graphite-dev team mailing list archive

[Sebastian Cohnen <question235941@xxxxxxxxxxxxxxxxxxxxx>]

New question #235941 on Graphite:
https://answers.launchpad.net/graphite/+question/235941

Hi,

I need a time-series database. And I have a multi-tenant requirement. I'd like to track metrics for a specific use case for a limited time and then access it for a very long period (indefinitely). The multi-tenant requirement should be easily solved by using a namespace.

The other thing is, that I need to track a number of metrics for a couple of hours and after that limited time period I only want to read the data back (no more updates, ever). The issue is, I want to be able to read this data indefinitely after I tracked the metrics without loosing precision.

What is the best way to achieve this (if at all possible) with graphite?

My current idea would be to specify  something like this in storage-schemas.conf:

[customer_projects]
pattern = projects\..*$
retentions = 1s:100y

I'm wondering when the aggregation (and reduction in resolution) is done. Is it scheduled? Depending on input / updates to that database?

Thanks for any input on that!


Best

Sebastian

-- 
You received this question notification because you are a member of
graphite-dev, which is an answer contact for Graphite.