Depressing Profit margins after calculating milage deduction - Page 2

I recommend against using more than 4 hard drives in a single computer, and recommend getting more computers as you scale. Here's why:

Bad drives during a Secure Erase will wreak havoc on your machines. Right now, I'm testing a pallet of hard drives for a poor guy who must have had something happen during shipping. The failures are slowing me to an absolute crawl. If you load 4 good hard drives into a wiper, you just tell it to wipe, and come back later. No big deal. When you have drives that have problems, tons of weird things can happen. You need to run a health check before you start a Secure Erase- if one of the drives dies during the erase, your computer may just sit there a LONG time before the erase finishes on the other drives. And, don't ever think of unplugging it! If you do that during a Secure Erase, you've just made bricks. The problem is that you can't always find faults in drives during the "Pre-Erase Testing". For example, in this pallet, I've had a few cases where one of my tester computers has been stuck for 10-12 hours just "Spinning its wheels". If your computers have more drives connected, you make a lock-up more likely, and you increase the financial risk if the power goes out on one of your machines.

Being in scrapping, you'll get computers that can handle wiping just fine- no worries if you scrapped some. My wipers are all HP towers with Pentium CPUs in them. Hardly something that would be worth much in resale, but for hard drive wiping, they handle it flawlessly! Well, as long as you feed them good drives that is

It's probably not a bad idea to run a SMART test before wiping. That's a pretty quick process for screening out a drive that is obviously bad. I sometimes start a surface test before wiping too. If you're going to have bad sectors ... they seem to show up more at the beginning of the scan for some reason.

Date of manufacture and drive capacity is another reason to send them directly to scrap. I guess it's the whole "bigger is better" mentality but customers seem to prefer larger drives. If the bigger drives pay better and sell more easily .... it makes the most sense to focus your best efforts there and cull the rest.

It's been my experience that as many as nine out of ten laptop drives coming into the shop have a problem of some kind. I think it's because the laptops tend to get dropped and bumped around. Minor head crashes into the platter aren't always fatal but they put bad sectors on the drive. Similar thing going on when the laptops don't power down correctly because the battery ran out. (Just my theory on why so many fail.)

3.5" HDD's out of skyboxes and other video equipment seem to have a high failure rate too.

JMO .... but it will get easier when solid state drives become more commonplace in the waste stream.


Afterthought:

You probably don't need the latest windows operating system for a drive wiping machine as long as it's isolated from the internet. Win 7 is a nice sturdy workhorse. It is easier to find apps that are made for pc than Mac or Linux. For whatever it's worth ... i was using XP on an 18 year old machine as a drive tester / wiper up until about a year ago.

I run Linux on mine, but if you're looking at just reselling hard drives to MTSV buyers, running a health check through something like CrystalDisk will usually weed out most of the failures. It won't be perfect, but that's not a big deal. When I get in hard drives for testing, I expect a few to fail. That's not a big deal. It's lots that have tons of problems that cause the trouble. For example, out of that pallet I mentioned earlier, the last box I tested was for 50 desktop drives. 31 of them failed. That's where you have problems. A lot of those 31 I was able to catch before wiping, but the ones that looked fine caused a lot of problems. I have six testing computers, and almost every day one of them gets locked up for a few hours. I'd much rather have a little 4-spot HP tied up in waiting than a 16 bay server. (Though I will admit, the allure of going to a server is tempting for those good lots- I could breeze through them a lot faster)

With SSDs, as long as the connector didn't get damaged in shipping, you should be good! The pallet I've mentioned in two posts already had some SSDs in it. I haven't tested them all out yet, but so far, NONE of them have failed. SSDs are super reliable.

For software, this is what I use: https://partedmagic.com - it used to be free software ages ago, but now it's about $10. It's well worth it IMO- it's a full Linux OS that can do pretty much whatever you need. The only thing I don't like about it is how it handles bad drives. I've never had a wiper eat drives from a lock-up, but I don't like the lock-ups that I've had on this last pallet. I know it's because I'm feeding the machines problem drives, but I still want something that just plain works no matter what.

