Hacker News Reader

Comment by danpalmer | Hacker News Reader

> One organization I knew of had 1,500 terabytes of data, with less than 2% ever having been accessed after it was first stored.

On a related note, probably a similar percentage of people claim on their car insurance. If only the rest realised they had "crap insurance" and were paying for nothing, they could save so much money!

This is obviously sarcasm, but I think it's important to remember that much of the data is stored because we don't know what we will need later. Photos of kids? Maybe that one will be The One that we end up framing? Miscellaneous business records? Maybe those will be the ones we have to dig out for a tax audit? Web pages on government sites? Maybe there will suddenly be an interest in obscure pages on public health policy if a global pandemic happens.

Complaining that data is mostly junk is not a particularly interesting conclusion without acknowledging this. Is there wastage? Yeah sure, but accuracy on what needs storing is directly traded off with time spent figuring that out, and often it's cheaper to store the data.

View on HN

This is exactly the problem. Consider an order processing system that processes a million orders a day, and retains them for ninety days. What percentage of those 89-day-old orders are actually needed at 89 days old? It could be quite low, maybe a couple thousand out of a million.

But if those orders aren't there, shit hits the fan. PCI compliance audits fail. The ability for customers to reconcile their charges with their purchases breaks. In that 0.01% of cases where the order was fraudulent, placed by mistake, or just didn't have what the customer thought it had in it, not having that data makes the order processor read as, if not malicious, at least incompetent.

The real question is, how much data do we need to store inefficiently, in a way that uses a lot of power and space?

I was just pondering this today, in terms of how much data and objects I create in a Google or Microsoft account, for example, and then they create a burden of cost and maintenance for me down the road. Especially deleting old emails and photos. That's arduous and sometimes poignant as I flush away my personal life story.

Cloud services make difficult and sometimes even byzantine processes for deleting stuff, and it's often impossible to operate en masse in order to clean up swaths of stuff quickly and efficiently. It's in their interest to retain everything at all costs, because more used storage can mean more profits. Cloud services also profit from unused storage, because if they're charging $20/year to 100,000 users who use 2% of their storage space, ka-ching!

It irks me to this day that standard or even advanced filesystems don't include "expiration dates" or "purge dates" on file objects. Wouldn't it be logical, if an organization has a "data-rentention policy" that mandates destruction after X date, that the filesystem simply purges it automatically? Why does this always get delegated to handmade userland cron jobs? Moreover, to my knowledge, nobody is really interested in devising a way to comb through backup media in order to [selectively] destroy data that shouldn't exist anymore. Not even the read-write media!

Google is now auto-deleting stuff like OTP SMS messages. I'd love it if auto-delete could be configurable account-wide, for more than just our web histories and Maps Timeline and stuff. Unfortunately, to "delete" cloud data means it still exists on backups anyway. But without deleting data in your cloud account, it becomes a juicier hacker target as it ages and accumulates personal stuff that shouldn't fall into the wrong hands. Likewise for any business, it behooves them to delete and destroy data that shouldn't be stolen. At least move it offline so that only physical access can restore it?

I will say that modern encryption techniques can make it easy to "destroy" data, simply by destroying the encryption keys. You can quickly render entire SSDs unreadable in the firmware itself with such a command. Bonus: sometimes it's even done on purpose!

But even deleting data presents a maintenance cost. So if 90% of an org's data is indeed crap, then 90% or more of your processing resources are going to be wasted on sifting through it at some later date. Imagine when your file formats and storage devices are obsolete, and some grunt needs to retrieve some record that's 30 years old, and 90% of your data was always crap. That grunt is hopefully paid by the hour. We really had this happen at a few of my jobs, where we had old reel-to-reel backup tapes and it was difficult enough to load the data into a modern SunOS machine.

https://m.xkcd.com/1683/

I suspect for most cloud providers you deleting data is cheaper because the data is not charged by the byte. But then they like having data, maybe just to train their AI models or for bragging rights to their investors.

For the expiration dates most modern file systems have the concept of arbitrary extended attributes per file. It's quite easy to add meta data like this yourself.

Unused services are always pure profit, and storage space is no exception. Providers can offer 100GB for like $24/year, because only like 2% of the subscribers will ever approach the limit, and so the extra space is never wasted but can be allocated to someone or something else.

