PagerDuty

Issues that use to "fall through the cracks" or receive no attention/acknowledgement at all have been almost eliminated totally. The time for a technical resource to be alerted to an incident and acknowledge it has gone from 15-25-35-45mins to 5 minutes or less. Teams are now proactively alerted to issues that may turn into something big before the customer notices any impact.

The software just works. We have had almost thirty thousand alerts flow through PagerDuty since we first started using it. The software completely replaced a rather broken SMS / Email alert chain overnight and has been by far the most valuable product we have purchased for keeping our system uptime levels where they need to be.

Lastly a personal experience which I will always cherish. It's a snowy winter day and you're on call, but you needed paperless, and cat food, and pedlyte and diapers and other bits, so you took your 2 year old to a big box store. You're all the way at the back of the store and your phone goes off with "Something's broken... Something's Broken... It's your Fault! ... It's your Fault!..." In the box store, with a full cart and a toddler.

Next thing you know... you're flying through the box store to the front, diapers falling out of the cart, sippy cup lost, toys missing, and toddler singing along with your phone since he now knows the words quite well. Get to the front, abandon the cart, the two children's voices still going full tilt, and get out to the parking lot just in the nick of time to production by ninja resizing a set of boxes as they are almost to the point of falling all over themselves.

No one noticed. Thanks to PagerDuty.

And now I'm *still* known as "The Something's Broken" dude by the managers of that store. They sing it to me every time I'm there 2 years later.

I have been using PagerDuty for 1 year now, and I'm with the product so far. The only negative thing about the PagerDuty is the big latency during the incidents.

It is quite a helpful tool. We use to get different services, we clearly got enough info on the incident row itself we do not have to go and check the whole incident.