StackOverflow2013

Note that there are some explanatory texts on larger screens.

plurals

POcassandra "data model failover", not looking for cluster failover
text
Body
copied!<p>Understand that Cassandra does not have a master-slave relationship which what I am after(peer to peer) and the data replication concept of Cassandra is one also the other feature I am looking forward for. </p> <p>However, is there a inbuilt function within Cassandra that will perform data model failover without the enhancement on the application. I am not after cluster failover but data model failover and possible of rollback of data model.</p> <p>And would also like to know if possible for database compare tool natively.</p> <p>Sorry for my bad English.</p> <p>Thanks.</p> <p>Etc:</p> <p>Data Model A is table1....table10 (for example)</p> <p>Data Model B is replica as Data Model A</p> <p>Scenario....</p> <p>Control Centre East - 3 nodes ( consistency must all table1....table10)</p> <p>Control Center West - 2 nodes ( eventual consistency table1....table10)</p> <p>Control Center North - 1 nodes ( consistency must be high for some tables table1...table5 and eventual consistency for the rest) </p> <p>Sequence of event</p> <ol> <li>All nodes read/write to Data Model set A (I am mean Client read and write Data Model set A). Data Model A is the one in production.</li> <li>Make changes to Data Model B (add new row, modify column value etc)</li> <li>Failover or changeover for all nodes from Data Model set A to Data Model set B. </li> <li>During the failover, some small amout of column value for identical row between Data Model A and B has to be copied over to from Datamodel A to B. </li> <li>After the failover, check if application giving wrong calculation(formula/output) based on the new Data Model B. </li> <li>Data model B is problematic, failover back to Data Model A for all nodes. Check/fix on Data Model B before running seq 2-3 again.</li> </ol> <p>Hopefully, it is not too long winded.</p>

Querying!

Guidance

An individual column

Larger individual text columns get their own page to allow for proper reading.

SQuiL has stopped working due to an internal error.

If you are curious you may find further information in the browser console, which is accessible through the devtools (F12).

Reload