icinga2/lib/icinga/checkable-notification.cpp

335 lines
11 KiB
C++

/* Icinga 2 | (c) 2012 Icinga GmbH | GPLv2+ */
#include "icinga/checkable.hpp"
#include "icinga/host.hpp"
#include "icinga/icingaapplication.hpp"
#include "icinga/service.hpp"
#include "base/dictionary.hpp"
#include "base/objectlock.hpp"
#include "base/logger.hpp"
#include "base/exception.hpp"
#include "base/context.hpp"
#include "base/convert.hpp"
#include "base/lazy-init.hpp"
#include "remote/apilistener.hpp"
using namespace icinga;
boost::signals2::signal<void (const Notification::Ptr&, const Checkable::Ptr&, const std::set<User::Ptr>&,
const NotificationType&, const CheckResult::Ptr&, const String&, const String&,
const MessageOrigin::Ptr&)> Checkable::OnNotificationSentToAllUsers;
boost::signals2::signal<void (const Notification::Ptr&, const Checkable::Ptr&, const User::Ptr&,
const NotificationType&, const CheckResult::Ptr&, const String&, const String&, const String&,
const MessageOrigin::Ptr&)> Checkable::OnNotificationSentToUser;
void Checkable::ResetNotificationNumbers()
{
for (const Notification::Ptr& notification : GetNotifications()) {
ObjectLock olock(notification);
notification->ResetNotificationNumber();
}
}
void Checkable::SendNotifications(NotificationType type, const CheckResult::Ptr& cr, const String& author, const String& text)
{
String checkableName = GetName();
CONTEXT("Sending notifications for object '" << checkableName << "'");
bool force = GetForceNextNotification();
SetForceNextNotification(false);
if (!IcingaApplication::GetInstance()->GetEnableNotifications() || !GetEnableNotifications()) {
if (!force) {
Log(LogInformation, "Checkable")
<< "Notifications are disabled for checkable '" << checkableName << "'.";
return;
}
}
std::set<Notification::Ptr> notifications = GetNotifications();
String notificationTypeName = Notification::NotificationTypeToString(type);
// Bail early if there are no notifications.
if (notifications.empty()) {
Log(LogNotice, "Checkable")
<< "Skipping checkable '" << checkableName << "' which doesn't have any notification objects configured.";
return;
}
Log(LogInformation, "Checkable")
<< "Checkable '" << checkableName << "' has " << notifications.size()
<< " notification(s). Checking filters for type '" << notificationTypeName << "', sends will be logged.";
for (const Notification::Ptr& notification : notifications) {
// Re-send stashed notifications from cold startup.
if (ApiListener::UpdatedObjectAuthority()) {
try {
if (!notification->IsPaused()) {
auto stashedNotifications (notification->GetStashedNotifications());
if (stashedNotifications->GetLength()) {
Log(LogNotice, "Notification")
<< "Notification '" << notification->GetName() << "': there are some stashed notifications. Stashing notification to preserve order.";
stashedNotifications->Add(new Dictionary({
{"notification_type", type},
{"cr", cr},
{"force", force},
{"reminder", false},
{"author", author},
{"text", text}
}));
} else {
notification->BeginExecuteNotification(type, cr, force, false, author, text);
}
} else {
Log(LogNotice, "Notification")
<< "Notification '" << notification->GetName() << "': HA cluster active, this endpoint does not have the authority (paused=true). Skipping.";
}
} catch (const std::exception& ex) {
Log(LogWarning, "Checkable")
<< "Exception occurred during notification '" << notification->GetName() << "' for checkable '"
<< GetName() << "': " << DiagnosticInformation(ex, false);
}
} else {
// Cold startup phase. Stash notification for later.
Log(LogNotice, "Notification")
<< "Notification '" << notification->GetName() << "': object authority hasn't been updated, yet. Stashing notification.";
notification->GetStashedNotifications()->Add(new Dictionary({
{"notification_type", type},
{"cr", cr},
{"force", force},
{"reminder", false},
{"author", author},
{"text", text}
}));
}
}
}
std::set<Notification::Ptr> Checkable::GetNotifications() const
{
std::unique_lock<std::mutex> lock(m_NotificationMutex);
return m_Notifications;
}
void Checkable::RegisterNotification(const Notification::Ptr& notification)
{
std::unique_lock<std::mutex> lock(m_NotificationMutex);
m_Notifications.insert(notification);
}
void Checkable::UnregisterNotification(const Notification::Ptr& notification)
{
std::unique_lock<std::mutex> lock(m_NotificationMutex);
m_Notifications.erase(notification);
}
void Checkable::FireSuppressedNotifications()
{
if (!IsActive())
return;
if (IsPaused())
return;
if (!GetEnableNotifications())
return;
int suppressed_types (GetSuppressedNotifications());
if (!suppressed_types)
return;
int subtract = 0;
{
LazyInit<bool> wasLastParentRecoveryRecent ([this]() {
auto cr (GetLastCheckResult());
if (!cr) {
return true;
}
auto threshold (cr->GetExecutionStart());
Host::Ptr host;
Service::Ptr service;
tie(host, service) = GetHostService(this);
if (service) {
ObjectLock oLock (host);
if (!host->GetProblem() && host->GetLastStateChange() >= threshold) {
return true;
}
}
for (auto& dep : GetDependencies()) {
auto parent (dep->GetParent());
ObjectLock oLock (parent);
if (!parent->GetProblem() && parent->GetLastStateChange() >= threshold) {
return true;
}
}
return false;
});
if (suppressed_types & (NotificationProblem|NotificationRecovery)) {
CheckResult::Ptr cr = GetLastCheckResult();
NotificationType type = cr && IsStateOK(cr->GetState()) ? NotificationRecovery : NotificationProblem;
bool state_suppressed = NotificationReasonSuppressed(NotificationProblem) || NotificationReasonSuppressed(NotificationRecovery);
/* Only process (i.e. send or dismiss) suppressed state notifications if the following conditions are met:
*
* 1. State notifications are not suppressed at the moment. State notifications must only be removed from
* the suppressed notifications bitset after the reason for the suppression is gone as these bits are
* used as a marker for when to set the state_before_suppression attribute.
