Open JoshuaMcManus opened 4 years ago
Should we use fdbcli status
or the new "special key space" to report this information?
The special key space provides more flexibility (users can query just a specific key for specific info);
I think the plan is for this to be a part of status, but it doesn't necessarily need to be limited to there.
Currently, a cluster's status reports the fault tolerance of the entire cluster, however when a single cluster is deployed using multi-region configuration a failure in the remote side will change the overall cluster's fault tolerance.
One could track the actual state of the cluster more easily if we tracked the fault tolerance of each instance type within each region. I.e. report the fault tolerance of storage processes in the primary region, storage processes in the remote region, log processes in the primary region, etc.