Timeline for Cassandra in multiple data center - EC2MultiRegionSnitch
Current License: CC BY-SA 3.0
6 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jun 21, 2013 at 10:06 | comment | added | Abzal Kalimbetov | local data is local to data center that is in one region. In other words, I do not want to replicate to all data centers, but I want one data center to have all replica of others. | |
| Jun 21, 2013 at 7:38 | history | migrated | from stackoverflow.com (revisions) | ||
| Jun 21, 2013 at 5:00 | comment | added | Nishant | what do you mean by "local data"? Look at consistency level. If you have set consistency level "LOCAL_QUORUM" the request will never wait for data to be returned from other data center. And, if you write with "LOCAL_QUORUM" data will be written to the data-center the client is connected to and then the replication will take place in background. So, connection speed between data-centers should not matter. You can tune it to performance requirement. | |
| Jun 21, 2013 at 4:57 | comment | added | Abzal Kalimbetov | cross-data center requests occur very rarely | |
| Jun 21, 2013 at 4:55 | comment | added | Abzal Kalimbetov | Hmm, that is not what I need, because network speed between data centers are slow. So I do not want replication occur between data centers. Can I configure Cassandra so that local data is stored local only and be accessed from other data centers? | |
| Jun 21, 2013 at 4:49 | history | answered | Nishant | CC BY-SA 3.0 |