Flequently Asked Question
Quick answers to common questions about Gluesync.
Licensing is available as an annual paid subscription per CPU core per data node or Data Under Management (DUM) / Unique Data Volume (UDV). Includes support and product updates.
Yes, vertical scaling increases the amount of resources per node. Horizontal scaling depends on the need: SQL to NoSQL and NoSQL to SQL are individual components each deployed in a docker container. Each node can work on multiple tables. Table workloads cannot be shared between nodes and are bound to a specific node that consumes the data from it.
Each GlueSync node manages its workload. If a node encounters a failure, it will be removed by the container orchestrator and a new one will handle the job.
GlueSync is a stateless service: the last checkpoints on both sides (SQL and NoSQL) are stored. If a failure occurs,GlueSync resumes from the point when the process or the container is restarted. You can opt to receive system alerts via email or grab them from the container logs.
Yes. Datasets are paginated per transaction and per row chunks. This can be fine-tuned using the GlueSync configuration file. Benchmark study demonstrated an optimized, tuned configuration file can process up to 2M rows per minute on 4 vCPUs and 8GB RAM.
Typically, on initialization, all data from a new table is replicated as a snapshot. Then only changed data is replicated. This is configurable, but the default is for continuous replication of changed data only.
GlueSync is lightweight and fine- tuned for NoSQL databases, supporting native vendor technologies. In addition, GlueSync can write to a RDBMS as well as reading from; it supports SQL views; and it can run an initial snapshot replication of all data and then process only changed data. Kafka does not provide these features.
An ODBC driver for NoSQL databases helps the database engine to translate the JSON document data model and enables your application layer to query that data structure like you are querying a common relational database (RDBMS). Unlike an ODBC driver, GlueSync is a real-time SQL to NoSQL two-way database replicator. This means it captures and transfers data from SQL tables or views and moves it to NoSQL JSON documents, and vice-versa.