Allows ACLK NG and Legacy to coexist (#11225)

41 files changed, 998 insertions, 1225 deletions

diff --git a/.github/workflows/tests.yml b/.github/workflows/tests.yml
index 7e47f12da4..72c789582a 100644
--- a/.github/workflows/tests.yml
+++ b/.github/workflows/tests.yml
@@ -52,7 +52,7 @@ jobs:
       - name: Configure
         run: |
           autoreconf -ivf
-          ./configure --without-aclk-ng
+          ./configure
       # XXX: Work-around for bug with libbson-1.0 in Ubuntu 18.04
       # See: https://bugs.launchpad.net/ubuntu/+source/libmongoc/+bug/1790771
       #      https://jira.mongodb.org/browse/CDRIVER-2818
diff --git a/CMakeLists.txt b/CMakeLists.txt
index ce737436ee..c5d4bc8f52 100644
--- a/CMakeLists.txt
+++ b/CMakeLists.txt
@@ -730,12 +730,22 @@ set(BACKENDS_PLUGIN_FILES
 set(CLAIM_PLUGIN_FILES
         claim/claim.c
         claim/claim.h
-        aclk/legacy/aclk_rrdhost_state.h
-        aclk/legacy/aclk_common.c
-        aclk/legacy/aclk_common.h
         )
 
-set(ACLK_PLUGIN_FILES
+set(ACLK_ALWAYS_BUILD
+        aclk/aclk_rrdhost_state.h
+        aclk/aclk_api.c
+        aclk/aclk_api.h
+        aclk/aclk_proxy.c
+        aclk/aclk_proxy.h
+        )
+
+set(ACLK_COMMON_FILES
+        aclk/aclk_collector_list.c
+        aclk/aclk_collector_list.h
+        )
+
+set(ACLK_LEGACY_FILES
         aclk/legacy/agent_cloud_link.c
         aclk/legacy/agent_cloud_link.h
         aclk/legacy/aclk_query.c
@@ -750,6 +760,40 @@ set(ACLK_PLUGIN_FILES
         aclk/legacy/aclk_stats.h
         aclk/legacy/aclk_rx_msgs.c
         aclk/legacy/aclk_rx_msgs.h
+        aclk/legacy/aclk_common.c
+        aclk/legacy/aclk_common.h
+        )
+
+set(ACLK_NG_FILES
+        aclk/aclk.c
+        aclk/aclk.h
+        aclk/aclk_util.c
+        aclk/aclk_util.h
+        aclk/aclk_stats.c
+        aclk/aclk_stats.h
+        aclk/aclk_query.c
+        aclk/aclk_query.h
+        aclk/aclk_query_queue.c
+        aclk/aclk_query_queue.h
+        aclk/aclk_otp.c
+        aclk/aclk_otp.h
+        aclk/aclk_tx_msgs.c
+        aclk/aclk_tx_msgs.h
+        aclk/aclk_rx_msgs.c
+        aclk/aclk_rx_msgs.h
+        aclk/https_client.c
+        aclk/https_client.h
+        mqtt_websockets/src/mqtt_wss_client.c
+        mqtt_websockets/src/include/mqtt_wss_client.h
+        mqtt_websockets/src/mqtt_wss_log.c
+        mqtt_websockets/src/include/mqtt_wss_log.h
+        mqtt_websockets/src/ws_client.c
+        mqtt_websockets/src/include/ws_client.h
+        mqtt_websockets/c-rbuf/src/ringbuffer.c
+        mqtt_websockets/c-rbuf/include/ringbuffer.h
+        mqtt_websockets/c-rbuf/src/ringbuffer_internal.h
+        mqtt_websockets/MQTT-C/src/mqtt.c
+        mqtt_websockets/MQTT-C/include/mqtt.h
         )
 
 set(SPAWN_PLUGIN_FILES
@@ -759,7 +803,7 @@ set(SPAWN_PLUGIN_FILES
         spawn/spawn.h
         )
 
