My new client in Boise pointed out their latency graphs spiked every day at 2 PM, which finally made me realize I'd been ignoring scheduled backup traffic on their primary circuit for months.

Has anyone else found a good way to automatically flag internal network events on ISP performance dashboards before a customer calls?

2 comments

2 Comments

wendy_carr2d ago

Why would you even want to automate that? Isn't the whole point of monitoring to catch the stuff you didn't plan for? If you start flagging every known internal event, your dashboards just become noise. You'll end up ignoring the real alerts because they're buried in scheduled task notes. Plus, figuring out what's "normal" for each client takes time you could spend fixing actual problems. Sometimes a customer call is the best alert system because it forces a real conversation.

jessicad982d ago

Actually the good monitoring tools let you tag those known events as "expected" so they don't even hit the main alert feed... they go to a separate log. That way your main dashboard stays clean for the real surprises. It's not about making noise, it's about teaching the system what to ignore so you can focus.