It's like gym memberships, ISP/telco service bundles, and amenities at your apartment complex. Anyone not using every possible service is wasting money, but it's impossible to purchase a bespoke service, so essentially everyone will waste money because they're chipping in money for services that someone else uses more than they do.

Here at home I don't ever use the gym, the racquetball courts, the doggy-doo supplies, the laundry room, or a parking space, and yet my rent (everyone's rent) includes upkeep for all of those things. I'm subsidizing all my neighbors and all the wear-and-tear damage they put on those common amenities. Likewise, everyone who's paying $24/year for storage, or any business that purchases big multi-terabyte storage media, they're paying for unused storage space and giving profit. It's practically impossible to rightsize your storage media, and you never want it undersized, and you can't simply shrink them and reclaim the resources you invested, you just keep adding on new ones and replacing the malfunctioning ones. So nearly everyone always owns or rents more space than they can realistically utilize.

Furthermore, you'll notice that I specified "automatic" destruction of data by expiration date. Of course it is trivial to tag any arbitrary file with arbitrary metadata, but the challenge is to create a filesystem that executes automatic data purges on schedule, rather than pushing it into a rickety old handmade cronjob in userland. I've never ever seen a filesystem with such a feature, nor does it seem that anyone's interested in doing so.

And here I thought that computers were useful for automating business logic and easing the burden on human effort. And this is me, manually sifting through emails and photos in order to manually delete each one with 3 dialog boxes intervening. It takes hours, days, weeks.

There’s a whole concept of records management in enterprises that manages the disposal of data. It’s far more complex than just purge dates as there’s often regulatory requirements and legal discovery issues so <2% of data is actually disposed due to perceived risk.

For personal data, the concept would be simpler but still has requirements like say tax records need to be kept 7 years.

Maybe park the data with the regulator as leaving it in your hands longer than you need it is asking for trouble.

Ah yes, when doing the manual Cron job I often say out loud to myself this is not why we made computers!

There is no clear path forwards. Perhaps regulation could be a solution but I doubt it would turn out nice.

I would love a feature like retention policies at the filesystem level. If it could somehow take data provenance into account, that would be even better, i.e. data that I've created directly is irreplaceable and should never be deleted unless I say so and should always be backed up (photographs, writing, project source code, art, game save files, my list of installed applications, etc.); data that I've created indirectly is high priority but may be deleted under certain circumstances or may be omitted from backups (browsing history, shell history, recently opened file lists, frequently used app lists); data that can be easily replaced is low priority and may be cleaned up at any time and need not be backed up at all, contingent on how replaceable it is (application files, cached data).

Not at the filesystem level, but Android does something that feels like that: an app can write files in its own space, and when you delete the app, it deletes its files.

Except when the app writes files outside of its own space, which is meant for stuff that should stay (like pictures).

Of course, some apps store the pictures in their private space, and you lose the pictures when you remove the app. And some apps write crap in the shared space. But that seems like a fundamental limitation (even if it was done at the filesystem): how do you make sure that the apps/software you use is doing it right?

See now, that's quite unrelated what I'm envisioning, or the topic of the OP's article, but that's definitely the way we're headed in terms of file handling.

"File-type strictly associated with app" is good for security and it's good for people's mental organization. Unfortunately you're referring strictly to local file storage, I assume, and practically nothing on my phone is in local storage but "in the cloud". The only local storage I use is for ringtones and certain audiobooks/prayers which I enjoy listening whether or not I'm connected to a network.

It would seem that data centers and cloud services are a long way from "file types strictly associated with app" and also the sort of behavior you describe would be really horrible and undesirable in a typical data-center setting. I mean, I can't even imagine how it'd be implemented, given that data storage is often remote or at least decoupled from the "application server", and it's definitely never a 1:1 relationship between those abstracted components.

Can you imagine needing to "delete the /bin/tar application" in order to purge every file ending in ".tar" or ".tgz"?

But it's true that the new mobile OS paradigm has advantages for cleanup. If you've got a 3rd-party app and you've created some big crap files with it, and you uninstall that app, then it's a clean sweep when it takes along all the crap and you don't need to worry about lingering loose ends, or files you simply can't open, because their handler's gone.

