<b>Q: One of my monitors keeps going up and down every few minutes. What is that?</b>

<b>Q: One of my monitors keeps going up and down every few minutes. What is that?</b>

A: That's flapping, and it usually means a real but intermittent problem, not a monitoring glitch. Common causes: one unhealthy server behind a load balancer (you only fail when the request lands on it), a connection pool maxing out under bursts, or a memory leak that triggers restarts on a cycle.

Don't just widen the threshold to silence it. Instead:
— Note the failure rate. Failing ~1 in N checks points straight at one bad node out of N behind the balancer.
— Check whether failures cluster at a time of day, which suggests a cron or traffic spike exhausting resources.

Use flap detection to stop the alert spam (alert once on "flapping detected"), then fix the root cause. Silencing without investigating just hides a partial outage your users are feeling.

Got a question? Drop it in the comments.