Jut Blog Instant Replay for your Data Using Pace in Juttle

Where would this world be without instant replays? Let's face it, sporting events would be half as long and not nearly as much fun. And when the game is over, what would we talk about at work the next day? Would we be stuck looking at photos of that just-barely offsides call from the ref?And, because I'm talking about instant replays I thought I'd start with a link to the "8 Most Decisive Instant Replay Moments in History." Before I go any further - you're welcome. I just gave you an excuse to watch some great old sports clips in the name of "data analytics." At the same time, I just got to quote Maxim Magazine as a definitive source of research. Hey, there's a first time for everything.Analytics Should Have Instant Reply Too!OK, on to that data stuff. As an operations data hub that excels at analytics of software systems, I think you'd be a bit surprised if Jut didn't let you pull up some old data, say from an hour ago, and show you a visualization. We'd agree. But sometimes a static picture just isn't enough. You want to replay the data, acting as if it were live.That's what 'pace' does. It's a Juttle processor that typically sits between your dataflow computation and your visualization, giving you control of how quickly you'd like to play back the data.Streaming Analytics with Historical Data ExampleLet's start with a sample program, one that grabs the last 5 minutes of data and then continues to stream in live data. This kind of "historical + live" streaming analysis is a great example of what Jut allows you to do easily.In this example, Jut pulls the historical 5 minutes off disk, visualizes it instantly, and then batches up new data in 2 second intervals for live updates to the chart.Making Historical Data "Live"But let's say we wanted to instead stream the entire 5 minutes, essentially acting as if it's streaming in right now. You can do that with the pace processor (here's the documentation). Pace is really simple. in its base form, it just plays back data at the same rate you've specified in your program. Compare the same visualization described above, with one that's been paced:With advanced options, you can do things like control playback speed. In the above example, I actually sped up replay by 4x using the -x option:.. |pace -x 4 .... //plays back data at 4x original speedAnd there's also one more option, time-shifting, which you can try out yourself on the juttle playground. -from // example: [-from :day:]Pace is just one of the little niceties of Jut. We don't believe instant replays should just be used for confirming outs on second base or replays of big-wave madness at Mavericks. Have a little fun and use pace to play with your data too!