Move Check Stats into CIB.

Fixes #5617
This commit is contained in:
Michael Friedrich 2014-02-10 14:56:39 +01:00
parent babfc128ba
commit 95f7de123e
3 changed files with 182 additions and 109 deletions

View File

@ -18,6 +18,11 @@
******************************************************************************/
#include "icinga/cib.h"
#include "icinga/service.h"
#include "base/objectlock.h"
#include "base/utility.h"
#include "base/dynamictype.h"
#include <boost/foreach.hpp>
using namespace icinga;
@ -43,3 +48,121 @@ int CIB::GetPassiveChecksStatistics(long timespan)
{
return m_PassiveChecksStatistics.GetValues(timespan);
}
ServiceCheckStatistics CIB::CalculateServiceCheckStats(void)
{
double min_latency = -1, max_latency = 0, sum_latency = 0;
int count_latency = 0;
double min_execution_time = -1, max_execution_time = 0, sum_execution_time = 0;
int count_execution_time = 0;
BOOST_FOREACH(const Service::Ptr& service, DynamicType::GetObjects<Service>()) {
ObjectLock olock(service);
CheckResult::Ptr cr = service->GetLastCheckResult();
/* latency */
double latency = Service::CalculateLatency(cr);
if (min_latency == -1 || latency < min_latency)
min_latency = latency;
if (latency > max_latency)
max_latency = latency;
sum_latency += latency;
count_latency++;
/* execution_time */
double execution_time = Service::CalculateExecutionTime(cr);
if (min_execution_time == -1 || execution_time < min_execution_time)
min_execution_time = execution_time;
if (execution_time > max_execution_time)
max_execution_time = execution_time;
sum_execution_time += execution_time;
count_execution_time++;
}
ServiceCheckStatistics scs = {0};
scs.min_latency = min_latency;
scs.max_latency = max_latency;
scs.avg_latency = sum_latency / count_latency;
scs.min_execution_time = min_execution_time;
scs.max_execution_time = max_execution_time;
scs.avg_execution_time = sum_execution_time / count_execution_time;
return scs;
}
ServiceStatistics CIB::CalculateServiceStats(void)
{
ServiceStatistics ss = {0};
BOOST_FOREACH(const Service::Ptr& service, DynamicType::GetObjects<Service>()) {
ObjectLock olock(service);
CheckResult::Ptr cr = service->GetLastCheckResult();
if (service->GetState() == StateOK)
ss.services_ok++;
if (service->GetState() == StateWarning)
ss.services_warning++;
if (service->GetState() == StateCritical)
ss.services_critical++;
if (service->GetState() == StateUnknown)
ss.services_unknown++;
if (!cr)
ss.services_pending++;
if (!service->IsReachable())
ss.services_unreachable++;
if (service->IsFlapping())
ss.services_flapping++;
if (service->IsInDowntime())
ss.services_in_downtime++;
if (service->IsAcknowledged())
ss.services_acknowledged++;
}
return ss;
}
HostStatistics CIB::CalculateHostStats(void)
{
HostStatistics hs = {0};
BOOST_FOREACH(const Host::Ptr& host, DynamicType::GetObjects<Host>()) {
ObjectLock olock(host);
if (host->GetState() == HostUp)
hs.hosts_up++;
if (host->GetState() == HostDown)
hs.hosts_down++;
if (host->GetState() == HostUnreachable)
hs.hosts_unreachable++;
Service::Ptr hc = host->GetCheckService();
if (!hc) {
hs.hosts_pending++;
continue; /* skip host service check counting */
}
if (!hc->GetLastCheckResult())
hs.hosts_pending++;
if (hc->IsFlapping())
hs.hosts_flapping++;
if (hc->IsInDowntime())
hs.hosts_in_downtime++;
if (hc->IsAcknowledged())
hs.hosts_acknowledged++;
}
return hs;
}

View File

@ -26,6 +26,37 @@
namespace icinga
{
typedef struct {
double min_latency;
double max_latency;
double avg_latency;
double min_execution_time;
double max_execution_time;
double avg_execution_time;
} ServiceCheckStatistics;
typedef struct {
double services_ok;
double services_warning;
double services_critical;
double services_unknown;
double services_pending;
double services_unreachable;
double services_flapping;
double services_in_downtime;
double services_acknowledged;
} ServiceStatistics;
typedef struct {
double hosts_up;
double hosts_down;
double hosts_unreachable;
double hosts_pending;
double hosts_flapping;
double hosts_in_downtime;
double hosts_acknowledged;
} HostStatistics;
/**
* Common Information Base class. Holds some statistics (and will likely be
* removed/refactored).
@ -41,6 +72,10 @@ public:
static void UpdatePassiveChecksStatistics(long tv, int num);
static int GetPassiveChecksStatistics(long timespan);
static ServiceCheckStatistics CalculateServiceCheckStats(void);
static ServiceStatistics CalculateServiceStats(void);
static HostStatistics CalculateHostStats(void);
private:
CIB(void);

