Hi Team,
We have a requirement for a real time highly available application where we are deploying NDB cluster (we use CGE) with 10+ nodes. The overall availability requirements is 99.999%.
With replication factor 2 per shard we are able to achieve 99.99% (with individual machine at 99%). However for higher mark, we need to have higher replication factor (replication to another cluster and bringing that up in case of failure would not be near real time for us, so that’s not an option we are considering).
With this context, please help me with my following questions
1) Will it be advisable to have replication factor of more than 2. Say 3 in my case; will that be supported by MySQL team.
2) If point 1 is not an option, how we can reach to 99.999% availability with MySQL cluster configuration and infrastructure layout (avoiding classic cascading failure scenario)
Thanks,
Amit
We have a requirement for a real time highly available application where we are deploying NDB cluster (we use CGE) with 10+ nodes. The overall availability requirements is 99.999%.
With replication factor 2 per shard we are able to achieve 99.99% (with individual machine at 99%). However for higher mark, we need to have higher replication factor (replication to another cluster and bringing that up in case of failure would not be near real time for us, so that’s not an option we are considering).
With this context, please help me with my following questions
1) Will it be advisable to have replication factor of more than 2. Say 3 in my case; will that be supported by MySQL team.
2) If point 1 is not an option, how we can reach to 99.999% availability with MySQL cluster configuration and infrastructure layout (avoiding classic cascading failure scenario)
Thanks,
Amit