Hard Drive Sentinel looks a lot like CrystalDiskInfo , but I've never used it in all fairness. Parted Magic can run health checks too, so I do it that way to make for "One-Stop drive wiping". Until this last lot, I really loved it. I still think it's a good software choice, and honestly, I don't think you'll have too many problems with failures that really gum up the works. When I test HDDs that came in from my E-Waste collection bins, I rarely have anything like that go wrong. Something happened to this pallet in shipping. Funny enough, I picked it up in Billings, and got a call the next day from Reddaway saying that they had a box that they thought came off the pallet. Sure enough, it did.

My gut says that I'm dealing with a mess that you won't have to. Other than this, Parted Magic hasn't missed a beat in almost 6 years of my use with it

I haven't used Easus, but I don't clone much. I deploy systems manually now, but I'm going to be setting up a PXE server and a cloning server this spring when I can get another room in the office building I'm in to house the darn things.

Linux works pretty simply, but Parted Magic isn't a full operating system in the same way that Windows is. Parted is a standalone OS, but all it's really good for is working with hard drives and doing some other minor testing. You wouldn't install it on a PC you wanted to just use like a desktop- you'd want something like Ubuntu for that. But, for hard drives.... whatever you want to do, it can do them.

I have a hunch that you guys wipe differently than I do. What a lot of people do for data erasure is something like dd, shred, or DBAN. Basically, those write 0's over all the data, making it a blank drive. I don't use this method, I use what's called "Secure Erase" (You can read about it here: https://www.lifewire.com/what-is-secure-erase-2626004 ). But, basically, Secure Erase is a command that works from the drive's firmware. It's a lot faster than using a 3 or 7 pass wipe (I don't believe one pass of zeros is good enough), and it's more secure, since the drive itself is doing the wiping, and not the software. You're basically telling the hard drive to blank itself, instead of telling software to do it.

The downside to Secure Erase is that it CANNOT be interrupted. If you have a power outage during a DBAN-style zero wipe, nothing bad happens. You just restart it later. If a power failure happens during a Secure Erase, your hard drive becomes a paperweight. This is where the problem of freezing comes in. The computer is totally responsive- I can move the mouse, open programs, and so-on. But, it doesn't matter since I have to wait for the program to finish it's wiping. When you get a problem drive in the bunch, oftentimes it'll wipe, but your system will "hang" with the dialogue window showing four bars that say "Finished". So, the drives SHOULD be fully wiped and done. But, I don't dare restart the PC. Instead, I just have to leave the PC until it loads the next dialogue. I've never had one that didn't load the next dialogue, but still- I don't like having a PC tied up for hours waiting. It's rare that I have this happen.... except for this last lot.

I've had a really high failure rate on that pallet. The packing of the drives I think is partially to blame, but the client says they've shipped them like that before with no issues (I know, isn't it fun when you get the first problem lot?), so my next thought is the trucking company. Considering that Reddaway had one of the boxes come off the pallet somehow, this really makes me think that they screwed up handling the pallet itself. A lot of the drives that failed are mechanical failures, not sector-based failures. Most drives from this lot that failed will make a buzzing noise instead of starting the motor, or they'll have the click of death, or they don't turn on at all. This really makes me think that something happened in shipping. All I can do is just slowly make my way through the lot and pray that the rest of the boxes do better.

I had a theory about RAM too. Parted shows you a RAM and CPU usage bar on the side of the screen. They say that you can run PM on 1GB RAM, but I run 2GB in my testers (I don't like pushing the limits). With 2GB installed, after taking some off to run graphics (I think 64MB), during a wipe, the OS uses 15% of the RAM. So, we should be good there. But, for curiosity's sake, I upped all the computers to 8GB of RAM. They booted a bit faster (To be expected), but they didn't wipe the drives any faster. It makes sense- the command to wipe is issued by HDD firmware, so you work at the speed of the HDD, not the speed of the software.

This spring, I need to redo a LOT of my facilities- I've grown a lot faster than I ever imagined, and I'm running into growing pains. It's taking me way too long to process hard drive lots, so I'm thinking about adding a server or two like JJinLV mentioned above. For trouble lots like this, they won't work at all. But, for good lots, it shouldn't be a problem. A little tip though: If you do Secure Erase, your PCs must be super stable- crashes kill drives. So, while checking machines, I use 80GB desktop drives. If your machine does fail somehow and eats them, who cares? They're scrap anyways....

