If a member goes down, its backup replica that also holds the same data, will dynamically redistribute the data including the ownership and locks on them to remaining live nodes. As a result, no data will get lost.
Several questions
a) If a backup of 1 is configured for a member in the cluster, does this mean that only one member in the cluster has a backup of that member? Or is it a backup backup?
b) Therefore, a) If the members and their backups are down, will there be data loss?
c) If there is a write-through strategy and the members and their backups fail and there is data loss, is there a mechanism for dynamically recovering data (using write-through mechanism or other methods)?
Thanks in advance
a) Configure the backup at the data structure level. So you can say this: This map has 1 synchronous backup… and the mapping is zero… or 2 or…
The backup can be synchronous (Meaning that there is a write of changes to the backup), or it can be a write backup (so the backup is completed at some point in time). Asynchronous and synchronous backups are independently configured at the data structure level.
b ) Yes. This is why if you have such high availability requirements, you can configure multiple backups.
c) You can use maploader/mapstore strategy to configure some data structures (e.g. map/queue); this Makes for example writing your changes to disk/database, etc.
The hazelcast documentation points out this
If a member goes down, its backup replica that also holds the same data, will dynamically redistribute the data including the ownership and locks on them to remaining live nodes. As a result, no data will get lost.
Several questions
a) If a backup of 1 is configured for a member in the cluster, does this mean that only 1 member in the cluster has a backup of that member? Or is it a backup backup?
b) Therefore, a) If the members and their backups are down, will there be data loss?
c) If there is a write-through strategy and the members and their backups fail and there is data loss, is there a mechanism for dynamically recovering data (using write-through mechanism or other methods)?
Thanks in advance
A few answers:
a) Configure backup at the data structure level. So you can say this: this map has 1 synchronous backup… and the map is zero… or 2 or…
The backup can be synchronous (meaning that there is a write of the changes to the backup) , It can also be a write backup (so the backup is completed at some point in time). Asynchronous and synchronous backups are independently configured at the data structure level.
b) Yes. This is why if you have such a high For the availability requirements, you can configure multiple backups.
c) You can use the maploader/mapstore strategy to configure some data structures (such as map/queue); this allows for example to write your changes to disk/database, etc. .