I have the following issue: I want to collect data from several locations (or servers). Now I want to store all collected data locally at that location (via carbon-cache, storage-schemas and so on), but in addition to that I want to aggregate (carbon-aggregator) this information (to reduce network load) and send it to another (main or central) location where all the important metrics from different locations are stored. The reason for this is, that internet connection to several locations is not always that reliable as some locations are in developing countries. Therefore a local backup with all the data would reduce the risk of losing important data. I thought of a setup where I use carbon-relay to duplicate the data. One goes into carbon-cache, the other one into carbon-aggregator which forwards it to the main location. Is this even possible? How would I achieve this configuration? Or what would be another possibility to achieve my goal. I appreciate any help.
3
votes
1 Answers
2
votes
Yes, you'd use carbon-relay to do this.
Configuring your carbon.conf to do relaying: http://graphite.readthedocs.org/en/latest/carbon-daemons.html
Then you need to configure relay-rules.conf: http://graphite.readthedocs.org/en/latest/config-carbon.html
Here's a more detailed explanation with an example: https://answers.launchpad.net/graphite/+question/126700
