Reputation: 1750
Although connection pooling has been mentioned as one of the features of the Cassandra's C++ driver(https://datastax.github.io/cpp-driver/), no details have been given regarding its implementation in C++. Can anybody help regarding the configuration of connection pools in the C++ program?
Upvotes: 1
Views: 403
Reputation: 87234
For high-level description of what connection pooling is, you can use documentation from Java driver (unfortunately there is no corresponding documentation for C++ driver).
C++ driver also have slightly different configuration options - for example, there is only one configuration for number of connections, while Java driver allows to specify different numbers for local & remote clusters.
You can use function cass_cluster_set_core_connections_per_host
to specify a minimal number of network connections to host, and cass_cluster_set_max_connections_per_host
to specify maximal number of connections to host - driver will resize pool if necessary. And function cass_cluster_set_max_concurrent_requests_threshold
specifies how many requests could be in-flight for one network connection. You can look to tests of C++ driver for examples.
But be very careful with changing these settings, as this may lead to increased resource consumption.
Upvotes: 2