View File

@ -42,120 +42,35 @@ CheckResult::Ptr IcingaCheckTask::ScriptFunc(const Service::Ptr&)
perfdata->Set("active_checks", CIB::GetActiveChecksStatistics(interval) / interval);
perfdata->Set("passive_checks", CIB::GetPassiveChecksStatistics(interval) / interval);
double min_latency = -1, max_latency = 0, sum_latency = 0;
int count_latency = 0;
double min_execution_time = -1, max_execution_time = 0, sum_execution_time = 0;
int count_execution_time = 0;
ServiceCheckStatistics scs = CIB::CalculateServiceCheckStats();
double services_ok, services_warn, services_crit, services_unknown, services_pending, services_unreachable,
services_flapping, services_in_downtime, services_acknowledged = 0;
double hosts_up, hosts_down, hosts_unreachable, hosts_pending, hosts_flapping, hosts_in_downtime, hosts_acknowledged = 0;
perfdata->Set("min_latency", scs.min_latency);
perfdata->Set("max_latency", scs.max_latency);
perfdata->Set("avg_latency", scs.avg_latency);
perfdata->Set("min_execution_time", scs.min_latency);
perfdata->Set("max_execution_time", scs.max_latency);
perfdata->Set("avg_execution_time", scs.avg_execution_time);
BOOST_FOREACH(const Service::Ptr& service, DynamicType::GetObjects<Service>()) {
ObjectLock olock(service);
ServiceStatistics ss = CIB::CalculateServiceStats();
CheckResult::Ptr cr = service->GetLastCheckResult();
perfdata->Set("num_services_ok", ss.services_ok);
perfdata->Set("num_services_warning", ss.services_warning);
perfdata->Set("num_services_critical", ss.services_critical);
perfdata->Set("num_services_unknown", ss.services_unknown);
perfdata->Set("num_services_pending", ss.services_pending);
perfdata->Set("num_services_unreachable", ss.services_unreachable);
perfdata->Set("num_services_flapping", ss.services_flapping);
perfdata->Set("num_services_in_downtime", ss.services_in_downtime);
perfdata->Set("num_services_acknowledged", ss.services_acknowledged);
/* latency */
double latency = Service::CalculateLatency(cr);
HostStatistics hs = CIB::CalculateHostStats();
if (min_latency == -1 || latency < min_latency)
min_latency = latency;
if (latency > max_latency)
max_latency = latency;
sum_latency += latency;
count_latency++;
/* execution_time */
double execution_time = Service::CalculateExecutionTime(cr);
if (min_execution_time == -1 || execution_time < min_execution_time)
min_execution_time = execution_time;
if (execution_time > max_execution_time)
max_execution_time = execution_time;
sum_execution_time += execution_time;
count_execution_time++;
/* states */
if (service->GetState() == StateOK)
services_ok++;
if (service->GetState() == StateWarning)
services_warn++;
if (service->GetState() == StateCritical)
services_crit++;
if (service->GetState() == StateUnknown)
services_unknown++;
/* pending, unreachable */
if (!cr)
services_pending++;
if (!service->IsReachable())
services_unreachable++;
/* flapping, downtime, acknowledgements */
if (service->IsFlapping())
services_flapping++;
if (service->IsInDowntime())
services_in_downtime++;
if (service->IsAcknowledged())
services_acknowledged++;
}
BOOST_FOREACH(const Host::Ptr& host, DynamicType::GetObjects<Host>()) {
ObjectLock olock(host);
if (host->GetState() == HostUp)
hosts_up++;
if (host->GetState() == HostDown)
hosts_down++;
if (host->GetState() == HostUnreachable)
hosts_unreachable++;
Service::Ptr hc = host->GetCheckService();
if (!hc) {
hosts_pending++;
continue; /* skip host service check counting */
}
if (!hc->GetLastCheckResult())
hosts_pending++;
if (hc->IsFlapping())
hosts_flapping++;
if (hc->IsInDowntime())
hosts_in_downtime++;
if (hc->IsAcknowledged())
hosts_acknowledged++;
}
perfdata->Set("min_latency", min_latency);
perfdata->Set("max_latency", max_latency);
perfdata->Set("avg_latency", sum_latency / count_latency);
perfdata->Set("min_execution_time", min_latency);
perfdata->Set("max_execution_time", max_latency);
perfdata->Set("avg_execution_time", sum_execution_time / count_execution_time);
perfdata->Set("num_services_ok", services_ok);
perfdata->Set("num_services_warn", services_warn);
perfdata->Set("num_services_crit", services_crit);
perfdata->Set("num_services_unknown", services_unknown);
perfdata->Set("num_services_pending", services_pending);
perfdata->Set("num_services_unreachable", services_unreachable);
perfdata->Set("num_services_flapping", services_flapping);
perfdata->Set("num_services_in_downtime", services_in_downtime);
perfdata->Set("num_services_acknowledged", services_acknowledged);
perfdata->Set("num_hosts_up", hosts_up);
perfdata->Set("num_hosts_down", hosts_down);
perfdata->Set("num_hosts_unreachable", hosts_unreachable);
perfdata->Set("num_hosts_flapping", hosts_flapping);
perfdata->Set("num_hosts_in_downtime", hosts_in_downtime);
perfdata->Set("num_hosts_acknowledged", hosts_acknowledged);
perfdata->Set("num_hosts_up", hs.hosts_up);
perfdata->Set("num_hosts_down", hs.hosts_down);
perfdata->Set("num_hosts_unreachable", hs.hosts_unreachable);
perfdata->Set("num_hosts_flapping", hs.hosts_flapping);
perfdata->Set("num_hosts_in_downtime", hs.hosts_in_downtime);
perfdata->Set("num_hosts_acknowledged", hs.hosts_acknowledged);
CheckResult::Ptr cr = make_shared<CheckResult>();
cr->SetOutput("Icinga 2 is running.");