-set(ACLK_STATIC_LIBS
+set(ACLK_LEGACY_STATIC_LIBS
         ${CMAKE_SOURCE_DIR}/externaldeps/mosquitto/libmosquitto.a
         ${CMAKE_SOURCE_DIR}/externaldeps/libwebsockets/libwebsockets.a
         )
@@ -967,21 +1011,21 @@ ENDIF()
 
 set(NETDATA_COMMON_LIBRARIES ${NETDATA_COMMON_LIBRARIES} m ${CMAKE_THREAD_LIBS_INIT})
 
-set(ACLK_CAN_BUILD 1)
+set(ACLK_LEGACY_CAN_BUILD 1)
 if(NOT EXISTS "${CMAKE_SOURCE_DIR}/externaldeps/mosquitto/libmosquitto.a")
     message(WARNING "Static build of mosquitto not found. Disabling ACLK")
-    set(ACLK_CAN_BUILD 0)
+    set(ACLK_LEGACY_CAN_BUILD 0)
 ENDIF()
 
 if(NOT EXISTS "${CMAKE_SOURCE_DIR}/externaldeps/libwebsockets/libwebsockets.a")
     message(WARNING "Static build of libwebsockets not found. Disabling ACLK")
-    set(ACLK_CAN_BUILD 0)
+    set(ACLK_LEGACY_CAN_BUILD 0)
 ENDIF()
 
-IF(ACLK_CAN_BUILD)
-    message(STATUS "agent-cloud-link: enabled")
-    list(APPEND NETDATA_FILES ${ACLK_PLUGIN_FILES})
-    list(APPEND NETDATA_COMMON_LIBRARIES ${ACLK_STATIC_LIBS})
+IF(ACLK_LEGACY_CAN_BUILD)
+    message(STATUS "agent-cloud-link Legacy: enabled")
+    list(APPEND NETDATA_FILES ${ACLK_LEGACY_FILES})
+    list(APPEND NETDATA_COMMON_LIBRARIES ${ACLK_LEGACY_STATIC_LIBS})
     include_directories(BEFORE ${CMAKE_SOURCE_DIR}/externaldeps/libwebsockets/include)
     IF(LINUX AND CAP_FOUND)
         list(APPEND NETDATA_COMMON_LIBRARIES ${CAP_LIBRARIES})
@@ -989,9 +1033,16 @@ IF(ACLK_CAN_BUILD)
         list(APPEND NETDATA_COMMON_CFLAGS ${CAP_CFLAGS_OTHER})
     ENDIF()
 ELSE()
-    message(STATUS "agent-cloud-link: disabled")
+    message(STATUS "agent-cloud-link Legacy: disabled")
 ENDIF()
 
+list(APPEND NETDATA_FILES ${ACLK_ALWAYS_BUILD})
+list(APPEND NETDATA_FILES ${ACLK_NG_FILES})
+list(APPEND NETDATA_FILES ${ACLK_COMMON_FILES})
+include_directories(BEFORE ${CMAKE_SOURCE_DIR}/mqtt_websockets/MQTT-C/include)
+include_directories(BEFORE ${CMAKE_SOURCE_DIR}/mqtt_websockets/src/include)
+include_directories(BEFORE ${CMAKE_SOURCE_DIR}/mqtt_websockets/c-rbuf/include)
+
 # -----------------------------------------------------------------------------
 # netdata
 
diff --git a/Makefile.am b/Makefile.am
index da23b4b15d..1c98e30b6a 100644
--- a/Makefile.am
+++ b/Makefile.am
@@ -114,7 +114,7 @@ SUBDIRS += \
     spawn \
     $(NULL)
 
