azure - How frequently the Traffic Manager monitors endpoints -
how traffic manager monitors endpoints? it's obvious it's not event driven (when endpoint down takes up-to 30 secs - 2.5 mins identify status of endpoint per observations). can configure frequency, cannot see configuration this.
is there relationship between traffic manager monitoring interval , ttl?
this may general question, real issue experience service downtime in fail on scenario (fail on of primary). understand effect in ttl until client dns cache expires calling cached endpoint. spent lot of time on , have narrowed down specific question.
issue there delay in traffic manager identifying endpoint status after it's stopped or started. need logical explanation this, not find azure reference explains this.
traffic manager settings
i need understand delay , plan down time.
i have gone through same issue. check link, explains monitoring behaviour
the monitoring system performs get, not receive response in 10 seconds or less. performs 3 more tries @ 30 second intervals. means @ most, takes approximately 1.5 minutes monitoring system detect when service becomes unavailable. if 1 of tries successful, number of tries reset. although not shown in diagram, if 200 ok message(s) come more 10 seconds after get, monitoring system still count failed check.
this explains 30-2 mins delay.
basically maximum delay 1.5 mins + ttl per details.
Comments
Post a Comment