Scoped storage is way more secure; it's also wonderful on my Androids that I can selectively grant access to files and media, rather than giving "the keys to the kingdom" to all these Social Media apps. I definitely never want to post my bank screenshots to Facebook and the possibility no longer looms!

These days I am sort of nonplussed when a web app demands an all-access pass to my Google Drive, because permissions can be appropriately scoped now, and the app can play in its own Drive "Sandbox" without having access to my spreadsheets and medical records. I just wish that there were some really good "cleanup" apps. Remember those "cleanup" programs that would come with antivirus for your PC, or Regclean, or whatever freeware utility would purport to liberate your 10MB HDD? Those were good times.

It's the same with social media posts--I often wish to really scrub my account, delete all posts and all comments and basically all activity, and while that's possible on Facebook it's arduous and it takes me weeks to scrub the major stuff after 4-5 years. But I only think of it on Shrovetide Monday or Fat Tuesday, and then I spend 3 weeks in Lent doing scrub, scrub, scrub. Sometimes the mouse/keyboard gestures are more repetetive & physically demanding than cleaning the toilet. Way too difficult, and that's by design.

> Especially deleting old emails and photos.

Google or apple can put a big button delete everything in their phones/accounts but then some prankster will do it to a family member and this gives bad PR. Let's be pragmatic.

> It's in their interest to retain everything at all costs, because more used storage can mean more profits.

As an user, I get the reverse. When I had local NAS then I dumped anything and everything - assuming.. this costs not much. Will clean up later. Once I moved to cloud that changed to If I put crap then it will cost me money! Keep it clean

Once upon a time Google had given generous unlimited to all education and workspace. They stopped it in the last 2 years and you can see most educational and companies are running a tight ship.

Due to backup costs in our organisation we are forcing people to use max of 100GB for emails.

> advanced filesystems don't include "expiration dates" or "purge dates" on file objects. Wouldn't it be logical, if an organization

Totally agree.

Outlook email service has some kind of keep only the last newsletter from this service etc.

> auto-delete could be configurable account-wide, for more than just our web histories

- features are designed with majority in mind - most users have some sort of nostalgia for reading or keeping old emails SMS etc

Assuming your somewhat methodical about labeling email in your non-primary tabs in Gmail, you can pretty much do periodic purges while not worrying much about your primary which, in my case at least, gets relative little traffic.

Unfortunately a lot of data retention policies aren’t so mechanical, but are of the form “delete after 7 years unless there is a legal hold in force”, which is usually just rare enough of an edge case that orgs evaluate it manually and hence only do periodic manual purges. But probably the main reason auto-delete isn’t popular is because a process that can delete your old data is one bug/misconfig away from deleting your new data too.

> The real question is, how much data do we need to store inefficiently, in a way that uses a lot of power and space?

This is indeed the critical question, and it's far from being trivial.

One issue we all hit is moving the data from the higher tier storage to the cheaper and more efficient one, which requires sync and paying for the transfer most of the time, but also handling two separate access and authorization process, backup and recovery system for data that absolutely needs to be accessible for the few years of legal retention, and can/must completely disappear afterwards.

In most orgs I've seen the cost of going through all that complexity is just not worth it, compared to "just" paying for the higher tier storage for the few years long lifetime of the data.

Seems like we should be solving this at the storage system layer, instead of per application?

  probably a similar percentage of people claim on their car insurance. 
  In that 0.01% of cases where the order was fraudulent, placed by mistake, 
  how much data do we need to store inefficiently, in a way that uses a lot of power and space?

I'm feeling, the real real question, as Sci-Fi as it gets, is, is the winning ticket data even data, OR is it more like a thumbnail of "the whole data" that is 98%+ worthless, than standalone piece of data?

The winning ticket ID, e.g. a "0x-636e-7461-4e7b", only makes sense in the context as one among the entire cohort of contestants; I can make one up like I did, but I can't walkout with the payout unless the rest of the lottery didn't exist.

Statistically, philosophically, technically, and all sorts of *-cally speaking, is the 2% data, the winning ticket datum, even data?