-if !ACLK_NG
+if ENABLE_ACLK
 SUBDIRS += \
     aclk/legacy \
     $(NULL)
@@ -540,7 +540,7 @@ PARSER_FILES = \
     $(NULL)
 
 if ACLK_NG
-ACLK_FILES = \
+ACLK_NG_FILES = \
     aclk/aclk.c \
     aclk/aclk.h \
     aclk/aclk_util.c \
@@ -551,8 +551,6 @@ ACLK_FILES = \
     aclk/aclk_query.h \
     aclk/aclk_query_queue.c \
     aclk/aclk_query_queue.h \
-    aclk/aclk_collector_list.c \
-    aclk/aclk_collector_list.h \
     aclk/aclk_otp.c \
     aclk/aclk_otp.h \
     aclk/aclk_tx_msgs.c \
@@ -573,17 +571,25 @@ ACLK_FILES = \
     mqtt_websockets/MQTT-C/src/mqtt.c \
     mqtt_websockets/MQTT-C/include/mqtt.h \
     $(NULL)
-else #ACLK_NG
-ACLK_FILES = \
-    aclk/legacy/aclk_rrdhost_state.h \
-    aclk/legacy/aclk_common.c \
-    aclk/legacy/aclk_common.h \
-    aclk/legacy/aclk_stats.c \
-    aclk/legacy/aclk_stats.h \
-    $(NULL)
+endif #ACLK_NG
 
 if ENABLE_ACLK
-ACLK_FILES += \
+ACLK_COMMON_FILES = \
+    aclk/aclk_collector_list.c \
+    aclk/aclk_collector_list.h \
+    $(NULL)
+endif
+
+ACLK_ALWAYS_BUILD_FILES = \
+    aclk/aclk_rrdhost_state.h \
+    aclk/aclk_api.c \
+    aclk/aclk_api.h \
+    aclk/aclk_proxy.c \
+    aclk/aclk_proxy.h \
+    $(NULL)
+
+if ACLK_LEGACY
+ACLK_LEGACY_FILES = \
     aclk/legacy/agent_cloud_link.c \
     aclk/legacy/agent_cloud_link.h \
     aclk/legacy/aclk_query.c \
@@ -596,9 +602,12 @@ ACLK_FILES += \
     aclk/legacy/aclk_lws_wss_client.h \
     aclk/legacy/aclk_lws_https_client.c \
     aclk/legacy/aclk_lws_https_client.h \
+    aclk/legacy/aclk_common.c \
+    aclk/legacy/aclk_common.h \
+    aclk/legacy/aclk_stats.c \
+    aclk/legacy/aclk_stats.h \
     $(NULL)
-endif #ENABLE_ACLK
-endif #ACLK_NG
+endif #ACLK_LEGACY
 
 SPAWN_PLUGIN_FILES = \
     spawn/spawn.c \
@@ -710,7 +719,10 @@ NETDATA_FILES = \
     $(WEB_PLUGIN_FILES) \
     $(CLAIM_FILES) \
     $(PARSER_FILES) \
-    $(ACLK_FILES) \
+    $(ACLK_ALWAYS_BUILD_FILES) \
+    $(ACLK_COMMON_FILES) \
+    $(ACLK_LEGACY_FILES) \
+    $(ACLK_NG_FILES) \
     $(SPAWN_PLUGIN_FILES) \
     $(NULL)
 
@@ -770,16 +782,14 @@ netdata_LDADD = \
     $(NETDATA_COMMON_LIBS) \
     $(NULL)
 
-if !ACLK_NG
-if ENABLE_ACLK
+if ACLK_LEGACY
 netdata_LDADD += \
     $(abs_top_srcdir)/externaldeps/mosquitto/libmosquitto.a \
     $(OPTIONAL_LIBCAP_LIBS) \
     $(OPTIONAL_LWS_LIBS) \
     $(NETDATA_COMMON_LIBS) \
     $(NULL)
-endif #ENABLE_ACLK
-endif #ACLK_NG
+endif #ACLK_LEGACY
 
 if ENABLE_CXX_LINKER
     netdata_LINK = $(CXXLD) $(CXXFLAGS) $(LDFLAGS) -o $@
diff --git a/aclk/aclk.c b/aclk/aclk.c
index 3e538a5f4c..7e8c1c32e0 100644
--- a/aclk/aclk.c
+++ b/aclk/aclk.c
@@ -13,6 +13,8 @@
 #include "aclk_collector_list.h"
 #include "https_client.h"
 
