gatus/watchdog/watchdog.go

67 lines
2.1 KiB
Go
Raw Normal View History

2019-09-05 01:37:13 +02:00
package watchdog
import (
"log"
"sync"
"time"
"github.com/TwinProduction/gatus/config"
"github.com/TwinProduction/gatus/core"
"github.com/TwinProduction/gatus/metric"
2020-12-31 12:49:41 +01:00
"github.com/TwinProduction/gatus/storage"
)
var (
// monitoringMutex is used to prevent multiple services from being evaluated at the same time.
// Without this, conditions using response time may become inaccurate.
monitoringMutex sync.Mutex
)
2020-08-18 02:25:29 +02:00
// Monitor loops over each services and starts a goroutine to monitor each services separately
func Monitor(cfg *config.Config) {
for _, service := range cfg.Services {
// To prevent multiple requests from running at the same time, we'll wait for a little bit before each iteration
time.Sleep(1111 * time.Millisecond)
go monitor(service)
}
}
2020-08-18 02:25:29 +02:00
// monitor monitors a single service in a loop
func monitor(service *core.Service) {
cfg := config.Get()
for {
if !cfg.DisableMonitoringLock {
// By placing the lock here, we prevent multiple services from being monitored at the exact same time, which
// could cause performance issues and return inaccurate results
monitoringMutex.Lock()
}
if cfg.Debug {
log.Printf("[watchdog][monitor] Monitoring group=%s; service=%s", service.Group, service.Name)
}
result := service.EvaluateHealth()
metric.PublishMetricsForService(service, result)
2020-11-27 00:09:01 +01:00
UpdateServiceStatuses(service, result)
log.Printf(
"[watchdog][monitor] Monitored group=%s; service=%s; success=%v; errors=%d; duration=%s",
service.Group,
service.Name,
result.Success,
len(result.Errors),
result.Duration.Round(time.Millisecond),
)
HandleAlerting(service, result)
if cfg.Debug {
log.Printf("[watchdog][monitor] Waiting for interval=%s before monitoring group=%s service=%s again", service.Interval, service.Group, service.Name)
}
if !cfg.DisableMonitoringLock {
monitoringMutex.Unlock()
}
time.Sleep(service.Interval)
}
}
2020-11-27 00:09:01 +01:00
// UpdateServiceStatuses updates the slice of service statuses
func UpdateServiceStatuses(service *core.Service, result *core.Result) {
2021-02-03 05:06:34 +01:00
storage.Get().Insert(service, result)
2020-11-27 00:09:01 +01:00
}