Nathaniel Cook

@joelegasse Makes sense, thanks.

@affo This sounds powerful, how does it differ from the `reduce` function that we have?

>we'll need to run a 2-3 day load test and confirm that this new data structure continues to scale well without slowing anything down This sounds great, there is a...

Thanks for the quick feedback I have update the PR to be off main branch. Additionally I updated all the tests and introduced a `Digest` type. Now scenario, transaction and...

Thanks for the hints on the failing tests. I still haven't fixed them yet but I have learned a few things: 1. The change to tdigest has caused several failures...

Thanks for the diff, I did something similar and got past that part of the test failures. > Did you already rebase against main where tests are working? Yes, and...

@Surtr04 Thanks for the thorough report. We are currently working on some optimizations to make the type of Flux query you are using significantly faster. We are specifically improving the...

@ayebayley Startup is likely blocked trying to talk to InfluxDB. Turn on debug logging to make sure, or otherwise disable InfluxDB.

@ayebayley This line here `[srv] 2016/10/25 22:43:06 D! opening service: *storage.Service` doesn't have a corresponded `opened` line, which means that the Kapacitor process is stuck opening that service. That service...

@ayebayley UDP will drop packets so you will have to make sure you are configuring your UDP kernel settings appropriately and monitoring any dropped data. What version are you using?