+#include "aclk_proxy.h"
+
 #ifdef ACLK_LOG_CONVERSATION_DIR
 #include <sys/types.h>
 #include <sys/stat.h>
@@ -21,19 +23,10 @@
 
 #define ACLK_STABLE_TIMEOUT 3 // Minimum delay to mark AGENT as stable
 
-//TODO remove most (as in 99.999999999%) of this crap
-int aclk_connected = 0;
-int aclk_disable_runtime = 0;
-int aclk_disable_single_updates = 0;
-int aclk_kill_link = 0;
-
 int aclk_pubacks_per_conn = 0; // How many PubAcks we got since MQTT conn est.
 
 time_t aclk_block_until = 0;
 
-usec_t aclk_session_us = 0;         // Used by the mqtt layer
-time_t aclk_session_sec = 0;        // Used by the mqtt layer
-
 aclk_env_t *aclk_env = NULL;
 
 mqtt_wss_client mqttwss_client;
@@ -43,22 +36,12 @@ netdata_mutex_t aclk_shared_state_mutex = NETDATA_MUTEX_INITIALIZER;
 #define ACLK_SHARED_STATE_UNLOCK netdata_mutex_unlock(&aclk_shared_state_mutex)
 
 struct aclk_shared_state aclk_shared_state = {
-    .agent_state = AGENT_INITIALIZING,
+    .agent_state = ACLK_HOST_INITIALIZING,
     .last_popcorn_interrupt = 0,
     .mqtt_shutdown_msg_id = -1,
     .mqtt_shutdown_msg_rcvd = 0
 };
 
-void aclk_single_update_disable()
-{
-    aclk_disable_single_updates = 1;
-}
-
-void aclk_single_update_enable()
-{
-    aclk_disable_single_updates = 0;
-}
-
 //ENDTODO
 
 static RSA *aclk_private_key = NULL;
@@ -301,7 +284,7 @@ static int handle_connection(mqtt_wss_client client)
 inline static int aclk_popcorn_check_bump()
 {
     ACLK_SHARED_STATE_LOCK;
-    if (unlikely(aclk_shared_state.agent_state == AGENT_INITIALIZING)) {
+    if (unlikely(aclk_shared_state.agent_state == ACLK_HOST_INITIALIZING)) {
         aclk_shared_state.last_popcorn_interrupt = now_realtime_sec();
         ACLK_SHARED_STATE_UNLOCK;
         return 1;
@@ -340,7 +323,7 @@ static inline void mqtt_connected_actions(mqtt_wss_client client)
     aclk_pubacks_per_conn = 0;
 
     ACLK_SHARED_STATE_LOCK;
-    if (aclk_shared_state.agent_state != AGENT_INITIALIZING) {
+    if (aclk_shared_state.agent_state != ACLK_HOST_INITIALIZING) {
         error("Sending `connect` payload immediately as popcorning was finished already.");
         queue_connect_payloads();
     }
@@ -360,13 +343,13 @@ static int wait_popcorning_finishes(mqtt_wss_client client, struct aclk_query_th
     int need_wait;
     while (!netdata_exit) {
         ACLK_SHARED_STATE_LOCK;
-        if (likely(aclk_shared_state.agent_state != AGENT_INITIALIZING)) {
+        if (likely(aclk_shared_state.agent_state != ACLK_HOST_INITIALIZING)) {
             ACLK_SHARED_STATE_UNLOCK;
             return 0;
         }
         elapsed = now_realtime_sec() - aclk_shared_state.last_popcorn_interrupt;
         if (elapsed >= ACLK_STABLE_TIMEOUT) {
-            aclk_shared_state.agent_state = AGENT_STABLE;
+            aclk_shared_state.agent_state = ACLK_HOST_STABLE;
             ACLK_SHARED_STATE_UNLOCK;
             error("ACLK localhost popocorn finished");
             if (unlikely(!query_threads->thread_list))
@@ -721,10 +704,10 @@ exit:
 // fix this in both old and new ACLK
 extern void health_alarm_entry2json_nolock(BUFFER *wb, ALARM_ENTRY *ae, RRDHOST *host);
 
