statistics daemon argument changes, documentation, cmake

CHANGELOG.md

@@ -10,18 +10,20 @@ to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 
 ### New
 
-- Added Stats ([!132](https://storage.bsc.es/gitlab/hpc/gekkofs/-/merge_requests/132)) gathering in servers
-  - Stats output can be enabled with --output-stats <filename>
-  - --enable-collection collects normal stats
-  - --enable-chunkstats collects extended chunk stats
+- Added statistics gathering on daemons ([!132](https://storage.bsc.es/gitlab/hpc/gekkofs/-/merge_requests/132)).
+    - Stats output can be enabled with:
+    - `--enable-collection` collects normal statistics.
+    - `--enable-chunkstats` collects extended chunk statistics.
+- Statistics output to file is controlled by `--output-stats <filename>`
+- Added Prometheus support for outputting
+  statistics ([!132](https://storage.bsc.es/gitlab/hpc/gekkofs/-/merge_requests/132)):
+    - Prometheus dependency optional and enabled at compile time with the CMake argument `GKFS_ENABLE_PROMETHEUS`.
+    - `--enable-prometheus` enables statistics pushing to Prometheus if statistics are enabled.
+    - `--prometheus-gateway` sets an IP and port for the Prometheus connection.
 - Added new experimental metadata backend:
   Parallax ([!110](https://storage.bsc.es/gitlab/hpc/gekkofs/-/merge_requests/110)).
     - Added support to use multiple metadata backends.
     - Added `--clean-rootdir-finish` argument to remove rootdir/metadir at the end when the daemon finishes.
-- Added Prometheus Output ([!132](https://storage.bsc.es/gitlab/hpc/gekkofs/-/merge_requests/132))
-  - New option to define gateway --prometheus-gateway <gateway:port>
-  - Prometheus output is optional with "GKFS_ENABLE_PROMETHEUS"
-  - --enable-prometheus creates a thread to push the metrics.
 
 ### Changed
 

src/common/CMakeLists.txt

@@ -52,18 +52,20 @@ target_sources(statistics
 if(GKFS_ENABLE_PROMETHEUS)
     find_package(CURL REQUIRED)
     find_package(prometheus-cpp REQUIRED)
-    set(PROMETHEUS_LIB
-    prometheus-cpp-pull
-    prometheus-cpp-push
-    prometheus-cpp-core
-    curl)
+    set(PROMETHEUS_LINK_LIBRARIES
+        prometheus-cpp::pull
+        prometheus-cpp::push
+        prometheus-cpp::core
+        curl)
+    target_include_directories(statistics PRIVATE ${prometheus-cpp_INCLUDE_DIR})
 endif()
 
   target_link_libraries(statistics
-  PRIVATE
-  ${PROMETHEUS_LIB}
+      PRIVATE
+      ${PROMETHEUS_LINK_LIBRARIES}
   )
 
+
 if(GKFS_ENABLE_CODE_COVERAGE)
   target_code_coverage(distributor AUTO)
   target_code_coverage(statistics AUTO)

src/common/statistics/stats.cpp

@@ -82,7 +82,9 @@ Stats::setup_Prometheus(const std::string& gateway_ip,
 
 Stats::Stats(bool enable_chunkstats, bool enable_prometheus,
              const std::string& stats_file,
-             const std::string& prometheus_gateway) {
+             const std::string& prometheus_gateway)
+    : enable_prometheus_(enable_prometheus),
+      enable_chunkstats_(enable_chunkstats) {
 
     // Init clocks
     start = std::chrono::steady_clock::now();
@@ -105,8 +107,6 @@ Stats::Stats(bool enable_chunkstats, bool enable_prometheus,
     setup_Prometheus(prometheus_gateway.substr(0, pos_separator),
                      prometheus_gateway.substr(pos_separator + 1));
 #endif
-    enable_chunkstats_ = enable_chunkstats;
-    enable_prometheus_ = enable_prometheus;
 
     if(!stats_file.empty() || enable_prometheus_) {
         output_thread_ = true;
@@ -331,7 +331,7 @@ Stats::output(std::chrono::seconds d, std::string file_output) {
 
         times++;
 
