A couple weeks ago I worked with our Big Data team to put together an Archiving and Tiered Storage webinar. One suggestion for the webinar title was “So Much Storage, So Little Time” because of all the technologies people need to think about when piecing together a comprehensive data storage strategy that often includes primary storage, backup storage and long-term archive storage.
Indeed, there are many exciting technologies to consider ranging from solid state, LTFS, Object Storage, intelligent tape vaulting, cloud-based backup and more. As technologies continue to develop, blurring the distinction between traditional use cases like backup and archive, it can be difficult to get clarity on the best strategy and the best technologies to meet your near-term andlong-term data retention and archiving requirements.
Rather than fueling the confusion profusion, we decided to offer some straightforward guidance by titling the webinar “4 Key Considerations for Archiving.” Here are a couple tidbits from the webinar that are good reminders for anyone managing data growth and long-term storage.
Clarify Your Data Requirements
- Data Type and Value: Is your data structured or unstructured, and what is the business value of your data long-term? Could you be mining that data in the future for business reasons?
- Access Requirements: How fast do you need access to the data? Seconds? Minutes? Hours? Days?
- Migration Strategy: Do you migrate solely based on age, or use a more stringent system based on multiple policies?
In other words, if you define archive as “data that doesn’t change,” you need to be clear on the performance requirements associated with this data throughout its life and how it will migrate from one phase of life to the next.
Understand the Tradeoffs of Different Technologies
- Technology Choices: Where does SSD high speed disk, secondary disk, object storage, near line tape and vaulted tape fit into your strategy?
Once you understand the tradeoffs associated with each type of technology, architecting your archive environment becomes pretty easy. Make sure you have the best tiers you can afford for the various stages of data life, and know how that data will move and migrate as it needs to. I find the following diagram useful to keep it all together in my mind:
View the Webinar
Check out the full webinar which goes into more depth on the above topics and includes relevant customer case studies.