The article touches on this issue by pointing out that:

>The Cloud is what happens when the cost of storing data is less than the cost of figuring out what to do with the crap

But I think that's wrong. The actual issue is that you often can't figure out "what to do with the crap" because the difference between useful data and crap data is determined at the point in time when you need it, not when you store it.

I'm relatively careful with deleting data, but even so, there were countless instances where I thought something was no longer needed and deleted it, only to figure out that, a month later, I needed it after all.

Also, the article has a few anecdotes like this:

>Scottish Enterprise had 753 pages on its website, with 47 pages getting 80% of visits

But that is completely orthogonal to the question of what is "crap." The fact that some data is rarely needed does not mean it's crap. If anything, the fact that it is rarely needed might make it more valuable because it might make it less likely that the same data can be found elsewhere.

I do cull photos and docs but it takes effort and you will make mistakes. On the other hand, intelligent culling makes it easier to find things.

It’s a trade off especially with digital where there’s not that much cost associated with holding onto things relative to the physical world. My whole house is basically in storage because of a kitchen fire and I’m planning to be pretty selective about what I move back in.

>intelligent culling makes it easier to find things

That's true, but things get easier to find over time. I have thousands of digital photographs from the 1990s that I burned on CDs and then copied to a NAS later. Today, for the first time, they're actually searchable because they're now indexed in Immich. So they're sorted by the person in them and searchable by their contents.

If I had culled the photos back then, I would have lost many photos that have become searchable (and thus valuable) only recently.

And geoencoding on especially phone photos makes manual metadata entry less important. I still there’s value to culling redundant and bad photos.

I have ~2.5TB of photos in iCloud, via Apple Photos. Excluding the various sized previews, I doubt many originals have been accessed in quite a while. I also have about 1TB Lightroom library archived to a different service; representing countless hours of photo processing work spanning over a decade. Haven’t touched that one in years. Neither are crap. (Yes, both have other backups and, yes, I’ve probably forgotten to sync one of them).

Crap is in the eye of the beholder...

Sometimes I wish there was a feasible system to audit and reduce redundant photos in my iCloud. I have an embarassing ton of pictures that I never look at just like everyone else. I wouldn’t mind a tool that said “here’s 10^x photos that are the most similar to all your other photos would you like to put them in the trash?”

Thing is, the tool needs you to pick the one instance that survives the purge. Which means looking through 10^x of mostly similar photos to pick one.

Removing duplicates isn't that complex already, and these tools exist so anyone can try it. It's just a truely grueling process.

Yeah, there have been a number of companies that say they'll offer AI Photo Culling, but I have yet to find one that actually does it.

Can you recommend any particular tools to do this?

If you’re using Apple Photos, the feature is there. It will detect both exact copies as well as (nearly) identical visual duplicates. Look under utilities.

Maybe it's different on the desktop, but on the iPhone it only detects the same photo that may have been saved at different resolutions or things like that. That's useful, but I really want something that can say use an embedding or something to group all 10 photos I shot of friends around the campfire and help me select one keeper and delete the other 9.

The feature works the same on iPhone. Looking through my duplicates just now, the detected duplicates that had a dozen or so photos to merge, are either high speed sports photography bursts or astrophotography—the rest were just mostly two photo duplicates. I suspect the threshold for determining a the photo as duplicate is too high for successive* photos to be considered such, as with your campfire example.

With that said, I’m surprised Apple hasn’t implemented a feature to group/bundle successive photos beneath what is determined to be “best.”

[*] Note: successive photos being separate shutter taps/actuations (potentially several seconds apart), where bursts are continuous (generally as fast as hardware allows, ms apart).

I download them 200 at a time then delete the selection.

I have about the same in iCloud and then dozens of drives of video footage. Beyond the "keep it in case you need it" aspect, another angle is that it's often cheaper to just keep everything than spend all the time it would take to do a thorough cull. Video is much slower to review, but also chews up far more space, so maybe that balances out.

If I knew the data i was going to need in the future i'd be in the future predicting business.

But there's an important piece there that is about data that should not have been stored in the first place. All the big data bullshit made people believe every single piece of data you could possibly store from a user is data that you should store and this is both a huge waste of resources and a huge liability, because now a data leak with useless to you PII could completely bankrupt your business.

