Database Consistency and Database Hot Backups

Dell PowerVault MD3600f/MD3620f Remote Replication Functional Guide

Database Consistency and Database Hot Backups

After suspending AWOC for database virtual disk replication, and, if a disaster occurs while resuming

RRthen the remote database virtual disks are not recoverable. This is due to the delta log processing of missing I/O in non-original write order. Once RR’s resume operations are complete, the virtual disk pairs are once again in synchronization, and the remote database virtual disks are recoverable in disaster recovery processing.

In order to have usable database virtual disks in the event that a disaster occurs in the middle of a resume operation, it is important to snapshot the remote database virtual disks at the very instant that they are suspended and before the replication is resumed and uses the delta log to re-synchronize. This requires the auto-resume feature to be turned off and a snapshot is taken at the time of suspension.

Once the resume operation completes, a second snapshot is taken. This provides the ability to recover from a primary site disaster. If a disaster occurs during resume operations, the first snapshot is used for recovery. Once the resume operation completes, the second snapshot is used for recovery.

In both cases, either snapshot will be database consistent and usable for database recovery in the event of a disaster. We can achieve this using Database Hot Backups.

All databases provide a hot backup operation that allows a storage array to snapshot the database virtual disks to create a point-in-time copy for backup and/or recovery purposes.

Database hot backups perform a number of functions that ensure recoverability using snapshots or copies of the database virtual disks. First, the database internal data (pointers, buffers, tablespace data, and recovery information) are all persisted to the database virtual disks. Enough information is therefore stored to ensure a successful recovery using the database virtual disks including the important logs.

Disaster Recovery using RR will use this same hot backup operation to create a remote snapshot of the database virtual disks. In essence, the remote snapshot is identical to the local snapshot.

During this hot backup window we will use RR’s suspend feature to flush the data written on the primary site virtual disks to their replication counter parts on the remote system; we can then take a snapshot of the remote virtual disks to create a recoverable set of database virtual disks. RR will then resume and the database will exit hot backup mode.

NOTE: Periodically performed, this process will create recovery snapshots (also called Recovery Points) which can be used as remote backup images, or to reprovision the database itself (a second copy residing at the remote site), and in the event of a disaster and the current set of virtual disks are not recoverable, the last recovery snapshot will be used instead.

Page 11

MD3620F, MD3600f specifications

Dell's MD3620F and MD3600F storage arrays are designed for businesses seeking dependable, high-performance storage solutions tailored to meet the demands of modern data management. These models offer a blend of scalability, efficiency, and advanced features, making them well-suited for a range of applications, from virtual environments to critical database operations.

One of the primary features of the MD3620F and MD3600F is their support for a hybrid storage architecture. This approach allows organizations to combine flash storage with traditional hard disk drives (HDDs) to optimize performance and cost. The MD3620F focuses heavily on all-flash configurations, delivering exceptional speed and lower latency, ideal for I/O-intensive workloads.

Scalability is another hallmark of these storage solutions. The MD3600F allows users to scale up to 120 drives, providing robust capacity to meet growing data needs. The MD3620F is engineered for environments where data availability and speed are paramount, with support for high IOPS (Input/Output Operations Per Second) and consistent performance under stress.

In terms of data management, both systems integrate with Dell’s proprietary software, which enhances operational efficiency through features like automated tiering. This technology optimizes data placement between different storage types based on access frequency, ensuring that hot data is readily available while reducing costs associated with less critical data.

Reliability is crucial for businesses that depend on their storage systems. The MD3620F and MD3600F come equipped with advanced data protection mechanisms, such as RAID support, which safeguards against data loss, and snapshot capabilities that provide point-in-time copies of data. Moreover, these solutions include hot-swappable components, ensuring minimal downtime for maintenance and repairs.

Connectivity is also a strong suit of these storage arrays. They support multiple connectivity options, including SAS and iSCSI, allowing for flexible integration into existing IT infrastructures. This adaptability enhances their appeal to a wide range of enterprise environments.

In summary, the Dell MD3620F and MD3600F stand out as robust storage solutions with a combination of hybrid and all-flash configurations, scalability, advanced data management tools, and enhanced reliability features. These characteristics make them ideal for organizations looking to enhance their data storage capabilities while maintaining high performance and efficiency.