Between adding PXE and cloning for deployment of laptops and desktops (Yes, Parted Magic can clone drives, but it's kind of clunky at doing it) and upgrading hard drive testing, it's going to be a heck of a Spring for me....

Originally Posted by matador

I haven't used Easus, but I don't clone much. I deploy systems manually now, but I'm going to be setting up a PXE server and a cloning server this spring when I can get another room in the office building I'm in to house the darn things.

Linux works pretty simply, but Parted Magic isn't a full operating system in the same way that Windows is. Parted is a standalone OS, but all it's really good for is working with hard drives and doing some other minor testing. You wouldn't install it on a PC you wanted to just use like a desktop- you'd want something like Ubuntu for that. But, for hard drives.... whatever you want to do, it can do them.

I have a hunch that you guys wipe differently than I do. What a lot of people do for data erasure is something like dd, shred, or DBAN. Basically, those write 0's over all the data, making it a blank drive. I don't use this method, I use what's called "Secure Erase" (You can read about it here: https://www.lifewire.com/what-is-secure-erase-2626004 ). But, basically, Secure Erase is a command that works from the drive's firmware. It's a lot faster than using a 3 or 7 pass wipe (I don't believe one pass of zeros is good enough), and it's more secure, since the drive itself is doing the wiping, and not the software. You're basically telling the hard drive to blank itself, instead of telling software to do it.

The downside to Secure Erase is that it CANNOT be interrupted. If you have a power outage during a DBAN-style zero wipe, nothing bad happens. You just restart it later. If a power failure happens during a Secure Erase, your hard drive becomes a paperweight. This is where the problem of freezing comes in. The computer is totally responsive- I can move the mouse, open programs, and so-on. But, it doesn't matter since I have to wait for the program to finish it's wiping. When you get a problem drive in the bunch, oftentimes it'll wipe, but your system will "hang" with the dialogue window showing four bars that say "Finished". So, the drives SHOULD be fully wiped and done. But, I don't dare restart the PC. Instead, I just have to leave the PC until it loads the next dialogue. I've never had one that didn't load the next dialogue, but still- I don't like having a PC tied up for hours waiting. It's rare that I have this happen.... except for this last lot.

I've had a really high failure rate on that pallet. The packing of the drives I think is partially to blame, but the client says they've shipped them like that before with no issues (I know, isn't it fun when you get the first problem lot?), so my next thought is the trucking company. Considering that Reddaway had one of the boxes come off the pallet somehow, this really makes me think that they screwed up handling the pallet itself. A lot of the drives that failed are mechanical failures, not sector-based failures. Most drives from this lot that failed will make a buzzing noise instead of starting the motor, or they'll have the click of death, or they don't turn on at all. This really makes me think that something happened in shipping. All I can do is just slowly make my way through the lot and pray that the rest of the boxes do better.

I had a theory about RAM too. Parted shows you a RAM and CPU usage bar on the side of the screen. They say that you can run PM on 1GB RAM, but I run 2GB in my testers (I don't like pushing the limits). With 2GB installed, after taking some off to run graphics (I think 64MB), during a wipe, the OS uses 15% of the RAM. So, we should be good there. But, for curiosity's sake, I upped all the computers to 8GB of RAM. They booted a bit faster (To be expected), but they didn't wipe the drives any faster. It makes sense- the command to wipe is issued by HDD firmware, so you work at the speed of the HDD, not the speed of the software.

This spring, I need to redo a LOT of my facilities- I've grown a lot faster than I ever imagined, and I'm running into growing pains. It's taking me way too long to process hard drive lots, so I'm thinking about adding a server or two like JJinLV mentioned above. For trouble lots like this, they won't work at all. But, for good lots, it shouldn't be a problem. A little tip though: If you do Secure Erase, your PCs must be super stable- crashes kill drives. So, while checking machines, I use 80GB desktop drives. If your machine does fail somehow and eats them, who cares? They're scrap anyways....

Between adding PXE and cloning for deployment of laptops and desktops (Yes, Parted Magic can clone drives, but it's kind of clunky at doing it) and upgrading hard drive testing, it's going to be a heck of a Spring for me....

