February 2017 Summaries

ScyllaDB 1.6.1 is a bugfix release of the ScyllaDB 1.6 stable branch, offering backward compatibility and supporting rolling upgrades, making it a recommended update for users. This release addresses specific bugs such as issues with Prometheus metrics and potential crashes under certain disk write conditions. Users are encouraged to upgrade from previous versions, and feedback on any encountered problems is welcomed by the ScyllaDB team.

ScyllaDB 1.5.1, a bugfix release for the ScyllaDB 1.5 stable branch, was announced by Tzach Livyatan on February 21, 2017, addressing several issues such as excessive disk usage during sstable resharding, installation errors due to existing users, and problems with the setup script not listing local SSD disks on Google Cloud VMs. This release supports backward compatibility and rolling upgrades, encouraging users to transition to the latest stable version, ScyllaDB 1.6. Key fixes include resolving RPC message tracing issues, optimizing Prometheus metric generation to reduce latency, and preventing potential crashes when disk writes are delayed. Users are encouraged to report any encountered problems, ensuring continuous improvement and support for the ScyllaDB platform.

Samsung's Memory Solutions Lab conducted a performance benchmark using the Yahoo! Cloud Serving Benchmark (YCSB) to compare ScyllaDB and Apache Cassandra, revealing that ScyllaDB significantly outperforms Cassandra. The test involved a ScyllaDB cluster with three servers, each utilizing NVMe SSDs, and demonstrated that ScyllaDB can perform 10 to 37 times better than Apache Cassandra when handling a 2TB dataset, even achieving high performance with a lower hit rate where SSDs handled a portion of requests. The study measured latency across different YCSB workloads, showing ScyllaDB's ability to maintain minimal latency while processing high input/output operations per second (IOPS) with a millisecond latency range between 0.6ms to 2.6ms. Samsung's results highlight ScyllaDB's capability to operate efficiently on high-end hardware, suggesting the use of larger machines to optimize performance and reduce total cost of ownership.

ScyllaDB, a distributed database known for its high performance, offers methods to efficiently perform full table scans, which are essential for data analytics despite being less common than operations on individual partitions. Traditional full table scans can be slow due to limited server and client parallelism, but ScyllaDB 1.6 introduces improvements that enhance performance by automatically tuning paging and reducing CPU consumption. The database utilizes a token function to evenly distribute data across nodes, allowing for parallel scans by dividing the token range into sub-ranges, which can be processed simultaneously by multiple threads. This approach maximizes the utilization of server nodes and client cores, though the number of parallel queries should be adjusted based on cluster size to prevent a "processing tail." Additionally, the use of message queues can coordinate these parallel scans effectively. ScyllaDB's improvements make it easier to query entire tables efficiently, but integrated tools like Presto and Apache Spark can also be used for full table scans, offering ease of use for ad-hoc queries at the potential cost of efficiency.

Ensuring the security of ScyllaDB clusters is crucial to prevent unauthorized access and potential data breaches, as highlighted by recent hacker attacks targeting various NoSQL databases. Key vulnerabilities include open ports such as CQL/Thrift, inter-node communication, JMX, REST API, and Prometheus client ports, which can allow attackers to connect and manipulate data if not properly secured. To safeguard against these risks, it is essential to close these ports through firewall configurations, utilize VPNs, enable SSL encryption for communication, and enforce client authentication with strong passwords. Additionally, modifying the replication strategy for the system_auth keyspace and managing user permissions can further enhance security by ensuring authentication data remains accessible and minimizing the potential impact of compromised credentials.

ScyllaDB 1.6, a minor release of the open-source NoSQL database compatible with Apache Cassandra, introduces several improvements, including superior performance, consistently low latency, and experimental Debian support. This release focuses on bug fixes for versions 1.6 and 1.5, urging users on older versions to upgrade. Notable features include an upgradable AMI, fetch size auto-tuning, and optimized multi-partition range scans to reduce CPU and I/O overhead. The release also brings changes to the metrics format and names, requiring users to update their monitoring tools accordingly. Key bug fixes address issues such as data streaming in repairs, bloom filter overhead reduction, and various installation and operational challenges, enhancing the overall stability and reliability of ScyllaDB.