Home > Storage > PowerMax and VMAX > Data Protection > Dell EMC PowerMax and VMAX All Flash: TimeFinder SnapVX Local Replication > Dell EMC ProtectPoint
Note: Dell EMC ProtectPoint has been renamed to Storage Direct and is included in PowerProtect, Data Protection Suite for Apps, and Data Protection Suite Enterprise Edition software.
Dell EMC ProtectPoint technology is a data protection solution that integrates primary storage, PowerMax, VMAX All Flash and XtremIO, with Data Domain protection storage, to accelerate backup and recovery, help organizations meet application protection SLAs, and minimize backup impact on applications, all while reducing cost and complexity. Using ProtectPoint, every backup is a full backup with the storage and bandwidth efficiency of an incremental backup, which simplifies recovery.
Configuration requirements and best practices:
The information in this section focuses on PowerMax and VMAX All Flash aspects of ProtectPoint implementations. See the following documents for more information about ProtectPoint and Data Domain requirements:
The minimum configuration requirements for a ProtectPoint implementation are as follows:
The following configuration options will increase performance and availability:
Copy throughput:
Copy throughput will be higher when there is a significant change rate and many copy streams available. This applies to backup and rollback processes.
Backups consisting of a minimum of 30 production devices will make optimal use of system resources and maximize the SnapVX throughput internal to the PowerMax or VMAX. Backing up fewer than 30 devices is supported but may not necessarily take less time, regardless of device capacity, because the process will not use as many resources. Backing up only a few devices at a time, as is often seen in proof-of-concept testing, will not maximize use of resources and therefore is not a true assessment of what may be seen in production.
The time for a backup to complete is affected by subscribed capacity of the TDEVs involved, not only allocated capacity. This is because SnapVX cycles through the entire subscribed size of the TDEV to determine which data needs to copy, and this is what is viewed from the SnapVX outputs that display the copy progress. Therefore, backup of a TDEV that is not heavily overprovisioned will likely take less time to complete than backup of a TDEV of the same allocated capacity but with a larger subscribed capacity.
Throughput of an incremental backup may seem less than that of a full backup, although the duration will be shorter. For instance, a differential copy of 5% changed data (in relation to the subscribed capacity) may typically reflect ~40% the throughput of a full copy but would complete nearly 80% faster.
Because SnapVX cycles through the entire provisioned size of the TDEV, the copy rate as viewed from the SnapVX point-of-view may not reflect the copy rate that is seen between the PowerMax or VMAX system and the Data Domain. Only actual data will be transferred from the PowerMax / VMAX to the Data Domain. Overall system activity influences the ProtectPoint backup and restore processes. Likewise, ProtectPoint backup and restore operations have an influence on the system. If the ProtectPoint backup needs to be slowed down beyond the level determined by default scheduling algorithms, the QoS settings of the encapsulated Data Domain backup devices (vDisks) can be modified. This will extend the ProtectPoint backup time while lessening the effect on overall system performance.
Multiple Data Domains can share DX ports. The DX resources will be shared if backups are run simultaneously.
When possible, schedule backups when the volumes being backed up are under light or moderate write loads. Snapshot overhead is most significant when the snapshot is newly established and nearly all writes need special handling. If the backup window must coincide with a busy time for the system, additional CPU resources may be required to support the increased load and should be considered when the system is being sized.