diff options
Diffstat (limited to 'src/daemon/main.c')
| -rw-r--r-- | src/daemon/main.c | 385 |
1 files changed, 215 insertions, 170 deletions
diff --git a/src/daemon/main.c b/src/daemon/main.c index 17fef8449..03ae7e003 100644 --- a/src/daemon/main.c +++ b/src/daemon/main.c @@ -6,6 +6,7 @@ #include "static_threads.h" #include "database/engine/page_test.h" +#include <curl/curl.h> #ifdef OS_WINDOWS #include "win_system-info.h" @@ -27,18 +28,7 @@ bool ieee754_doubles = false; time_t netdata_start_time = 0; struct netdata_static_thread *static_threads; -struct config netdata_config = { - .first_section = NULL, - .last_section = NULL, - .mutex = NETDATA_MUTEX_INITIALIZER, - .index = { - .avl_tree = { - .root = NULL, - .compar = appconfig_section_compare - }, - .rwlock = AVL_LOCK_INITIALIZER - } -}; +struct config netdata_config = APPCONFIG_INITIALIZER; typedef struct service_thread { pid_t tid; @@ -326,6 +316,7 @@ void web_client_cache_destroy(void); void netdata_cleanup_and_exit(int ret, const char *action, const char *action_result, const char *action_data) { netdata_exit = 1; + usec_t shutdown_start_time = now_monotonic_usec(); watcher_shutdown_begin(); nd_log_limits_unlimited(); @@ -361,7 +352,7 @@ void netdata_cleanup_and_exit(int ret, const char *action, const char *action_re watcher_step_complete(WATCHER_STEP_ID_CLOSE_WEBRTC_CONNECTIONS); service_signal_exit(SERVICE_MAINTENANCE | ABILITY_DATA_QUERIES | ABILITY_WEB_REQUESTS | - ABILITY_STREAMING_CONNECTIONS | SERVICE_ACLK | SERVICE_ACLKSYNC); + ABILITY_STREAMING_CONNECTIONS | SERVICE_ACLK); watcher_step_complete(WATCHER_STEP_ID_DISABLE_MAINTENANCE_NEW_QUERIES_NEW_WEB_REQUESTS_NEW_STREAMING_CONNECTIONS_AND_ACLK); service_wait_exit(SERVICE_MAINTENANCE, 3 * USEC_PER_SEC); @@ -474,21 +465,22 @@ void netdata_cleanup_and_exit(int ret, const char *action, const char *action_re #endif } + // Don't register a shutdown event if we crashed + if (!ret) + add_agent_event(EVENT_AGENT_SHUTDOWN_TIME, (int64_t)(now_monotonic_usec() - shutdown_start_time)); sqlite_close_databases(); watcher_step_complete(WATCHER_STEP_ID_CLOSE_SQL_DATABASES); sqlite_library_shutdown(); // unlink the pid - if(pidfile[0]) { + if(pidfile && *pidfile) { if(unlink(pidfile) != 0) netdata_log_error("EXIT: cannot unlink pidfile '%s'.", pidfile); } watcher_step_complete(WATCHER_STEP_ID_REMOVE_PID_FILE); -#ifdef ENABLE_HTTPS netdata_ssl_cleanup(); -#endif watcher_step_complete(WATCHER_STEP_ID_FREE_OPENSSL_STRUCTURES); (void) unlink(agent_incomplete_shutdown_file); @@ -496,6 +488,7 @@ void netdata_cleanup_and_exit(int ret, const char *action, const char *action_re watcher_shutdown_end(); watcher_thread_stop(); + curl_global_cleanup(); #ifdef OS_WINDOWS return; @@ -527,12 +520,12 @@ void web_server_threading_selection(void) { int make_dns_decision(const char *section_name, const char *config_name, const char *default_value, SIMPLE_PATTERN *p) { - char *value = config_get(section_name,config_name,default_value); + const char *value = config_get(section_name,config_name,default_value); if(!strcmp("yes",value)) return 1; if(!strcmp("no",value)) return 0; - if(strcmp("heuristic",value)) + if(strcmp("heuristic",value) != 0) netdata_log_error("Invalid configuration option '%s' for '%s'/'%s'. Valid options are 'yes', 'no' and 'heuristic'. Proceeding with 'heuristic'", value, section_name, config_name); @@ -542,11 +535,13 @@ int make_dns_decision(const char *section_name, const char *config_name, const c void web_server_config_options(void) { web_client_timeout = - (int)config_get_number(CONFIG_SECTION_WEB, "disconnect idle clients after seconds", web_client_timeout); + (int)config_get_duration_seconds(CONFIG_SECTION_WEB, "disconnect idle clients after", web_client_timeout); + web_client_first_request_timeout = - (int)config_get_number(CONFIG_SECTION_WEB, "timeout for first request", web_client_first_request_timeout); + (int)config_get_duration_seconds(CONFIG_SECTION_WEB, "timeout for first request", web_client_first_request_timeout); + web_client_streaming_rate_t = - config_get_number(CONFIG_SECTION_WEB, "accept a streaming request every seconds", web_client_streaming_rate_t); + config_get_duration_seconds(CONFIG_SECTION_WEB, "accept a streaming request every", web_client_streaming_rate_t); respect_web_browser_do_not_track_policy = config_get_boolean(CONFIG_SECTION_WEB, "respect do not track policy", respect_web_browser_do_not_track_policy); @@ -595,7 +590,7 @@ void web_server_config_options(void) web_enable_gzip = config_get_boolean(CONFIG_SECTION_WEB, "enable gzip compression", web_enable_gzip); - char *s = config_get(CONFIG_SECTION_WEB, "gzip compression strategy", "default"); + const char *s = config_get(CONFIG_SECTION_WEB, "gzip compression strategy", "default"); if(!strcmp(s, "default")) web_gzip_strategy = Z_DEFAULT_STRATEGY; else if(!strcmp(s, "filtered")) @@ -807,8 +802,6 @@ int help(int exitcode) { " are enabled or not, in JSON format.