News

4/4/2011
02:37 PM
George Crump
George Crump
Commentary
50%
50%

The Passive Primary Storage Tier

Primary storage, especially the storage that supports file based data, is challenging to deal with. In most environments file based data, data created by office productivity and rich media applications can number in the double digit TB's, can bring backup applications to a crawl and cause countless retention issues. The challenge is that 80% or more of this data had an active life span of a month but can't be deleted "just in case". What is needed is to create a passive tier of primary stora

Primary storage, especially the storage that supports file based data, is challenging to deal with. In most environments file based data, data created by office productivity and rich media applications can number in the double digit TB's, can bring backup applications to a crawl and cause countless retention issues. The challenge is that 80% or more of this data had an active life span of a month but can't be deleted "just in case". What is needed is to create a passive tier of primary storage.The ability to automatically move data between tiers of storage is now built into file systems and storage controllers. Moving data to a slower less expensive tier is nothing new. The problem is that even these cheaper, more power efficient tiers of storage are not deserving of data that basically was useful for a month and is only being held on to "just in case" or to meet some legal requirement. This data needs to be identified and moved out of the active primary storage environment so that backup systems and IT budgets have to account for them. The best location for this passive primary tier of storage may actually be tape...Wait, let me explain.

Remember this is data that was active for a short period of time but needs to be retained for one reason or another. Chances are extremely high that most of it will never be accessed again but chances are that some, maybe 1% of it might be, so you just can't delete it and you will need to be able to find some of it. That means that the data does not have to be made available instantly, the moment it is decided that it is needed again, but it does need to be instantly searchable and the retrieval needs to occur within a few minutes.

Modern tape is ideal for this use case. Tape has improved in its reliability, its performance and increased its lead in cost for capacity. Most tape libraries can select a tape cartridge, load it and position it to the right location in less than a minute. With tape we can get the data back and as we indicated in a recent briefing note, tape library vendors are adding off-line verification capabilities. The only missing component then is finding that data. The simplest solution for that is to never lose it or in the case of a file system make it look like it is still online.

Several file systems exist today that can combine disk and tape into a single entity that the users interact with as if it were any other file system. As files are first created and saved they are first stored on high speed storage, then as they age they are migrated off to slower storage, maybe a MAID device and then to a tape tier. With the performance of tape being what it is now, it could be possible to eliminate that middle tier. It depends on your comfort level.

You may be wondering if I just snuck an HSM (hierarchical storage management) or ILM (Integrated lifecycle management) design in on you here. To some extent I did. The difference is the level of integration that these file systems can perform and how seamless the movement back and forth between disk and tape is. The other difference is how early in the process you can move to tape now. In the past tape was the place to put data after it was really, really old. Now given its performance you may want to consider it the moment it becomes inactive.

The tape market is so vibrant that it now has its own event, "The Tape Summit", in which tape leaders will gather to further detail these architectures to advance tape's role in the enterprise. The goal is important not only to the tape suppliers but also to you the users. If the negatives that surround tape can be overcome and the clear economic advantages of tape be widened your capacity acquisition costs will go down and power efficiency will go up.

Track us on Twitter: http://twitter.com/storageswiss

Subscribe to our RSS feed.

George Crump is lead analyst of Storage Switzerland, an IT analyst firm focused on the storage and virtualization segments. Find Storage Switzerland's disclosure statement here.

Comment  | 
Print  | 
More Insights
Comments
Newest First  |  Oldest First  |  Threaded View
'Hidden Tunnels' Help Hackers Launch Financial Services Attacks
Kelly Sheridan, Staff Editor, Dark Reading,  6/20/2018
Inside a SamSam Ransomware Attack
Ajit Sancheti, CEO and Co-Founder, Preempt,  6/20/2018
Tesla Employee Steals, Sabotages Company Data
Jai Vijayan, Freelance writer,  6/19/2018
Register for Dark Reading Newsletters
White Papers
Video
Cartoon
Current Issue
Flash Poll
Twitter Feed
Dark Reading - Bug Report
Bug Report
Enterprise Vulnerabilities
From DHS/US-CERT's National Vulnerability Database
CVE-2018-12716
PUBLISHED: 2018-06-25
The API service on Google Home and Chromecast devices before mid-July 2018 does not prevent DNS rebinding attacks from reading the scan_results JSON data, which allows remote attackers to determine the physical location of most web browsers by leveraging the presence of one of these devices on its l...
CVE-2018-12705
PUBLISHED: 2018-06-24
DIGISOL DG-BR4000NG devices have XSS via the SSID (it is validated only on the client side).
CVE-2018-12706
PUBLISHED: 2018-06-24
DIGISOL DG-BR4000NG devices have a Buffer Overflow via a long Authorization HTTP header.
CVE-2018-12714
PUBLISHED: 2018-06-24
An issue was discovered in the Linux kernel through 4.17.2. The filter parsing in kernel/trace/trace_events_filter.c could be called with no filter, which is an N=0 case when it expected at least one line to have been read, thus making the N-1 index invalid. This allows attackers to cause a denial o...
CVE-2018-12713
PUBLISHED: 2018-06-24
GIMP through 2.10.2 makes g_get_tmp_dir calls to establish temporary filenames, which may result in a filename that already exists, as demonstrated by the gimp_write_and_read_file function in app/tests/test-xcf.c. This might be leveraged by attackers to overwrite files or read file content that was ...