"Photos of kids? Maybe that one will be The One that we end up framing? Miscellaneous business records? Maybe those will be the ones we have to dig out for a tax audit? Web pages on government sites? Maybe there will suddenly be an interest in obscure pages on public health policy if a global pandemic happens."

Perhaps I am missing something, but these examples all sound like candidates for _offline_ storage, for which no third party custodian or data center is required.

The price of large capacity NVMe SSDs continues to fall.

The amount of energy, the resource requirements, not to mention the environmental and community impact, of running a car insurance company are miniscule in comparison to that of running a data center.

Offline storage is only easy on the very small and the very large scale. On the very small scale you can get archival-grade CD/DVD/Blurays. On the very large scale you dedicate climate-controlled rooms to an automatic tape library.

In between is a vast gulf where your only good options are disks that you have to occasionally spin up to check integrity, have the disk trigger a data refresh if it's an SSD, and replace any disks that failed and rebuild their data from redundancy (raid or whatever you prefer). HDDs die even when off [1] and flash storage in SSDs is only rated to hold data for three years without power.

Sure, you can roll this yourself, but it can easily go wrong. Easier to either keep the disks spinning or pay someone with a tape archive (e.g. AWS deep glacier)

1 https://www.tomshardware.com/pc-components/storage/twenty-pe...

Are you going to be able to find the data you're looking for if judges demand it?

Data that ends up in that sea of crap is very often poorly labeled.

Data that you cannot find again is useless.

If you take more than 3 minutes to find a picture you wanted to show me, it doesn't deserve to be showed anymore.

"Better to have and not need, than need and not have."

Storage is cheap, very cheap.

For your car note, insurance is only worth it if the thing being insured can ruin you if something were to happen to it. In the case of cars, you can potentially get ruined from the value of the other party's car. But if you live somewhere where most people drive normal cars, it might be worth it not having insurance. Our culture of insure everything, from your iphone to house is a market failure. If your house were to burn down, the value of the asset is still mostly concentrated in the land.

True, when it comes to data, there is a strong case to be a compulsive hoarder. It's cheap to store; often free and data is often marginally more useful and interesting than most generic mass-produced old junk which sits in garage gathering dust.

Technology for sorting through unordered piles of data also gets better over time. When I started storing digital photographs, face detection wasn't a thing, but it is now, so those tens of thousands of photos from the 90s suddenly became more useful to me.

> …I think it's important to remember that much of the data is stored because we don't know what we will need later.

This is the basic argument that every hoarder uses to justify their hoarding.

I think you miss the point of the article, which precisely says that we store a ton of data we don't need to store in the first place.

Photos of kids are obviously not part of that "crap" (even if we have too many of them and it would be worth triaging for our own sake).

The question is: what makes for most of that data? Is that all business records, or is that storing stuff because we can? I've worked in multiple startups, all were tracking users as much as they could. Adding tools that collect data is easy, and storing that data is cheap. "We may need it later". Never needed any of that crap, and it was invading our users' privacy.

In any case, I think the article raises a good point: it's so cheap to store crap that we don't even think about it. And it's bad for the environment. Just like it's so cheap to take a plane that we don't think about it, even when taking the train takes almost the same time.

Photos of kids? Maybe that one will be The One that we end up framing?

Look through the pictures with your kids (and with your kids) and pick out the best ones. Delete the rest.

The photo apps really don't have a good "sorting through the shoe box" UX. It would be great if I could in Google Photos, for example:

1. Bulk favorite photos in Google Photos for long term archiving

2. Set a retention policy of months/years for photos based on metadata.

3. Have a UX to quickly sort the week or months photos via a swipe left/right UX.

I go back and look at old photos all the time (like when I think of a trip, and look back to it, and then just keep scrolling for a while). The pictures bring back more memories and that's to me a great joy - and on the other hand I find no value in deleting the photos that are not "the best" (I'm talking about good enough photos, I delete blurry, useless or duplicate ones immediately) - what would it gain? A few dollars less storage cost over the next decade? The satisfaction of digital cleaning?

The problem with that is that what you might find important or beautiful changes with time.

That random photo of your messy living room might be trash right now, but beloved in 20 years when you want to remember your old home.

Reply on news.ycombinator.com