\n\n" " -W simple-pattern pattern string\n" " Check if string matches pattern and exit.\n\n" - " -W \"claim -token=TOKEN -rooms=ROOM1,ROOM2\"\n" - " Claim the agent to the workspace rooms pointed to by TOKEN and ROOM*.\n\n" #ifdef OS_WINDOWS " -W perflibdump [key]\n" " Dump the Windows Performance Counters Registry in JSON.\n\n" @@ -825,7 +818,6 @@ int help(int exitcode) { return exitcode; } -#ifdef ENABLE_HTTPS static void security_init(){ char filename[FILENAME_MAX + 1]; snprintfz(filename, FILENAME_MAX, "%s/ssl/key.pem",netdata_configured_user_config_dir); @@ -839,14 +831,13 @@ static void security_init(){ netdata_ssl_initialize_openssl(); } -#endif static void log_init(void) { nd_log_set_facility(config_get(CONFIG_SECTION_LOGS, "facility", "daemon")); time_t period = ND_LOG_DEFAULT_THROTTLE_PERIOD; size_t logs = ND_LOG_DEFAULT_THROTTLE_LOGS; - period = config_get_number(CONFIG_SECTION_LOGS, "logs flood protection period", period); + period = config_get_duration_seconds(CONFIG_SECTION_LOGS, "logs flood protection period", period); logs = (unsigned long)config_get_number(CONFIG_SECTION_LOGS, "logs to trigger flood protection", (long long int)logs); nd_log_set_flood_protection(logs, period); @@ -856,50 +847,75 @@ static void log_init(void) { nd_log_set_priority_level(config_get(CONFIG_SECTION_LOGS, "level", netdata_log_level)); char filename[FILENAME_MAX + 1]; + char* os_default_method = NULL; +#if defined(OS_LINUX) + os_default_method = is_stderr_connected_to_journal() /* || nd_log_journal_socket_available() */ ? "journal" : NULL; +#elif defined(OS_WINDOWS) +#if defined(HAVE_ETW) + os_default_method = "etw"; +#elif defined(HAVE_WEL) + os_default_method = "wel"; +#endif +#endif + +#if defined(OS_WINDOWS) + // on windows, debug log goes to windows events + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); +#else snprintfz(filename, FILENAME_MAX, "%s/debug.log", netdata_configured_log_dir); +#endif + nd_log_set_user_settings(NDLS_DEBUG, config_get(CONFIG_SECTION_LOGS, "debug", filename)); - bool with_journal = is_stderr_connected_to_journal() /* || nd_log_journal_socket_available() */; - if(with_journal) - snprintfz(filename, FILENAME_MAX, "journal"); + if(os_default_method) + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); else snprintfz(filename, FILENAME_MAX, "%s/daemon.log", netdata_configured_log_dir); nd_log_set_user_settings(NDLS_DAEMON, config_get(CONFIG_SECTION_LOGS, "daemon", filename)); - if(with_journal) - snprintfz(filename, FILENAME_MAX, "journal"); + if(os_default_method) + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); else snprintfz(filename, FILENAME_MAX, "%s/collector.log", netdata_configured_log_dir); nd_log_set_user_settings(NDLS_COLLECTORS, config_get(CONFIG_SECTION_LOGS, "collector", filename)); +#if defined(OS_WINDOWS) + // on windows, access log goes to windows events + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); +#else snprintfz(filename, FILENAME_MAX, "%s/access.log", netdata_configured_log_dir); +#endif nd_log_set_user_settings(NDLS_ACCESS, config_get(CONFIG_SECTION_LOGS, "access", filename)); - if(with_journal) - snprintfz(filename, FILENAME_MAX, "journal"); + if(os_default_method) + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); else snprintfz(filename, FILENAME_MAX, "%s/health.log", netdata_configured_log_dir); nd_log_set_user_settings(NDLS_HEALTH, config_get(CONFIG_SECTION_LOGS, "health", filename)); -#ifdef ENABLE_ACLK aclklog_enabled = config_get_boolean(CONFIG_SECTION_CLOUD, "conversation log", CONFIG_BOOLEAN_NO); if (aclklog_enabled) { +#if defined(OS_WINDOWS) + // on windows, aclk log goes to windows events + snprintfz(filename, FILENAME_MAX, "%s", os_default_method); +#else snprintfz(filename, FILENAME_MAX, "%s/aclk.log", netdata_configured_log_dir); +#endif nd_log_set_user_settings(NDLS_ACLK, config_get(CONFIG_SECTION_CLOUD, "conversation log file", filename)); } -#endif + + aclk_config_get_query_scope(); } -char *initialize_lock_directory_path(char *prefix) -{ +static const char *get_varlib_subdir_from_config(const char *prefix, const char *dir) { char filename[FILENAME_MAX + 1]; - snprintfz(filename, FILENAME_MAX, "%s/lock", prefix); - - return