-        if(enable_chunkstats_ and of) {
+        if(enable_chunkstats_ && of) {
             if(times % 4 == 0)
                 output_map(of.value());
         }
@@ -340,7 +340,7 @@ Stats::output(std::chrono::seconds d, std::string file_output) {
             gateway->Push();
         }
 #endif
-        while(running and a < d) {
+        while(running && a < d) {
             a += 1s;
             std::this_thread::sleep_for(1s);
         }

src/daemon/daemon.cpp

@@ -654,40 +654,60 @@ parse_input(const cli_options& opts, const CLI::App& desc) {
         GKFS_DATA->parallax_size_md(stoi(opts.parallax_size));
     }
 
-    if(desc.count("--output-stats")) {
-        auto stats_file = opts.stats_file;
-        GKFS_DATA->stats_file(stats_file);
-        GKFS_DATA->spdlogger()->debug("{}() Stats Enabled: '{}'", __func__,
-                                      stats_file);
-    } else {
-        GKFS_DATA->stats_file("");
-        GKFS_DATA->spdlogger()->debug("{}() Stats Output Disabled", __func__);
-    }
-
+    /*
+     * Statistics collection arguments
+     */
     if(desc.count("--enable-collection")) {
         GKFS_DATA->enable_stats(true);
-        GKFS_DATA->spdlogger()->debug("{}() Collection Enabled", __func__);
+        GKFS_DATA->spdlogger()->info("{}() Statistic collection enabled",
+                                     __func__);
     }
-
     if(desc.count("--enable-chunkstats")) {
         GKFS_DATA->enable_chunkstats(true);
-        GKFS_DATA->spdlogger()->debug("{}() ChunkStats Enabled", __func__);
+        GKFS_DATA->spdlogger()->info("{}() Chunk statistic collection enabled",
+                                     __func__);
     }
 
 #ifdef GKFS_ENABLE_PROMETHEUS
     if(desc.count("--enable-prometheus")) {
         GKFS_DATA->enable_prometheus(true);
-        GKFS_DATA->spdlogger()->debug("{}() Prometheus Enabled", __func__);
+        if(GKFS_DATA->enable_stats() || GKFS_DATA->enable_chunkstats())
+            GKFS_DATA->spdlogger()->info(
+                    "{}() Statistics output to Prometheus enabled", __func__);
+        else
+            GKFS_DATA->spdlogger()->warn(
+                    "{}() Prometheus statistic output enabled but no stat collection is enabled. There will be no output to Prometheus",
+                    __func__);
     }
 
-
     if(desc.count("--prometheus-gateway")) {
         auto gateway = opts.prometheus_gateway;
         GKFS_DATA->prometheus_gateway(gateway);
-        GKFS_DATA->spdlogger()->debug("{}() Prometheus Gateway: '{}'", __func__,
-                                      gateway);
+        if(GKFS_DATA->enable_prometheus())
+            GKFS_DATA->spdlogger()->info("{}() Prometheus gateway set to '{}'",
+                                         __func__, gateway);
+        else
+            GKFS_DATA->spdlogger()->warn(
+                    "{}() Prometheus gateway was set but Prometheus is disabled.");
     }
 #endif
+
+    if(desc.count("--output-stats")) {
+        auto stats_file = opts.stats_file;
+        GKFS_DATA->stats_file(stats_file);
+        if(GKFS_DATA->enable_stats() || GKFS_DATA->enable_chunkstats())
+            GKFS_DATA->spdlogger()->info(
+                    "{}() Statistics are written to file '{}'", __func__,
+                    stats_file);
+        else
+            GKFS_DATA->spdlogger()->warn(
+                    "{}() --output-stats argument used but no stat collection is enabled. There will be no output to file '{}'",
+                    __func__, stats_file);
+    } else {
+        GKFS_DATA->stats_file("");
+        GKFS_DATA->spdlogger()->debug("{}() Statistics output disabled",
+                                      __func__);
+    }
 }
 
