Graphite

Possible to precompute rollup data?

Asked by ZachG

Hi! I have a dataset in graphite which is broken down by many fields. When I try and do a query along the lines of:

X.*.Y.*.Z.*.*

it can take a very long time to generate a graph. Is there anyway to precompute this specific query? As in every time I send data to carbon and it notices that it fits a certain pattern it just adds that to data to a specific data point?

If this is not presently possible I am likely to add this feature and would very much welcome feedback and ideas on the best way to go about doing it.

It would also be great if the maintainers would give input as to the necessary conditions to perhaps have this feature incorporated in the upstream branch.

Thanks
-Zach

Question information

Language:
English Edit question
Status:
Solved
For:
Graphite Edit question
Assignee:
No assignee Edit question
Solved by:
ZachG
Solved:
Last query:
Last reply:
Revision history for this message
ZachG (zgold550) said : 		#1

I have implemented this feature, it works well. You define a json file of template patterns which match incomming paths and if an incomming path matches it gets added to the aggregate for that 60 second time period. If the maintainers are interested in this feature I am happy to offer patches.

Revision history for this message
chrismd (chrismd) said : 		#2

Patches are always welcome, can you email it to me? Thanks.