The relationship between memory and the amount of data

Apr 23, 2012

I note fragmentation of data in the table is stored in each node, this can reduce memory consumption?

The relationship between memory and the amount of data

Apr 23, 2012

I note fragmentation of data in the table is stored in each node, this can reduce memory consumption?

Database Partitioning

Apr 23, 2012

Hi,

Depends on your k-factor, the whole data set may be partitioned onto multiple nodes.

Intro to how VoltDB Works

For example, if you have 2 nodes and k=0, roughly half of your data, assuming no skew, will be stored on each node. if you have 2 nodes and k=1, each node has a complete copy of your data.

Ning

Depends on your k-factor, the whole data set may be partitioned onto multiple nodes.

Intro to how VoltDB Works

For example, if you have 2 nodes and k=0, roughly half of your data, assuming no skew, will be stored on each node. if you have 2 nodes and k=1, each node has a complete copy of your data.

Ning

Reply to comment

Apr 24, 2012

I understand:

When I have 2 nodes, k=0, if node2 crashed, then node1 should automatically stop.

When I have 2 nodes, k=1, if node2 crashed, then node1 still running.

So the correct？

When I have 2 nodes, k=0, if node2 crashed, then node1 should automatically stop.

When I have 2 nodes, k=1, if node2 crashed, then node1 still running.

So the correct？

Yes, that is correct. Ning

Apr 24, 2012

I understand:

When I have 2 nodes, k=0, if node2 crashed, then node1 should automatically stop.

When I have 2 nodes, k=1, if node2 crashed, then node1 still running.

So the correct？

Yes, that is correct.

Ning