• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Data lifespan

bigboxes

Lifer
Twice I've asked Anand this question in a comment in response of a SSD review, but have gotten no reply. What I want to know is information about data lifespan. I've heard that data can safely reside on traditional HDDs for about 10 years. I figure that eventually, as prices fall and capacities increase, we will see a move to SSD for data storage. Whether it is a file server or backup drives that are stored unused I want to know the lifespan of data on these drives. I'm not sure if it makes a difference whether a drive is being used or is sitting on a shelf in storage. What I'd like to know is the data lifespan from the moment it is written. So, I'd like the following information if anyone knows:

HDD Data lifespan
Drives in use:
Drives in storage:

SSD Data lifespan
Drives in use:
Drives in storage:
 
I think it will be hard to quantify in a length of time.

three main factors will be:
1 errors before writing (bad data is written to storage)
2 errors in stored data (physical or otherwise, may or may not be fixed by the file system)
3 device failures

you will need to look at the chances of all of these, how long you plan to store the data. how much data, what file system you are using (for example ZFS is better with some errors) and then at last you will need to look at this and figure what % chance of losing data you will find acceptable

for some information you can read here:
ZDNet: Data corruption is worse than you know

The bottom line
CERN found an overall byte error rate of 3 * 10^7, a rate considerably higher than numbers like 10^14 or 10^12 spec’d for components would suggest. This isn’t sinister.

It’s the BER of each link in the chain from CPU to disk and back again plus the fact that for some traffic, such as transferring a byte from the network to a disk, requires 6 memory r/w operations. That really pumps up the data volume and with it the likelihood of encountering an error.
and here is information about hard drive ecc correction error rates:
http://www.storagereview.com/guide/errorRead.html

but you will need to find a source more specific to your questions as per usage (stored drive vs in use) etc
 
Last edited:
I think the OP is asking about data retention time, unpowered and powered, not data corruption.

For powered data retention time, it will depend on the specific SSD and firmware, since the details of the wear-leveling and any static-data shuffling that it does will be relevant.

For unpowered data retention time, you only need to know the spec for the underlying flash chips. The industry standard is documented in JEDEC JESD47F, "Data Retention for Non-Volatile Memory". Details on the tests that are to be done to verify the data-retention time are in JESD22-A117B "EEPROM program/erase endurance and data retention test".

To simplify and summarize, the flash memory makers will qualify their parts by performing program/erase cycles and high-temperature accelerated aging to simulate years of use in a matter of days.

JEDEC JESD47F specifies that the parts should have at least a 10 year data retention when new, and at least 1 year after they have reached maximum program/erase cycle count. Between 10% and 100% of maximum program/erase cycle count, the minimum data-retention time should be estimated as:

1yr / (fraction of cycles used)

So, if you have an SSD with 80GiB of flash with a max program/erase cycle count of 5000, and you estimate that the write amplification is 2, then after you have written 40,000GiB (about 43TB) you will have used up 1000 cycles. Therefore the minimum data-retention time would be 5 years = 1yr / (1000/5000)
 
Last edited:
I think the OP is asking about data retention time, unpowered and powered, not data corruption.

You are correct. As we move more and more into the digital age we need a cost-effective way to store our data. I don't want to find out ten years down the road that I no longer have my pictures, documents, videos and music. We need to know how often we need to replace our storage drives in order to maintain data integrity. I appreciate all the replies.
 
You are correct. As we move more and more into the digital age we need a cost-effective way to store our data. I don't want to find out ten years down the road that I no longer have my pictures, documents, videos and music. We need to know how often we need to replace our storage drives in order to maintain data integrity. I appreciate all the replies.

Only active (e.g. online, cloud-based, whatever) storage can really solve that problem. Redundant, active storage that is constantly rotated between different file formats as appropriate.

There is really no way to store digital data for an appreciably long time in an offline state. This isn't just because of media degradation -- it has to do with being able to have software to actually interpret the data. Anything more substantial than a text file is victim to this. Will JPGs, DOCs, and MP3s be relevant or even readable (without specialized software) in 20, 30 years? My best prediction is no.
 
Anything more substantial than a text file is victim to this. Will JPGs, DOCs, and MP3s be relevant or even readable (without specialized software) in 20, 30 years? My best prediction is no.
Not really, JPGs are a documented standard the same goes for MP3s and PDFs. So those are obviously in the clear, but even proprietary formats like docs will be easily readable with a VM and the adequate old software - the chances for every version of MS Word vanishing in the next 20 years are pretty small 😉 Although it'd be obviously a good idea to store the necessary software with the data for comfort (at worst you've got to create a new image if your favorite format gets discontinued and you've got to transition to the newer standard)

One thing that would be more of a problem are old hardware standards, which aren't impossible to get but a lot more pricey than the software (see old storage tapes - well and even floppy drives by now)
 
Last edited:
Back
Top