 /**
@@ -755,24 +775,25 @@ main(int argc, const char* argv[]) {
     desc.add_option("--parallaxsize", opts.parallax_size,
                     "parallaxdb - metadata file size in GB (default 8GB), "
                     "used only with new files");
-    desc.add_option(
-                "--output-stats", opts.stats_file,
-                "Creates a thread that outputs the server stats each 10s, to the file specified");
     desc.add_flag(
                 "--enable-collection",
-                "Enables collection of normal stats, independent of the output-stats option");
+                "Enables collection of general statistics. "
+                "Output requires either the --output-stats or --enable-prometheus argument.");
     desc.add_flag(
                 "--enable-chunkstats",
-                "Enables collection of chunkstats stats, independent of the output-stats option")
-                ;
+                "Enables collection of data chunk statistics in I/O operations."
+                "Output requires either the --output-stats or --enable-prometheus argument.");
+    desc.add_option(
+                "--output-stats", opts.stats_file,
+                "Creates a thread that outputs the server stats each 10s to the specified file.");
     #ifdef GKFS_ENABLE_PROMETHEUS
     desc.add_flag(
                 "--enable-prometheus",
-                "Enables prometheus output, enables thread");
+                "Enables prometheus output and a corresponding thread.");
 
     desc.add_option(
                 "--prometheus-gateway", opts.prometheus_gateway,
-                "Defines the prometheus gateway, default is 127.0.0.1:9091");
+                "Defines the prometheus gateway <ip:port> (Default 127.0.0.1:9091).");
     #endif
 
     desc.add_flag("--version", "Print version and exit.");

src/daemon/handler/srv_data.cpp

@@ -114,10 +114,7 @@ rpc_srv_write(hg_handle_t handle) {
             "{}() path: '{}' chunk_start '{}' chunk_end '{}' chunk_n '{}' total_chunk_size '{}' bulk_size: '{}' offset: '{}'",
             __func__, in.path, in.chunk_start, in.chunk_end, in.chunk_n,
             in.total_chunk_size, bulk_size, in.offset);
-    if(GKFS_DATA->enable_stats()) {
-        GKFS_DATA->stats()->add_value_size(
-                gkfs::utils::Stats::SizeOp::write_size, bulk_size);
-    }
+
 
 #ifdef GKFS_ENABLE_AGIOS
     int* data;
@@ -352,7 +349,13 @@ rpc_srv_write(hg_handle_t handle) {
      */
     GKFS_DATA->spdlogger()->debug("{}() Sending output response {}", __func__,
                                   out.err);
-    return gkfs::rpc::cleanup_respond(&handle, &in, &out, &bulk_handle);
+    auto handler_ret =
+            gkfs::rpc::cleanup_respond(&handle, &in, &out, &bulk_handle);
+    if(GKFS_DATA->enable_stats()) {
+        GKFS_DATA->stats()->add_value_size(
+                gkfs::utils::Stats::SizeOp::write_size, bulk_size);
+    }
+    return handler_ret;
 }
 
 /**
@@ -414,10 +417,6 @@ rpc_srv_read(hg_handle_t handle) {
             "{}() path: '{}' chunk_start '{}' chunk_end '{}' chunk_n '{}' total_chunk_size '{}' bulk_size: '{}' offset: '{}'",
             __func__, in.path, in.chunk_start, in.chunk_end, in.chunk_n,
             in.total_chunk_size, bulk_size, in.offset);
-    if(GKFS_DATA->enable_stats()) {
-        GKFS_DATA->stats()->add_value_size(
-                gkfs::utils::Stats::SizeOp::read_size, bulk_size);
-    }
 
 #ifdef GKFS_ENABLE_AGIOS
     int* data;
@@ -619,7 +618,13 @@ rpc_srv_read(hg_handle_t handle) {
      */
     GKFS_DATA->spdlogger()->debug("{}() Sending output response, err: {}",
                                   __func__, out.err);
-    return gkfs::rpc::cleanup_respond(&handle, &in, &out, &bulk_handle);
+    auto handler_ret =
+            gkfs::rpc::cleanup_respond(&handle, &in, &out, &bulk_handle);
+    if(GKFS_DATA->enable_stats()) {
+        GKFS_DATA->stats()->add_value_size(
+                gkfs::utils::Stats::SizeOp::read_size, bulk_size);
+    }
+    return handler_ret;
 }