-void aclk_alarm_reload(void)
+void ng_aclk_alarm_reload(void)
 {
     ACLK_SHARED_STATE_LOCK;
-    if (unlikely(aclk_shared_state.agent_state == AGENT_INITIALIZING)) {
+    if (unlikely(aclk_shared_state.agent_state == ACLK_HOST_INITIALIZING)) {
         ACLK_SHARED_STATE_UNLOCK;
         return;
     }
@@ -733,7 +716,7 @@ void aclk_alarm_reload(void)
     aclk_queue_query(aclk_query_new(METADATA_ALARMS));
 }
 
-int aclk_update_alarm(RRDHOST *host, ALARM_ENTRY *ae)
+int ng_aclk_update_alarm(RRDHOST *host, ALARM_ENTRY *ae)
 {
     BUFFER *local_buffer;
     json_object *msg;
@@ -742,7 +725,7 @@ int aclk_update_alarm(RRDHOST *host, ALARM_ENTRY *ae)
         return 0;
 
     ACLK_SHARED_STATE_LOCK;
-    if (unlikely(aclk_shared_state.agent_state == AGENT_INITIALIZING)) {
+    if (unlikely(aclk_shared_state.agent_state == ACLK_HOST_INITIALIZING)) {
         ACLK_SHARED_STATE_UNLOCK;
         return 0;
     }
@@ -764,7 +747,7 @@ int aclk_update_alarm(RRDHOST *host, ALARM_ENTRY *ae)
     return 0;
 }
 
-int aclk_update_chart(RRDHOST *host, char *chart_name, int create)
+int ng_aclk_update_chart(RRDHOST *host, char *chart_name, int create)
 {
     struct aclk_query *query;
 
@@ -788,7 +771,7 @@ int aclk_update_chart(RRDHOST *host, char *chart_name, int create)
  * Add a new collector to the list
  * If it exists, update the chart count
  */
-void aclk_add_collector(RRDHOST *host, const char *plugin_name, const char *module_name)
+void ng_aclk_add_collector(RRDHOST *host, const char *plugin_name, const char *module_name)
 {
     struct aclk_query *query;
     struct _collector *tmp_collector;
@@ -831,7 +814,7 @@ void aclk_add_collector(RRDHOST *host, const char *plugin_name, const char *modu
  * This function will release the memory used and schedule
  * a cloud update
  */
-void aclk_del_collector(RRDHOST *host, const char *plugin_name, const char *module_name)
+void ng_aclk_del_collector(RRDHOST *host, const char *plugin_name, const char *module_name)
 {
     struct aclk_query *query;
     struct _collector *tmp_collector;
@@ -871,27 +854,3 @@ void aclk_del_collector(RRDHOST *host, const char *plugin_name, const char *modu
     query->data.metadata_alarms.initial_on_connect = 0;
     aclk_queue_query(query);
 }
-
-struct label *add_aclk_host_labels(struct label *label) {
-#ifdef ENABLE_ACLK
-    ACLK_PROXY_TYPE aclk_proxy;
-    char *proxy_str;
-    aclk_get_proxy(&aclk_proxy);
-
-    switch(aclk_proxy) {
-        case PROXY_TYPE_SOCKS5:
-            proxy_str = "SOCKS5";
-            break;