Hi all,
I have a very big, GKE cluster (near 300 nodes) with nearly 3,000 applications deployed to it.
Obviously, all these applications do produce logs, and in some cases, we do have very verbose ones.
At the moment, all logs will go to the Default Bucket, but recently we noticed that in the bill, the item referring to logs was growing massively. So now, I have to find a way to reduce them.
Well, we do have applications that are not very critical enough to have logs ingested. And in here I include, for example, monitoring-related applications.
When some is error is happening, is good to know what the issue is, and we can inspect pods at the moment, but no need on keeping logs for more than that. I know i can configure some exclusions on the sync associated with the Default Bucket, and they will stop being visible on Stack Driver; however, they are still visible when we inspect the pods directly.
I know i can go app by app, and review all logs, but for the time being, I need a quick solution to cut costs.
So my question in here is,
do these logs still have a cost to me, even if they are dropped, by sync exclusion?
regards
Kubernetes version: 1.26
Cloud being used: GKE - GCP
Installation method: GCLOUD SDK