config_get(CONFIG_SECTION_DIRECTORIES, "lock", filename); + snprintfz(filename, FILENAME_MAX, "%s/%s", prefix, dir); + return config_get(CONFIG_SECTION_DIRECTORIES, dir, filename); } static void backwards_compatible_config() { // move [global] options to the [web] section + config_move(CONFIG_SECTION_GLOBAL, "http port listen backlog", CONFIG_SECTION_WEB, "listen backlog"); @@ -1003,7 +1019,10 @@ static void backwards_compatible_config() { CONFIG_SECTION_PLUGINS, "statsd"); config_move(CONFIG_SECTION_GLOBAL, "memory mode", - CONFIG_SECTION_DB, "mode"); + CONFIG_SECTION_DB, "db"); + + config_move(CONFIG_SECTION_DB, "mode", + CONFIG_SECTION_DB, "db"); config_move(CONFIG_SECTION_GLOBAL, "history", CONFIG_SECTION_DB, "retention"); @@ -1012,7 +1031,13 @@ static void backwards_compatible_config() { CONFIG_SECTION_DB, "update every"); config_move(CONFIG_SECTION_GLOBAL, "page cache size", - CONFIG_SECTION_DB, "dbengine page cache size MB"); + CONFIG_SECTION_DB, "dbengine page cache size"); + + config_move(CONFIG_SECTION_DB, "dbengine page cache size MB", + CONFIG_SECTION_DB, "dbengine page cache size"); + + config_move(CONFIG_SECTION_DB, "dbengine extent cache size MB", + CONFIG_SECTION_DB, "dbengine extent cache size"); config_move(CONFIG_SECTION_DB, "page cache size", CONFIG_SECTION_DB, "dbengine page cache size MB"); @@ -1023,30 +1048,6 @@ static void backwards_compatible_config() { config_move(CONFIG_SECTION_DB, "page cache with malloc", CONFIG_SECTION_DB, "dbengine page cache with malloc"); - config_move(CONFIG_SECTION_GLOBAL, "dbengine disk space", - CONFIG_SECTION_DB, "dbengine disk space MB"); - - config_move(CONFIG_SECTION_GLOBAL, "dbengine multihost disk space", - CONFIG_SECTION_DB, "dbengine multihost disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine disk space MB", - CONFIG_SECTION_DB, "dbengine multihost disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine multihost disk space MB", - CONFIG_SECTION_DB, "dbengine tier 0 disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine tier 1 multihost disk space MB", - CONFIG_SECTION_DB, "dbengine tier 1 disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine tier 2 multihost disk space MB", - CONFIG_SECTION_DB, "dbengine tier 2 disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine tier 3 multihost disk space MB", - CONFIG_SECTION_DB, "dbengine tier 3 disk space MB"); - - config_move(CONFIG_SECTION_DB, "dbengine tier 4 multihost disk space MB", - CONFIG_SECTION_DB, "dbengine tier 4 disk space MB"); - config_move(CONFIG_SECTION_GLOBAL, "memory deduplication (ksm)", CONFIG_SECTION_DB, "memory deduplication (ksm)"); @@ -1060,17 +1061,67 @@ static void backwards_compatible_config() { CONFIG_SECTION_DB, "dbengine pages per extent"); config_move(CONFIG_SECTION_GLOBAL, "cleanup obsolete charts after seconds", - CONFIG_SECTION_DB, "cleanup obsolete charts after secs"); + CONFIG_SECTION_DB, "cleanup obsolete charts after"); + + config_move(CONFIG_SECTION_DB, "cleanup obsolete charts after secs", + CONFIG_SECTION_DB, "cleanup obsolete charts after"); config_move(CONFIG_SECTION_GLOBAL, "gap when lost iterations above", CONFIG_SECTION_DB, "gap when lost iterations above"); config_move(CONFIG_SECTION_GLOBAL, "cleanup orphan hosts after seconds", - CONFIG_SECTION_DB, "cleanup orphan hosts after secs"); + CONFIG_SECTION_DB, "cleanup orphan hosts after"); + + config_move(CONFIG_SECTION_DB, "cleanup orphan hosts after secs", + CONFIG_SECTION_DB, "cleanup orphan hosts after"); + + config_move(CONFIG_SECTION_DB, "cleanup ephemeral hosts after secs", + CONFIG_SECTION_DB, "cleanup ephemeral hosts after"); + + config_move(CONFIG_SECTION_DB, "seconds to replicate", + CONFIG_SECTION_DB, "replication period"); + + config_move(CONFIG_SECTION_DB, "seconds per replication step", + CONFIG_SECTION_DB, "replication step"); config_move(CONFIG_SECTION_GLOBAL, "enable zero metrics", CONFIG_SECTION_DB, "enable zero metrics"); + // ---------------------------------------------------------------------------------------------------------------- + + config_move(CONFIG_SECTION_GLOBAL, "dbengine disk space", + CONFIG_SECTION_DB, "dbengine tier 0 retention size"); + + config_move(CONFIG_SECTION_GLOBAL, "dbengine multihost disk space", + CONFIG_SECTION_DB, "dbengine tier 0 retention size"); + + config_move(CONFIG_SECTION_DB, "dbengine disk space MB", + CONFIG_SECTION_DB, "dbengine tier 0 retention size"); + + for(size_t tier = 0; tier < RRD_STORAGE_TIERS ;tier++) { + char old_config[128], new_config[128]; + + snprintfz(old_config, sizeof(old_config), "dbengine tier %zu retention days", tier); + snprintfz(new_config, sizeof(new_config), "dbengine