Loki HA cluster with caching

Grafana Loki
1

I have a couple of questions regarding Loki HA cluster with caching.

  1. How to make Loki fault tolerant, what is the best High Availability solution ?

  2. Can we just use Loki instances that access the same storage backend (which holds the logs)?

  3. Is there any local data stored for Loki? If so, how do we sync that data in a HA scenario and how can we use caching in a HA scenario. Redis, memcached, … ?

Can anybody help me please with above questions.

Thanks.

2

Hello @keerthi ,

I know this is an old question but I thought I’d bump it as I’m looking at the same things at the moment.

There are many ways to deploy Loki. I have chose the current recommended method of Single Store Loki (boltdb-shipper index type) (at least as far as I can tell, this is the recommended way at the moment, I could be wrong)

The above link also goes through HA, as in which services are recommended to have persistent storage to prevent data loss, etc. Running several instances of each service should ensure the rest of HA.

With the above configuration, all long term storage is in your object store.

That takes care of 1. and 2.

I’m just about to start looking at caching. Both Redis and memcached are supported. The cache is there to decrease load on the object store (I think). Can’t say much more than that for now.

If you have figured out answers to your questions it would be interesting to hear your findings :slight_smile:

3

This topic was automatically closed 365 days after the last reply. New replies are no longer allowed.