Missed out on the previous parts of our Straight Talk series? Check out the introduction to data backup and deduplication, backup to tape, disk staging, enter data deduplication and sizing a disk backup system.
"A well educated buyer will make good decisions. The more you know and the more questions you ask, the better equipped you will be to choose the right solution for the long run." Bill Andrews, president and CEO, ExaGrid Systems
In the previous sections, we demonstrated why you need a strategic approach for backup. In this section, we provide a checklist of what to consider and questions to ask when selecting a vendor to ensure you choose the right solution for both the short and long term.
A number of questions should be asked before you acquire a disk-based backup system to avoid the following possible undesired outcomes:
- Certain features in your backup application may not be supported
- You may not have disk backup support for other utilities in your environment
- Disk usage may rise quickly with each backup due to a poor deduplication ratio, and within 6-18 months you could need additional capacity. The poor deduplication ratio can be a result of the actual algorithm, certain data types that do not deduplicate, or both
- Purchase of additional WAN bandwidth may be required between the primary site and DR site because of a poor deduplication ratio. The poor deduplication may be due to the actual algorithm, certain data types that do not deduplicate, or both
- You may be faced with a forklift upgrade where the front-end controller needs to be upgraded if the system does not add processing resources with disk as data grows
- Your restores could be slow
- Your tape copies could be slow and could take several days to a week to complete
- Time to recover from a disaster could be excessive
- Vendor technical support delays in troubleshooting or resolving problems may result in downtime
Here are some of the questions you need to ask vendors about their disk backup solution and support:
Backup application support
- Does the solution support your backup application?
- If you are thinking of changing backup applications in the future, are those other backup applications also supported?
- For databases, does the solution support direct to disk for Oracle RMAN dumps or SQL dumps?
- For environments with some UNIX, does the solution support direct UNIX TAR files?
- If it is a Symantec Backup Exec or NetBackup environment, which levels of OST are supported?
- If it is a Symantec NetBackup environment, are advanced features, including AIR and optimised synthetics, supported?
- If it is an IBM Tivoli Storage Manager environment, are progressive incrementals supported? How is reclamation performed?
- If the environment includes a backup application that performs deduplication (e.g. Veeam Backup and Replication), does the disk-based backup product work with that deduplication and further deduplicate the data?
- Can you bring in data from multiple backup applications and utilities? If so, which ones?
- If the vendor also sells their own backup application, does their disk backup solution offer robust, complete capabilities to support other companies' backup applications?
Deduplication approach/architecture, impact on backup and recovery operations
- What method of deduplication does the solution use, since disk usage and bandwidth needs will vary greatly? (Zone with byte-level; 8KB block with variable length splitting; 4KB fixed block; 64KB or 128KB fixed block)
- How long does it take for a traditional full backup or image backup to restore?
- Does the solution support instant recovery techniques, such as Instant VM Recovery?
- How long does it take to make a full backup offsite tape copy?
- As data grows, does the system scale by adding processor, memory and bandwidth resources along with more disk capacity? If not, then as data grows, how quickly will the backup window re-expand?
- Is it possible to speak with several customer references to verify the restore, recovery, tape copy, and scalability capabilities above?
- How is RPO (recovery point objective) achieved at the disaster recovery site?
- How is RTO (recovery time objective) achieved at the disaster recovery site?
System sizing and pricing
- When sizing, did the vendor ask you the following for sizing the system? Weekend full backup amount; weekly backup rotation (incrementals, rolling fulls, differentials, etc.); annual data growth rate
- How much of your data is comprised of each of the following? Compressed; encrypted; unstructured; database
- What retention in weeks, months and years is required onsite and offsite?
- Are any large data amounts expected to be added in the short term due to factors such as bringing on new customers, merging businesses, or other changes?
- If more than one location is to be backed up, how much data is to be backed up at each location so additional system sizing can be taken into consideration for multi-site cross protection for disaster recovery?
- Are you turning on deduplication with the backup application such that the overall deduplication rate will be impacted? If so, what does that mean to disk and bandwidth usage?
- At your current data growth rate, when will you have to buy additional capacity?
- If the solution is a controller/disk-shelf architecture (scale-up), at your current data growth rate, when will you need to upgrade the front-end controller via forklift upgrade? What will that upgrade cost?
- Is the vendor giving you a low price up front to gain your business, knowing it will keep selling you more and raise the price later? Or is the vendor willing to put in writing that the price you pay up front is the same that you will pay moving forward
Management/reporting and technical support
- What level of user- and management-level reporting is provided?
- Can the system be easily installed with phone support or does it require an installation team and associated costs?
- Is support handled by senior-level second-line staff to quickly resolve problems, or is support handled by junior-level first-line staff unable to quickly resolve problems?
- Is technical support staff skilled in specific backup applications and assigned based on expertise with your backup application and environment?
- Is technical support based in the country where the system is deployed and managed to avoid poor quality connections and language barriers?
- Does the technical support and disk backup system include automated health-check emails from the system to the vendor to proactively alert about capacity issues, hardware diagnostics, or other problems?
The above list is not meant to be all-inclusive, but rather to demonstrate the complexity of disk-based backup and why it is important to not view it as merely a NAS target, disk, and deduplication. A disk backup system requires software to make the solution work properly in all existing backup environments, allow for fast backups, restores and cost-effective scalability, as well as operate reliably in a real-world production environment, and be properly supported.
The following is highly recommended when acquiring a disk-based backup product:
- Ask all of the important questions
- Receive an in-depth product presentation
- Ask for actual customer user stories with the companies and IT staff named
- Ask to speak with 3-5 references
- Do a side-by-side test if you have the staff, lab and time
- Use a large amount of data to stress the systems, as all products look good with a small backup job
- Ask about the support model and ask the provided customer references how satisfied they are with the vendor's support
- Get a future price guarantee in writing to ensure you are not getting a low-ball price up front only to have the price increased later
This guide explains the various backup complexities, enabling you to ask the right questions and make the right decision for your specific environment and requirements. Stay tuned for the next part of this guide, which will be live on ITProPortal shortly.