster();

        {

        if (initial_sync_ok_.load()) {

            std::shared_lock<std::shared_mutex> cguard(cluster_op_mutex_);

            auto sids = local_.store_ids();

            if (!sids.empty()) {

                // Count total albums across all stores

            std::size_t total_albums = 0;

            for (const auto& sid : sids)

                total_albums += local_.list_albums(sid).size();

            std::cerr << "[CLUSTER-SYNC] initial sync OK on attempt " << attempt

                      << ": " << sids.size() << " stores, " << total_albums << " albums\n";

                if (total_albums > 0) break;  // fully synced

            }

            break;

        }

        std::cerr << "[CLUSTER-SYNC] initial sync attempt " << attempt

                  << "/5 incomplete, retrying in 2s\n";
@@ -2999,19 +2996,21 @@ void ClusterMediaBackend::sync_from_cluster() { 
    bool fetch_ok = cluster_.fetch("index", index_data);

    if (!fetch_ok || index_data.empty()) {

        if (!has_index) {

            std::cerr << "[CLUSTER-SYNC] fresh empty cluster detected\n";

            std::cerr << "[CLUSTER-SYNC] no index exists on any peer — fresh/empty cluster\n";

            initial_sync_ok_.store(true);

        } else {

            std::cerr << "[CLUSTER-SYNC] index fetch "

                      << (fetch_ok ? "returned empty data" : "failed (exception in cluster layer)")

                  << "\n";

                      << " (has_index=true, peers report having group)\n";

        }

        return;

    }

    std::cerr << "[CLUSTER-SYNC] index fetched OK, " << index_data.size() << " bytes\n";

    {

        std::unique_lock<std::shared_mutex> cguard(cluster_op_mutex_);

        local_.load_index_from_buffer(index_data);

    }

    initial_sync_ok_.store(true);



    // Snapshot store IDs under shared lock

    std::vector<std::string> sids;