Details
-
Bug
-
Resolution: Fixed
-
Major
-
None
-
2.0
-
None
-
Security Level: Public
-
None
Description
- I think it's quite confusing to refer to "replica partitions" instead of vbuckets as they are known throughout the rest of the docs
- I also think it's quite confusing to use the term "replica node" when a) we explicitly do NOT have replica nodes and b) customers are frequently confused because other technologies DO have master/slave
- Typo: "Notice that a replica node may also be simultaneously by handling"
- Typo: "...in a disk write queue to eventually stored on disk at the replica node"
- Typo and very confusing: "If one of the nodes in the system, the replica vBuckets are enabled in place of the vBuckets that are unavailable at the failed node."
- The last paragraph should be "introduced" better to note that it is referring to handling failovers. It's also slightly incorrect in that it doesn't mention the fact that the cluster will push-out an updated map upon failover (the client doesn't have to request a new one)
- I think the failover should be pointed to in a different section where it can be handled more completely.