* 2. The checkable is in a hard state. Soft states represent a state where we are not certain yet about
* the actual state and wait with sending notifications. If we want to immediately send a notification,
* we might send a recovery notification for something that just started failing or a problem
* notification which might be for an intermittent problem that would have never received a
* notification if there was no suppression as it still was in a soft state. Both cases aren't ideal so
* better wait until we are certain.
* 3. The checkable isn't likely checked soon. For example, if a downtime ended, give the checkable a
* chance to recover afterwards before sending a notification.
* 4. No parent recovered recently. Similar to the previous condition, give the checkable a chance to
* recover after one of its dependencies recovered before sending a notification.
*
* If any of these conditions is not met, processing the suppressed notification is further delayed.
*/
if (!state_suppressed && GetStateType() == StateTypeHard && !IsLikelyToBeCheckedSoon() && !wasLastParentRecoveryRecent.Get()) {
if (cr->GetState() != GetStateBeforeSuppression()) {
Checkable::OnNotificationsRequested(this, type, cr, "", "", nullptr);
}
subtract |= NotificationRecovery|NotificationProblem;
}
}
for (auto type : {NotificationFlappingStart, NotificationFlappingEnd}) {
if (suppressed_types & type) {
bool still_applies = NotificationReasonApplies(type);
if (still_applies) {
if (!NotificationReasonSuppressed(type) && !IsLikelyToBeCheckedSoon() && !wasLastParentRecoveryRecent.Get()) {
Checkable::OnNotificationsRequested(this, type, GetLastCheckResult(), "", "", nullptr);
subtract |= type;
}
} else {
subtract |= type;
}
}
}
}
if (subtract) {
ObjectLock olock (this);
int suppressed_types_before (GetSuppressedNotifications());
int suppressed_types_after (suppressed_types_before & ~subtract);
if (suppressed_types_after != suppressed_types_before) {
SetSuppressedNotifications(suppressed_types_after);
}
}
}
/**
* Re-sends all notifications previously suppressed by e.g. downtimes if the notification reason still applies.
*/
void Checkable::FireSuppressedNotificationsTimer(const Timer * const&)
{
for (auto& host : ConfigType::GetObjectsByType<Host>()) {
host->FireSuppressedNotifications();
}
for (auto& service : ConfigType::GetObjectsByType<Service>()) {
service->FireSuppressedNotifications();
}
}
/**
* Returns whether sending a notification of type type right now would represent *this' current state correctly.
*
* @param type The type of notification to send (or not to send).
*
* @return Whether to send the notification.
*/
bool Checkable::NotificationReasonApplies(NotificationType type)
{
switch (type) {
case NotificationProblem:
{
auto cr (GetLastCheckResult());
return cr && !IsStateOK(cr->GetState());
}
case NotificationRecovery:
{
auto cr (GetLastCheckResult());
return cr && IsStateOK(cr->GetState());
}
case NotificationFlappingStart:
return IsFlapping();
case NotificationFlappingEnd:
return !IsFlapping();
default:
VERIFY(!"Checkable#NotificationReasonStillApplies(): given type not implemented");
return false;
}
}
/**
* Checks if notifications of a given type should be suppressed for this Checkable at the moment.
*
* @param type The notification type for which to query the suppression status.
*
* @return true if no notification of this type should be sent.
*/
bool Checkable::NotificationReasonSuppressed(NotificationType type)
{
switch (type) {
case NotificationProblem:
case NotificationRecovery:
return !IsReachable(DependencyNotification) || IsInDowntime() || IsAcknowledged();
case NotificationFlappingStart:
case NotificationFlappingEnd:
return IsInDowntime();
default:
return false;
}
}
/**
* E.g. we're going to re-send a stashed problem notification as *this is still not ok.
* But if the next check result recovers *this soon, we would send a recovery notification soon after the problem one.
* This is not desired, especially for lots of checkables at once.
* Because of that if there's likely to be a check result soon,
* we delay the re-sending of the stashed notification until the next check.
* That check either doesn't change anything and we finally re-send the stashed problem notification
* or recovers *this and we drop the stashed notification.
*
* @return Whether *this is likely to be checked soon
*/
bool Checkable::IsLikelyToBeCheckedSoon()
{
if (!GetEnableActiveChecks()) {
return false;
}
// One minute unless the check interval is too short so the next check will always run during the next minute.
auto threshold (GetCheckInterval() - 10);
if (threshold > 60) {
threshold = 60;
} else if (threshold < 0) {
threshold = 0;
}
return GetNextCheck() <= Utility::GetTime() + threshold;
}