#34 Are we going to lose the data if any of the data node fails, where the distribution mechanism is followed to store the data?

[Lakshmi Prameela]

Hi,

I am investigating on Postgres XL cluster advantages/disadvantages to use this for our project.

I have done the XL set-up, created cluster set-up with one GTM, one Coordinator, three datanodes. Executed all basic SQL commands.

checking the High availability, Replication, Load balancing features.

From the documentation, came to know that, can make use of stand by server in case of fail over..

I have raised with few questions. Kindly provide me the clarifications.

1. If the data is distributed across the nodes of a particular table, if any of the datanode fails, are we going to lose the data of that particular node? If not, how can we recover them?

2. If the primary datanode fails, is any other datanode will act as primary node automatically? is this taken care in Postgres XL? If not, how to do this.

3. How to recover the data if any crash happens to database. There is some procedure available in the documentation, but I have observed that is not updated from the Postgres XL point of view.