tailhook / cantal Goto Github PK

We use auto-generated ID when there is no such file. But it turns out to be a problem, because sometimes file is written after cantal is started and we see cantal with wrong machine id.

We need to require it if -cluster-name is specified (i.e. we expect that it's clustered cantal not a standalone one)

As advertised in this writeup

To make scheduling decisions we need two things: absolute value of some metric and a trend.

Currently, we fetch all the datapoints and compute math on top. But it's better to aggregate values in cantal itself. We already have aggregation function facility. We just need a few functions:

1. Gradient value over N data points fit using least squares
2. Moving average over N data points
3. Maximum value over N data points (probably applied after moving average)

It states:

Description: Package created with checkinstall 1.6.2

Hi!
I would very much like to be able to get the values of individual metrics via JSON, passing the key and get metric value and last update timestamp as result.

At the main page, in the charts, January has number 00, not 01.
img

Cantal tries to track all the processes in the system. And it keeps the data for 1-2 hours.

If there are two much short-lived processes in the system they occupy too much memory, but the data they generate is pretty much useless. We should probably clean short-lived processes faster to drop CPU usage by cantal on such system.

This is needed if you use --bind option, but also need to access cantal at localhost

It was in old version of cantal, and wasn't implemented since a refactoring of UI.

Currently, to add a host to a cluster you need to use curl:

curl -X POST -d "{\"addr\":\"1.2.3.4:22682\"}" http://localhost:22682/add_host.json