hotfix(health): increase timeout values and periods

- run small check every 60s, from 15s
- small check (icmp/dns) initial timeout from 3s to 10s
- small check (icmp/dns) timeout increase from 1s to 10s
- full check initial timeout increased from 10s to 20s
- full check extra timeout increase from 3s to 10s
This commit is contained in:
Quentin McGaw
2025-10-19 23:27:02 +00:00
parent 36fe349b70
commit 31a36a9250

View File

@@ -83,7 +83,7 @@ func (c *Checker) Start(ctx context.Context) (runError <-chan error, err error)
done := make(chan struct{})
c.done = done
c.smallCheckName = "ICMP echo"
const smallCheckPeriod = 15 * time.Second
const smallCheckPeriod = time.Minute
smallCheckTimer := time.NewTimer(smallCheckPeriod)
const fullCheckPeriod = 5 * time.Minute
fullCheckTimer := time.NewTimer(fullCheckPeriod)
@@ -131,8 +131,8 @@ func (c *Checker) smallPeriodicCheck(ctx context.Context) error {
ip := c.icmpTarget
c.configMutex.Unlock()
const maxTries = 3
const timeout = 3 * time.Second
const extraTryTime = time.Second // 1s added for each subsequent retry
const timeout = 10 * time.Second
const extraTryTime = 10 * time.Second // 10s added for each subsequent retry
check := func(ctx context.Context) error {
if c.icmpNotPermitted {
return c.dnsClient.Check(ctx)
@@ -151,10 +151,10 @@ func (c *Checker) smallPeriodicCheck(ctx context.Context) error {
func (c *Checker) fullPeriodicCheck(ctx context.Context) error {
const maxTries = 2
// 10s timeout in case the connection is under stress
// 20s timeout in case the connection is under stress
// See https://github.com/qdm12/gluetun/issues/2270
const timeout = 10 * time.Second
const extraTryTime = 3 * time.Second // 3s added for each subsequent retry
const timeout = 20 * time.Second
const extraTryTime = 10 * time.Second // 10s added for each subsequent retry
check := func(ctx context.Context) error {
return tcpTLSCheck(ctx, c.dialer, c.tlsDialAddr)
}