Cassandra – In-Memory Tables Explained

cassandranosql

I am planning to use Cassandra in memory tables for my particular work load. THe problem is I am worried if the node fails then what will happen to the data in memory that was not flushed to disk ?

Would I loose that data, or not ?

Best Answer

Would I loose that data, or not ?

Good question! As outlined in the DataStax documentation section on The Write Path To Compaction, you would not lose data.

As you can see, when Cassandra processes a write operation, data is written to both the commit log (on-disk) and to the memtable (which you referred to). In case of a "plug out of the wall" event, the writes are still in the commit log. When Cassandra restarts on a node, it first replays the commit log, and verifies that all of its data have been written to disk. If it finds data in the commit log that was not flushed to disk, it writes to disk at that time.

Related Solutions

A good way to copy data from one Cassandra ColumnFamily to another on the same Keyspace (like SQL’s INSERT INTO)

cqlsh -k mykeyspace -e 'COPY fromTable(columnNames) TO STDOUT' | head -n -1 | 
cqlsh -k mykeyspace -e 'COPY toTable(columnNames) FROM STDIN'

Cassandra – Cluster Monitoring

You can use check Java Heap memory for each node. Like Total Java Heap Memory and Using Java Heap Memory.
One most important is CPU Utilization for each node.
Set alerts for errors. system.log have lots of information about.
You may set alerts for data disk and log disk.
Heartbeat check of servers like if you dont receive in few min it will get alerts to you.
Also dropped mutations and hinted hindoff clear alerts.

Basically you need to start observe system.log and will get more and more error for monitoring.

Best Answer

Related Solutions

A good way to copy data from one Cassandra ColumnFamily to another on the same Keyspace (like SQL’s INSERT INTO)

Cassandra – Cluster Monitoring

Related Question