What is Database Replication and its Uses
Database replication refers to the frequent copying of data residing in a database from one server or computer to another. This results in a distributed database that all users can rapidly access to carry out their tasks, but without obstructing other users. There are two types of database replication:
- Synchronous database replication: data is copied from the client server to the model server, and then copied to all of the replica servers. This takes place prior to the client being notified that the data-base has been copied. In this way, the primary copy and the replica always remain in sync (hence the label).
- Asynchronous database replication: data is copied to the replica after the data has first been written to the primary storage. This can occur in near-real-time, or more commonly on a scheduled basis. In this way, updating the primary copy and the replica copy occurs asynchronously (hence the label).
Data replication supports two important cloud storage and security-related use cases:
- Disaster recovery (DR) and high availability (HA): A replicated database is essentially a live backup, which preserves a precise copy of a database on a separate system. As such, it is a ready-made source in the event that the source system fails or loses data. Furthermore, as required the source (primary) and target (secondary) systems can reverse roles, which provides failover and failback support.
- Automated backup and restore: Data replication can also be used as a method to automate backup and restore operations, but without requiring intermediate storage.
What is Cloud Data Migration?
Essentially, cloud storage migration is the process of shifting data, applications, and other business elements to a cloud computing environment. However, these days enterprises must be cautious of repeating past mistakes by investing in appliances that are inherently inflexible, and unable to extend to the cloud. Indeed, native cloud-based database warehouses are not a panacea!
Cloud data migration is highly complex, and there are crucial factors regarding aspects like data governance and privacy that block enterprises from implementing a wholesale cutover to the cloud. At the same time, for many enterprises investing in existing legacy infrastructure is a non-starter, because it is either excessively costly, or outright impossible. And even when it is feasible, the updated infrastructure has inherent flexibility and compatibility issues.
The good news is that enterprises can get the best of both worlds with Yellowbrick’s unified hybrid cloud solution, and move their workloads where their business requires: on-premises and/or in the cloud, instead of one or the other. Below, we highlight some ways that we have enhanced our solution.
New Key Features & Announcements from Yellowbrick
Recently, we announced business continuity enhancements — richer backup/restore and database replication functionality and the GA of Cloud DR, a low-cost disaster recovery service for Yellowbrick data warehouses — that formed the backbone of Yellowbrick Release 4.0. In Release 4.1, which is now available, we’re following up with optimizations for improved real time ingestion, better support for doing bulk loads into Yellowbrick from cloud object stores, and enhanced security, including support for column level encryption and new fine-grained access control functionality.
Together, these new features are a continuation of our commitments to helping customers get richer, more accurate insights from their data far more quickly and reliably than they can with other approaches, with enterprise-class security baked in.
- Improved real time ingestion rate for data feeds: Yellowbrick now supports even higher real-time ingest rates of up to millions of rows per second. Unlike competing cloud-only data warehouses that require micro-batching or bulk loads, Yellowbrick can ingest live, streaming data in small transactions from data sources such as Apache Kafka or Oracle Golden Gate. The latest transactions can be analyzed alongside historical data instantly, without the multi-minute lags of competing databases.
- Deeper support for cloud data migrations: Yellowbrick is able to ingest data at line rates, up to 10TB/hr or more for cloud data migrations. The database now supports bulk loads from Azure Blob containers and object stores from other S3-compatible providers, and has enhanced support for loading from AWS S3 buckets — which makes it easier to migrate data into Yellowbrick from cloud-based repositories, for example.
- Column encryption and fine-grained access control: Yellowbrick offers column level encryption support, particularly relevant for the healthcare industry and other customers managing PII. Starting in Release 4.1, encrypted columns can be declared in the schema, then automatically loaded, stored, and queried in a protected format. Only authorized users may decrypt the data in these columns. Several encryption options are provided, including deterministic and randomized OFB (Output Feedback) block-cipher algorithms.
Furthermore, a range of new GRANT and REVOKE commands are now available at different levels to provide mechanisms for creating a fine-grained security model. You can grant access privileges (ACLs) on various database objects (or on the system) to one or more users or roles. Objects can be individually secured or in the context of a schema, a database or the entire system can be secured. As time goes on, we’ll continue to invest in richer support for fine-grained access control.
More on the Way You can see the complete details about Release 4.1 in the Product Documentation here. Stay tuned as we continue to invest in new features that widen Yellowbrick’s performance lead, improve visibility into internals, bring more cloud-native features to amplify the benefits of hybrid architecture,streamline and leverage data replication and more!