tier %zu retention time", tier); + config_move(CONFIG_SECTION_DB, old_config, + CONFIG_SECTION_DB, new_config); + + if(tier == 0) + snprintfz(old_config, sizeof(old_config), "dbengine multihost disk space MB"); + else + snprintfz(old_config, sizeof(old_config), "dbengine tier %zu multihost disk space MB", tier); + snprintfz(new_config, sizeof(new_config), "dbengine tier %zu retention size", tier); + config_move(CONFIG_SECTION_DB, old_config, + CONFIG_SECTION_DB, new_config); + + snprintfz(old_config, sizeof(old_config), "dbengine tier %zu disk space MB", tier); + snprintfz(new_config, sizeof(new_config), "dbengine tier %zu retention size", tier); + config_move(CONFIG_SECTION_DB, old_config, + CONFIG_SECTION_DB, new_config); + } + + // ---------------------------------------------------------------------------------------------------------------- + config_move(CONFIG_SECTION_LOGS, "error", CONFIG_SECTION_LOGS, "daemon"); @@ -1082,11 +1133,42 @@ static void backwards_compatible_config() { config_move(CONFIG_SECTION_LOGS, "errors flood protection period", CONFIG_SECTION_LOGS, "logs flood protection period"); + config_move(CONFIG_SECTION_HEALTH, "is ephemeral", CONFIG_SECTION_GLOBAL, "is ephemeral node"); config_move(CONFIG_SECTION_HEALTH, "has unstable connection", CONFIG_SECTION_GLOBAL, "has unstable connection"); + + config_move(CONFIG_SECTION_HEALTH, "run at least every seconds", + CONFIG_SECTION_HEALTH, "run at least every"); + + config_move(CONFIG_SECTION_HEALTH, "postpone alarms during hibernation for seconds", + CONFIG_SECTION_HEALTH, "postpone alarms during hibernation for"); + + config_move(CONFIG_SECTION_HEALTH, "health log history", + CONFIG_SECTION_HEALTH, "health log retention"); + + config_move(CONFIG_SECTION_REGISTRY, "registry expire idle persons days", + CONFIG_SECTION_REGISTRY, "registry expire idle persons"); + + config_move(CONFIG_SECTION_WEB, "disconnect idle clients after seconds", + CONFIG_SECTION_WEB, "disconnect idle clients after"); + + config_move(CONFIG_SECTION_WEB, "accept a streaming request every seconds", + CONFIG_SECTION_WEB, "accept a streaming request every"); + + config_move(CONFIG_SECTION_STATSD, "set charts as obsolete after secs", + CONFIG_SECTION_STATSD, "set charts as obsolete after"); + + config_move(CONFIG_SECTION_STATSD, "disconnect idle tcp clients after seconds", + CONFIG_SECTION_STATSD, "disconnect idle tcp clients after"); + + config_move("plugin:idlejitter", "loop time in ms", + "plugin:idlejitter", "loop time"); + + config_move("plugin:proc:/sys/class/infiniband", "refresh ports state every seconds", + "plugin:proc:/sys/class/infiniband", "refresh ports state every"); } static int get_hostname(char *buf, size_t buf_size) { @@ -1119,7 +1201,7 @@ static void get_netdata_configured_variables() // get the hostname netdata_configured_host_prefix = config_get(CONFIG_SECTION_GLOBAL, "host access prefix", ""); - verify_netdata_host_prefix(true); + (void) verify_netdata_host_prefix(true); char buf[HOSTNAME_MAX + 1]; if (get_hostname(buf, HOSTNAME_MAX)) @@ -1131,22 +1213,22 @@ static void get_netdata_configured_variables() // ------------------------------------------------------------------------ // get default database update frequency - default_rrd_update_every = (int) config_get_number(CONFIG_SECTION_DB, "update every", UPDATE_EVERY); + default_rrd_update_every = (int) config_get_duration_seconds(CONFIG_SECTION_DB, "update every", UPDATE_EVERY); if(default_rrd_update_every < 1 || default_rrd_update_every > 600) { netdata_log_error("Invalid data collection frequency (update every) %d given. Defaulting to %d.", default_rrd_update_every, UPDATE_EVERY); default_rrd_update_every = UPDATE_EVERY; - config_set_number(CONFIG_SECTION_DB, "update every", default_rrd_update_every); + config_set_duration_seconds(CONFIG_SECTION_DB, "update every", default_rrd_update_every); } // ------------------------------------------------------------------------ - // get default memory mode for the database + // get the database selection { - const char *mode = config_get(CONFIG_SECTION_DB, "mode", rrd_memory_mode_name(default_rrd_memory_mode)); + const char *mode = config_get(CONFIG_SECTION_DB, "db", rrd_memory_mode_name(default_rrd_memory_mode)); default_rrd_memory_mode = rrd_memory_mode_id(mode); if(strcmp(mode, rrd_memory_mode_name(default_rrd_memory_mode)) != 0) { netdata_log_error("Invalid memory mode '%s' given. Using '%s'", mode, rrd_memory_mode_name(default_rrd_memory_mode)); - config_set(CONFIG_SECTION_DB, "mode", rrd_memory_mode_name(default_rrd_memory_mode)); + config_set(CONFIG_SECTION_DB, "db", rrd_memory_mode_name(default_rrd_memory_mode)); } } @@ -1175,7 +1257,8 @@ static void get_netdata_configured_variables() netdata_configured_cache_dir = config_get(CONFIG_SECTION_DIRECTORIES, "cache", netdata_configured_cache_dir); netdata_configured_varlib_dir = config_get(CONFIG_SECTION_DIRECTORIES, "lib", netdata_configured_varlib_dir); - netdata_configured_lock_dir = initialize_lock_directory_path(netdata_configured_varlib_dir); + netdata_configured_lock_dir = get_varlib_subdir_from_config(netdata_configured_varlib_dir, "lock"); + netdata_configured_cloud_dir = get_varlib_subdir_from_config(netdata_configured_varlib_dir, "cloud.d"); { pluginsd_initialize_plugin_directories(); @@ -1199,17 +1282,19 @@ static void get_netdata_configured_variables() // ------------------------------------------------------------------------ // get default Database Engine page cache size in MiB - default_rrdeng_page_cache_mb = (int) config_get_number(CONFIG_SECTION_DB, "dbengine page cache size MB", default_rrdeng_page_cache_mb); - default_rrdeng_extent_cache_mb = (int) config_get_number(CONFIG_SECTION_DB, "dbengine extent cache size MB", default_rrdeng_extent_cache_mb); + default_rrdeng_page_cache_mb = (int) config_get_size_mb(CONFIG_SECTION_DB, "dbengine page cache size", default_rrdeng_page_cache_mb); + default_rrdeng_extent_cache_mb = (int) config_get_size_mb(CONFIG_SECTION_DB, "dbengine extent cache size", default_rrdeng_extent_cache_mb); db_engine_journal_check = config_get_boolean(CONFIG_SECTION_DB, "dbengine enable journal integrity check", CONFIG_BOOLEAN_NO); - if(default_rrdeng_extent_cache_mb < 0) + if(default_rrdeng_extent_cache_mb < 0) { default_rrdeng_extent_cache_mb = 0; + config_set_size_mb(CONFIG_SECTION_DB, "dbengine extent cache size", default_rrdeng_extent_cache_mb); + } if(default_rrdeng_page_cache_mb < RRDENG_MIN_PAGE_CACHE_SIZE_MB) { netdata_log_error("Invalid page cache size %d given. Defaulting to %d.", default_rrdeng_page_cache_mb, RRDENG_MIN_PAGE_CACHE_SIZE_MB); default_rrdeng_page_cache_mb = RRDENG_MIN_PAGE_CACHE_SIZE_MB; - config_set_number(CONFIG_SECTION_DB, "dbengine page cache size MB", default_rrdeng_page_cache_mb); + config_set_size_mb(CONFIG_SECTION_DB, "dbengine page cache size", default_rrdeng_page_cache_mb); } // ------------------------------------------------------------------------ @@ -1242,28 +1327,24 @@ static void get_netdata_configured_variables() // get KSM settings #ifdef MADV_MERGEABLE - enable_ksm = config_get_boolean(CONFIG_SECTION_DB, "memory deduplication (ksm)", enable_ksm); + enable_ksm = config_get_boolean_ondemand(CONFIG_SECTION_DB, "memory deduplication (ksm)", enable_ksm); #endif // -------------------------------------------------------------------- - // metric correlations - enable_metric_correlations = config_get_boolean(CONFIG_SECTION_GLOBAL, "enable metric correlations", enable_metric_correlations); - default_metric_correlations_method = weights_string_to_method(config_get( - CONFIG_SECTION_GLOBAL, "metric correlations method", - weights_method_to_string(default_metric_correlations_method))); + rrdhost_free_ephemeral_time_s = + config_get_duration_seconds(CONFIG_SECTION_DB, "cleanup ephemeral hosts after", rrdhost_free_ephemeral_time_s); - // -------------------------------------------------------------------- + rrdset_free_obsolete_time_s = + config_get_duration_seconds(CONFIG_SECTION_DB, "cleanup obsolete charts after", rrdset_free_obsolete_time_s); - rrdset_free_obsolete_time_s = config_get_number(CONFIG_SECTION_DB, "cleanup obsolete charts after secs", rrdset_free_obsolete_time_s); - rrdhost_free_ephemeral_time_s = config_get_number(CONFIG_SECTION_DB, "cleanup ephemeral hosts after secs", rrdhost_free_ephemeral_time_s); // Current chart locking and invalidation scheme doesn't prevent Netdata from segmentation faults if a short // cleanup delay is set. Extensive stress tests showed that 10 seconds is quite a safe delay. Look at // https://github.com/netdata/netdata/pull/11222#issuecomment-868367920 for more information. if (rrdset_free_obsolete_time_s < 10) { rrdset_free_obsolete_time_s = 10; - netdata_log_info("The \"cleanup obsolete charts after seconds\" option was set to 10 seconds."); - config_set_number(CONFIG_SECTION_DB, "cleanup obsolete charts after secs", rrdset_free_obsolete_time_s); + netdata_log_info("The \"cleanup obsolete charts after\" option was set to 10 seconds."); + config_set_duration_seconds(CONFIG_SECTION_DB, "cleanup obsolete charts after", rrdset_free_obsolete_time_s); } gap_when_lost_iterations_above = (int)config_get_number(CONFIG_SECTION_DB, "gap when lost iterations above", gap_when_lost_iterations_above); @@ -1276,14 +1357,13 @@ static void get_netdata_configured_variables() // -------------------------------------------------------------------- // get various system parameters - os_get_system_HZ(); os_get_system_cpus_uncached(); os_get_system_pid_max(); } -static void post_conf_load(char **user) +static void post_conf_load(const char **user) { // -------------------------------------------------------------------- // get the user we should run @@ -1298,7 +1378,7 @@ static void post_conf_load(char **user) } } -static bool load_netdata_conf(char *filename, char overwrite_used, char **user) { +static bool load_netdata_conf(char *filename, char overwrite_used, const char **user) { errno_clear(); int ret = 0; @@ -1309,14 +1389,14 @@ static bool load_netdata_conf(char *filename, char overwrite_used, char **user) netdata_log_error("CONFIG: cannot load config file '%s'.", filename); } else { - filename = strdupz_path_subpath(netdata_configured_user_config_dir, "netdata.conf"); + filename = filename_from_path_entry_strdupz(netdata_configured_user_config_dir, "netdata.conf"); ret = config_load(filename, overwrite_used, NULL); if(!ret) { netdata_log_info("CONFIG: cannot load user config '%s'. Will try the stock version.", filename); freez(filename); - filename = strdupz_path_subpath(netdata_configured_stock_config_dir, "netdata.conf"); + filename = filename_from_path_entry_strdupz(netdata_configured_stock_config_dir, "netdata.conf"); ret = config_load(filename, overwrite_used, NULL); if(!ret) netdata_log_info("CONFIG: cannot load stock config '%s'. Running with internal defaults.", filename); @@ -1351,7 +1431,7 @@ int get_system_info(struct rrdhost_system_info *system_info) { char line[200 + 1]; // Removed the double strlens, if the Coverity tainted string warning reappears I'll revert. // One time init code, but I'm curious about the warning... - while (fgets(line, 200, instance->child_stdout_fp) != NULL) { + while (fgets(line, 200, spawn_popen_stdout(instance)) != NULL) { char *value=line; while (*value && *value != '=') value++; if (*value=='=') { @@ -1366,7 +1446,7 @@ int get_system_info(struct rrdhost_system_info *system_info) { if(unlikely(rrdhost_set_system_info_variable(system_info, line, value))) { netdata_log_error("Unexpected environment variable %s=%s", line, value); } else { - setenv(line, value, 1); + nd_setenv(line, value, 1); } } } @@ -1405,12 +1485,13 @@ int unittest_rrdpush_compressions(void); int uuid_unittest(void); int progress_unittest(void); int dyncfg_unittest(void); +bool netdata_random_session_id_generate(void); #ifdef OS_WINDOWS int windows_perflib_dump(const char *key); #endif -int unittest_prepare_rrd(char **user) { +int unittest_prepare_rrd(const char **user) { post_conf_load(user); get_netdata_configured_variables(); default_rrd_update_every = 1; @@ -1422,13 +1503,12 @@ int unittest_prepare_rrd(char **user) { fprintf(stderr, "rrd_init failed for unittest\n"); return 1; } - default_rrdpush_enabled = 0; + stream_conf_send_enabled = 0; return 0; } int netdata_main(int argc, char **argv) { - clocks_init(); string_init(); analytics_init(); @@ -1441,7 +1521,7 @@ int netdata_main(int argc, char **argv) { int config_loaded = 0; bool close_open_fds = true; size_t default_stacksize; - char *user = NULL; + const char *user = NULL; #ifdef OS_WINDOWS int dont_fork = 1; @@ -1455,6 +1535,8 @@ int netdata_main(int argc, char **argv) { // set the name for logging program_name = "netdata"; + curl_global_init(CURL_GLOBAL_ALL); + // parse options { int num_opts = sizeof(option_definitions) / sizeof(struct option_def); @@ -1483,7 +1565,7 @@ int netdata_main(int argc, char **argv) { } else { netdata_log_debug(D_OPTIONS, "Configuration loaded from %s.", optarg); - load_cloud_conf(1); + cloud_conf_load(1); config_loaded = 1; } break; @@ -1499,8 +1581,7 @@ int netdata_main(int argc, char **argv) { config_set(CONFIG_SECTION_WEB, "bind to", optarg); break; case 'P': - strncpy(pidfile, optarg, FILENAME_MAX); - pidfile[FILENAME_MAX] = '\0'; + pidfile = strdupz(optarg); break; case 'p': config_set(CONFIG_SECTION_GLOBAL, "default port", optarg); @@ -1522,7 +1603,6 @@ int netdata_main(int argc, char **argv) { { char* stacksize_string = "stacksize="; char* debug_flags_string = "debug_flags="; - char* claim_string = "claim"; #ifdef ENABLE_DBENGINE char* createdataset_string = "createdataset="; char* stresstest_string = "stresstest="; @@ -1791,7 +1871,7 @@ int netdata_main(int argc, char **argv) { // so the caller can use -c netdata.conf before or // after this parameter to prevent or allow overwriting // variables at netdata.conf - config_set_default(section, key, value); + config_set_default_raw_value(section, key, value); // fprintf(stderr, "SET section '%s', key '%s', value '%s'\n", section, key, value); } @@ -1824,7 +1904,7 @@ int netdata_main(int argc, char **argv) { // so the caller can use -c netdata.conf before or // after this parameter to prevent or allow overwriting // variables at netdata.conf - appconfig_set_default(tmp_config, section, key, value); + appconfig_set_default_raw_value(tmp_config, section, key, value); // fprintf(stderr, "SET section '%s', key '%s', value '%s'\n", section, key, value); } @@ -1870,7 +1950,7 @@ int netdata_main(int argc, char **argv) { if(!config_loaded) { fprintf(stderr, "warning: no configuration file has been loaded. Use -c CONFIG_FILE, before -W get. Using default config.\n"); load_netdata_conf(NULL, 0, &user); - load_cloud_conf(1); + cloud_conf_load(1); } get_netdata_configured_variables(); @@ -1884,10 +1964,6 @@ int netdata_main(int argc, char **argv) { printf("%s\n", value); return 0; } - else if(strncmp(optarg, claim_string, strlen(claim_string)) == 0) { - /* will trigger a claiming attempt when the agent is initialized */ - claiming_pending_arguments = optarg + strlen(claim_string); - } else if(strcmp(optarg, "buildinfo") == 0) { print_build_info(); return 0; @@ -1919,18 +1995,18 @@ int netdata_main(int argc, char **argv) { if (close_open_fds == true) { // close all open file descriptors, except the standard ones // the caller may have left open files (lxc-attach has this issue) - os_close_all_non_std_open_fds_except(NULL, 0); + os_close_all_non_std_open_fds_except(NULL, 0, 0); } if(!config_loaded) { load_netdata_conf(NULL, 0, &user); - load_cloud_conf(0); + cloud_conf_load(0); } // ------------------------------------------------------------------------ // initialize netdata { - char *pmax = config_get(CONFIG_SECTION_GLOBAL, "glibc malloc arena max for plugins", "1"); + const char *pmax = config_get(CONFIG_SECTION_GLOBAL, "glibc malloc arena max for plugins", "1"); if(pmax && *pmax) setenv("MALLOC_ARENA_MAX", pmax, 1); @@ -1970,7 +2046,8 @@ int netdata_main(int argc, char **argv) { // prepare configuration environment variables for the plugins get_netdata_configured_variables(); - set_global_environment(); + set_environment_for_plugins_and_scripts(); + analytics_reset(); // work while we are cd into config_dir // to allow the plugins refer to their config @@ -1986,8 +2063,8 @@ int netdata_main(int argc, char **argv) { // -------------------------------------------------------------------- // get the debugging flags from the configuration file - char *flags = config_get(CONFIG_SECTION_LOGS, "debug flags", "0x0000000000000000"); - setenv("NETDATA_DEBUG_FLAGS", flags, 1); + const char *flags = config_get(CONFIG_SECTION_LOGS, "debug flags", "0x0000000000000000"); + nd_setenv("NETDATA_DEBUG_FLAGS", flags, 1); debug_flags = strtoull(flags, NULL, 0); netdata_log_debug(D_OPTIONS, "Debug flags set to '0x%" PRIX64 "'.", debug_flags); @@ -2013,16 +2090,10 @@ int netdata_main(int argc, char **argv) { nd_log_initialize(); netdata_log_info("Netdata agent version '%s' is starting", NETDATA_VERSION); - ieee754_doubles = is_system_ieee754_double(); - if(!ieee754_doubles) - globally_disabled_capabilities |= STREAM_CAP_IEEE754; - - aral_judy_init(); + check_local_streaming_capabilities(); get_system_timezone(); - bearer_tokens_init(); - replication_initialize(); rrd_functions_inflight_init(); @@ -2030,9 +2101,7 @@ int netdata_main(int argc, char **argv) { // -------------------------------------------------------------------- // get the certificate and start security -#ifdef ENABLE_HTTPS security_init(); -#endif // -------------------------------------------------------------------- // This is the safest place to start the SILENCERS structure @@ -2040,12 +2109,6 @@ int netdata_main(int argc, char **argv) { health_set_silencers_filename(); health_initialize_global_silencers(); -// // -------------------------------------------------------------------- -// // Initialize ML configuration -// -// delta_startup_time("initialize ML"); -// ml_init(); - // -------------------------------------------------------------------- // setup process signals @@ -2053,8 +2116,7 @@ int netdata_main(int argc, char **argv) { // this causes the threads to block signals. delta_startup_time("initialize signals"); - signals_block(); - signals_init(); // setup the signals we want to use + nd_initialize_signals(); // setup