From 112b5b91647c3dea45cc1c9bc364df526c8012f1 Mon Sep 17 00:00:00 2001
From: Daniel Baumann <daniel.baumann@progress-linux.org>
Date: Wed, 26 Jan 2022 19:05:15 +0100
Subject: Merging upstream version 1.33.0.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>
---
 ml/Config.cc              | 14 +++++++-------
 ml/Config.h               |  3 +++
 ml/Host.cc                |  4 ++++
 ml/kmeans/SamplesBuffer.h |  5 +++--
 ml/ml-dummy.c             |  5 ++++-
 ml/ml.cc                  | 14 +++++++++++++-
 ml/ml.h                   |  1 +
 7 files changed, 35 insertions(+), 11 deletions(-)

(limited to 'ml')

diff --git a/ml/Config.cc b/ml/Config.cc
index f48f9b39f..df0a2825e 100644
--- a/ml/Config.cc
+++ b/ml/Config.cc
@@ -47,12 +47,6 @@ void Config::readMLConfig(void) {
     double ADWindowRateThreshold = config_get_float(ConfigSectionML, "window minimum anomaly rate", 0.25);
     double ADDimensionRateThreshold = config_get_float(ConfigSectionML, "anomaly event min dimension rate threshold", 0.05);
 
-    std::string HostsToSkip = config_get(ConfigSectionML, "hosts to skip from training", "!*");
-    std::string ChartsToSkip = config_get(ConfigSectionML, "charts to skip from training",
-            "!system.* !cpu.* !mem.* !disk.* !disk_* "
-            "!ip.* !ipv4.* !ipv6.* !net.* !net_* !netfilter.* "
-            "!services.* !apps.* !groups.* !user.* !ebpf.* !netdata.* *");
-
     std::stringstream SS;
     SS << netdata_configured_cache_dir << "/anomaly-detection.db";
     Cfg.AnomalyDBPath = SS.str();
@@ -123,6 +117,12 @@ void Config::readMLConfig(void) {
     Cfg.ADWindowRateThreshold = ADWindowRateThreshold;
     Cfg.ADDimensionRateThreshold = ADDimensionRateThreshold;
 
-    Cfg.SP_HostsToSkip = simple_pattern_create(HostsToSkip.c_str(), NULL, SIMPLE_PATTERN_EXACT);
+    Cfg.HostsToSkip = config_get(ConfigSectionML, "hosts to skip from training", "!*");
+    Cfg.SP_HostsToSkip = simple_pattern_create(Cfg.HostsToSkip.c_str(), NULL, SIMPLE_PATTERN_EXACT);
+
+    Cfg.ChartsToSkip = config_get(ConfigSectionML, "charts to skip from training",
+            "!system.* !cpu.* !mem.* !disk.* !disk_* "
+            "!ip.* !ipv4.* !ipv6.* !net.* !net_* !netfilter.* "
+            "!services.* !apps.* !groups.* !user.* !ebpf.* !netdata.* *");
     Cfg.SP_ChartsToSkip = simple_pattern_create(ChartsToSkip.c_str(), NULL, SIMPLE_PATTERN_EXACT);
 }
diff --git a/ml/Config.h b/ml/Config.h
index f29bae3a6..58ecce53f 100644
--- a/ml/Config.h
+++ b/ml/Config.h
@@ -30,7 +30,10 @@ public:
     double ADWindowRateThreshold;
     double ADDimensionRateThreshold;
 
+    std::string HostsToSkip;
     SIMPLE_PATTERN *SP_HostsToSkip;
+
+    std::string ChartsToSkip;
     SIMPLE_PATTERN *SP_ChartsToSkip;
 
     std::string AnomalyDBPath;
diff --git a/ml/Host.cc b/ml/Host.cc
index d26ff2ae4..b632710a4 100644
--- a/ml/Host.cc
+++ b/ml/Host.cc
@@ -282,6 +282,9 @@ void RrdHost::getConfigAsJson(nlohmann::json &Json) const {
     Json["idle-window-size"] = Cfg.ADIdleWindowSize;
     Json["window-rate-threshold"] = Cfg.ADWindowRateThreshold;
     Json["dimension-rate-threshold"] = Cfg.ADDimensionRateThreshold;
+
+    Json["hosts-to-skip"] = Cfg.HostsToSkip;
+    Json["charts-to-skip"] = Cfg.ChartsToSkip;
 }
 