Interesting that you use the Secure Erase command. Do all drives support those commands? I thought that not all drives supported those commands, although maybe I am mixing up those sets of commands with S.M.A.R.T .... whatever.

If you are doing this on pallet scale would it makes sense to buy like an enterprise grade wiping station/enclosure that is specifically built hardware/software for wiping drives? Something like this thing for example:

https://www.45drives.com/products/data-destruction/

Although, I suppose you could probably build an equally capable machine out of stuff you come across like servers and such (they are like $10k... but at a large enough scale it could be worth it..... and a pallet.... sounds like it is approaching that scale lol)

Not all drives can do a Secure Erase, but most can. Some older ones can't, but I don't see those very often. For those, I have to use a triple zero-pass wipe system. But, those drives are becoming more and more rare. I'd say that at least 95% of all the drives I see can do a Secure Erase. The ones that can't usually are the low capacity ones that won't be MTSV much longer anyways....

--------------------

I've looked into those enterprise things a little. For example, one of the products I saw was the White Canyon FlexStation. They say that it can take 24 drives without caddies. Honestly, that doesn't impress me much- here's why:

Caddies aren't a big deal- I don't use them anyways. I could easily fabricate the case they've built with less than $100 of stuff at the Home Depot. I had to do a lot of digging to find a price on the thing, but it's about $14,000 .... PLUS licenses to actually get the privilege to use their wiping software. For me, that's an insane amount of money that I can't justify (Even if I had $14,000 burning a hole in my pocket... which I wish was the case )

I'm using HP towers and some monitors that came in for E-Waste recycling. So, I'm out nothing in hardware costs. Parted Magic is a one time purchase- it can be used an indefinite number of times, with no further cost to me. I have six computers, so I can wipe 24 drives at once. That's equal in capacity to one FlexStation, but $13,989 cheaper.

I think JJinLV has a point with using servers if you're doing volume. I could probably benefit from a few of them. Doing a basic eBay search shows an example server - an HP DL320s for about $250 - is a LOT cheaper than the FlexStation. Granted, that model can only take 12 hard drives, but two of them equals a FlexStation for way less than $1000. And again, I don't have to pay-per-wipe.

I don't like how the cables become a spaghetti mess with my machines, and they take up a LOT of space. But, I know that they'll work. I've run a few thousand hard drives through my testers. I believe a lot in "If it ain't broke, don't fix it". I don't know a lot about servers, but I'd like to try it out when I have some more room to set up one of my racks and a server to play with. I talked with another kinda-local SMF guy about servers- when he gets a SATA server in, I'll give it a shot and see if it works better. Right now, I've got so much stuff going on that it's super exciting, but at the same time, I can't worry about upgrading systems that are doing their job well

Originally Posted by hills

Whew ... 10 k. is a pretty big investment for a small business.

Look forward to the future though. I think it's a matter of timing and when the next wave of change will crest. The trend with personal computing has been from large to small. Traditional hard drives will always have some market share but that's shrinking. Many of the smaller devices have already gone to SSD chips that are soldered right to the board. In a few years ... there may not be very many traditional HDD's available for wiping and resale.

It might be best not to invest in a technology that may soon become obsolete ?

Yea I don't think it is worth the money, or a wise investment, was curious if he had ever looked into it or not.

I do think storage as we see it will change tremendously. I don't think devices will come with much (expandable) space on them in the future, and there won't really be a need to.apple has already done away with expandable storage on their devices. Things may come with like 500gb or 1tb in an embedded flash chip like you said, but with internet speed increasing continually and quickly, everything soon will be cloud storage since it'll be cheaper for Amazon or Google or Microsoft to hold all the data and stream it back and forth whenever it is needed. If the internet is fast enough, there is really no need to for a huge amount of local storage .... We arent there yet.... But probably eventually!


For example, your phone camera could take a picture hold it in ram until it gets it uploaded and verified it's uploaded then flush from ram. No real need for local storage there. Or for example, playing a videogame, while walking around it could just be loading the world around you in real-time if the internet is fast enough. Which it is very close to right now...

The only people who will need local storage is businesses or individuals who want it for "privacy" .... But even that, could not be the case... Most businesses are already in or headed to something like AWS or azure and not having their own storage anyway.