3 ms·
Requiring a dedicated volume with at least 50 GB of space and 1500 privisioned IOPS just for etcd [1] seems excessive to me. How big a cluster, e.g. how many no
by mwcampbell 7y ago
Requiring a dedicated volume with at least 50 GB of space and 1500 privisioned IOPS just for etcd [1] seems excessive to me. How big a cluster, e.g. how many nodes and pods, is this for?
[1]: https://gravitational.com/gravity/docs/ver/6.x/requirements/#etcd-disk https://gravitational.com/gravity/docs/ver/6.x/requirements/...
- alexk 7y agoWe are staying on the safe side here on our recommendations. We have seen many scenarios when high latency on volumes on AWS creates a lot of problems even for small 3 node clusters.
- AaronFriel 7y agoEven small clusters with K8s operators doing complex work (e.g.: autoscaling) can generate a lot of etcd traffic updating state. I'm not sure how (etcd) stateless Istio and other meshes are, too.