 std::pair<Dimension *, Duration<double>>
@@ -441,6 +444,7 @@ void DetectableHost::detect() {
 }
 
 void DetectableHost::getDetectionInfoAsJson(nlohmann::json &Json) const {
+    Json["version"] = 1;
     Json["anomalous-dimensions"] = NumAnomalousDimensions;
     Json["normal-dimensions"] = NumNormalDimensions;
     Json["total-dimensions"] = NumAnomalousDimensions + NumNormalDimensions;
diff --git a/ml/kmeans/SamplesBuffer.h b/ml/kmeans/SamplesBuffer.h
index fccd216d5..5c3cb1a97 100644
--- a/ml/kmeans/SamplesBuffer.h
+++ b/ml/kmeans/SamplesBuffer.h
@@ -20,8 +20,9 @@ public:
     Sample(CalculatedNumber *Buf, size_t N) : CNs(Buf), NumDims(N) {}
 
     void initDSample(DSample &DS) const {
-        for (size_t Idx = 0; Idx != NumDims; Idx++)
-            DS(Idx) = CNs[Idx];
+        for (size_t Idx = 0; Idx != NumDims; Idx++) {
+            DS(Idx) = std::abs(CNs[Idx]);
+        }
     }
 
     void add(const Sample &RHS) const {
diff --git a/ml/ml-dummy.c b/ml/ml-dummy.c
index 795b80c34..222f0b0f6 100644
--- a/ml/ml-dummy.c
+++ b/ml/ml-dummy.c
@@ -10,7 +10,10 @@ void ml_new_host(RRDHOST *RH) { (void) RH; }
 
 void ml_delete_host(RRDHOST *RH) { (void) RH; }
 
-char *ml_get_host_info(RRDHOST *RH) { (void) RH; }
+char *ml_get_host_info(RRDHOST *RH) {
+    (void) RH;
+    return NULL;
+}
 
 void ml_new_dimension(RRDDIM *RD) { (void) RD; }
 
diff --git a/ml/ml.cc b/ml/ml.cc
index 857d23d33..cfda3d28f 100644
--- a/ml/ml.cc
+++ b/ml/ml.cc
@@ -75,7 +75,6 @@ char *ml_get_host_info(RRDHOST *RH) {
     if (RH && RH->ml_host) {
         Host *H = static_cast<Host *>(RH->ml_host);
         H->getConfigAsJson(ConfigJson);
-        H->getDetectionInfoAsJson(ConfigJson);
     } else {
         ConfigJson["enabled"] = false;
     }
@@ -83,6 +82,19 @@ char *ml_get_host_info(RRDHOST *RH) {
     return strdup(ConfigJson.dump(2, '\t').c_str());
 }
 
+char *ml_get_host_runtime_info(RRDHOST *RH) {
+    nlohmann::json ConfigJson;
+
+    if (RH && RH->ml_host) {
+        Host *H = static_cast<Host *>(RH->ml_host);
+        H->getDetectionInfoAsJson(ConfigJson);
+    } else {
+        return nullptr;
+    }
+
+    return strdup(ConfigJson.dump(1, '\t').c_str());
+}
+
 bool ml_is_anomalous(RRDDIM *RD, double Value, bool Exists) {
     Dimension *D = static_cast<Dimension *>(RD->state->ml_dimension);
     if (!D)
diff --git a/ml/ml.h b/ml/ml.h
index 96448453c..783bfabb9 100644
--- a/ml/ml.h
+++ b/ml/ml.h
@@ -18,6 +18,7 @@ void ml_new_host(RRDHOST *RH);
 void ml_delete_host(RRDHOST *RH);
 
 char *ml_get_host_info(RRDHOST *RH);
+char *ml_get_host_runtime_info(RRDHOST *RH);
 
 void ml_new_dimension(RRDDIM *RD);
 void ml_delete_dimension(RRDDIM *RD);
-- 
cgit v1.2.3