the signals we want to use // -------------------------------------------------------------------- // check which threads are enabled and initialize them @@ -2086,7 +2148,7 @@ int netdata_main(int argc, char **argv) { st->init_routine(); if(st->env_name) - setenv(st->env_name, st->enabled?"YES":"NO", 1); + nd_setenv(st->env_name, st->enabled?"YES":"NO", 1); if(st->global_variable) *st->global_variable = (st->enabled) ? true : false; @@ -2097,7 +2159,7 @@ int netdata_main(int argc, char **argv) { delta_startup_time("initialize web server"); - web_client_api_v1_init(); + nd_web_api_init(); web_server_threading_selection(); if(web_server_mode != WEB_SERVER_MODE_NONE) { @@ -2165,7 +2227,7 @@ int netdata_main(int argc, char **argv) { netdata_configured_home_dir = config_get(CONFIG_SECTION_DIRECTORIES, "home", pw->pw_dir); } - setenv("HOME", netdata_configured_home_dir, 1); + nd_setenv("HOME", netdata_configured_home_dir, 1); dyncfg_init(true); @@ -2173,11 +2235,12 @@ int netdata_main(int argc, char **argv) { delta_startup_time("initialize threads after fork"); - netdata_threads_init_after_fork((size_t)config_get_number(CONFIG_SECTION_GLOBAL, "pthread stack size", (long)default_stacksize)); + netdata_threads_init_after_fork((size_t)config_get_size_bytes(CONFIG_SECTION_GLOBAL, "pthread stack size", default_stacksize)); // initialize internal registry delta_startup_time("initialize registry"); registry_init(); + cloud_conf_init_after_registry(); netdata_random_session_id_generate(); // ------------------------------------------------------------------------ @@ -2203,7 +2266,7 @@ int netdata_main(int argc, char **argv) { delta_startup_time("initialize RRD structures"); if(rrd_init(netdata_configured_hostname, system_info, false)) { - set_late_global_environment(system_info); + set_late_analytics_variables(system_info); fatal("Cannot initialize localhost instance with name '%s'.", netdata_configured_hostname); } @@ -2219,15 +2282,10 @@ int netdata_main(int argc, char **argv) { if (fd >= 0) close(fd); - // ------------------------------------------------------------------------ // Claim netdata agent to a cloud endpoint delta_startup_time("collect claiming info"); - - if (claiming_pending_arguments) - claim_agent(claiming_pending_arguments, false, NULL); - load_claiming_state(); // ------------------------------------------------------------------------ @@ -2242,11 +2300,13 @@ int netdata_main(int argc, char **argv) { // ------------------------------------------------------------------------ // spawn the threads + bearer_tokens_init(); + delta_startup_time("start the static threads"); web_server_config_options(); - set_late_global_environment(system_info); + set_late_analytics_variables(system_info); for (i = 0; static_threads[i].name != NULL ; i++) { struct netdata_static_thread *st = &static_threads[i]; @@ -2269,7 +2329,13 @@ int netdata_main(int argc, char **argv) { delta_startup_time("ready"); usec_t ready_ut = now_monotonic_usec(); - netdata_log_info("NETDATA STARTUP: completed in %llu ms. Enjoy real-time performance monitoring!", (ready_ut - started_ut) / USEC_PER_MS); + add_agent_event(EVENT_AGENT_START_TIME, (int64_t ) (ready_ut - started_ut)); + usec_t median_start_time = get_agent_event_time_median(EVENT_AGENT_START_TIME); + netdata_log_info( + "NETDATA STARTUP: completed in %llu ms (median start up time is %llu ms). Enjoy real-time performance monitoring!", + (ready_ut - started_ut) / USEC_PER_MS, median_start_time / USEC_PER_MS); + + cleanup_agent_event_log(); netdata_ready = true; analytics_statistic_t start_statistic = { "START", "-", "-" }; @@ -2295,28 +2361,7 @@ int netdata_main(int argc, char **argv) { } } - // ------------------------------------------------------------------------ - // Report ACLK build failure -#ifndef ENABLE_ACLK - netdata_log_error("This agent doesn't have ACLK."); - char filename[FILENAME_MAX + 1]; - snprintfz(filename, FILENAME_MAX, "%s/.aclk_report_sent", netdata_configured_varlib_dir); - if (netdata_anonymous_statistics_enabled > 0 && access(filename, F_OK)) { // -1 -> not initialized - analytics_statistic_t statistic = { "ACLK_DISABLED", "-", "-" }; - analytics_statistic_send(&statistic); - - int fd = open(filename, O_WRONLY | O_CREAT | O_TRUNC, 444); - if (fd == -1) - netdata_log_error("Cannot create file '%s'. Please fix this.", filename); - else - close(fd); - } -#endif - webrtc_initialize(); - - signals_unblock(); - return 10; } @@ -2327,7 +2372,7 @@ int main(int argc, char *argv[]) if (rc != 10) return rc; - signals_handle(); + nd_process_signals(); return 1; } #endif |