Maintain fast read access to Cassandra SSTables by tuning the use of the key
cache separately for each table.
Cassandra’s key cache stores a map of partition keys to row index entries, which enables
fast read access into SSTables.
-
Check the current cache size by using Cassandra's nodetool info utility.
The following nodetool info snippet shows sample key cache
metrics:
root@ip-10-123-5-62:/usr/local/tomcat/cassandra/bin# ./nodetool info
ID : 8ae22738-98eb-4ed1-8b15-0af50afc5943
Gossip active : true
Thrift active : true
Native Transport active: true
Load : 230.73 GB
Generation No : 1550753940
Uptime (seconds) : 634324
Heap Memory (MB) : 3151.44 / 10240.00
Off Heap Memory (MB) : 444.33
Data Center : us-east
Rack : 1c
Exceptions : 0
Key Cache : entries 1286950, size 98.19 MB, capacity 300 MB, 83295 hits, 83591 requests, 0.996 recent hit rate, 14400 save period in seconds
Row Cache : entries 0, size 0 bytes, capacity 0 bytes, 0 hits, 0 requests, NaN recent hit rate, 0 save period in seconds
Counter Cache : entries 0, size 0 bytes, capacity 50 MB, 0 hits, 0 requests, NaN recent hit rate, 7200 save period in seconds
Token : (invoke with -T/--tokens to see all 256 tokens)
-
If the
size parameter roughly equals the
capacity parameter, increase the cache size in the
prconfig/dnode/yaml/key_cache_size_in_mb/default
dynamic system setting, depending on your needs.
The key_cache_size_in_mb setting indicates the maximum
amount of memory for the key cache across all tables. The default value is
either 5% of the total JVM heap, or 100 MB, whichever is lower.