Hard Drive Stats for 2019

I have ~10mbps upload here in the US, and my backup was looking to take about a month for about 3ish TB of data. One thing that helped is that with default settings of only 1 backup thread, the Windows client was unable to saturate my upload bandwidth. Upping it to 4-6 threads allowed it to keep enough data moving to actually saturate my upload bandwidth and brought my backup down to like a week.

mherrmann5• 6 years ago

Does anyone here have experiences with BackBlaze's B2 service for hosting files? I'm considering switching to it from S3 because it is much cheaper. (I need to transfer 2-3TB / month, usually in 2-3 bursts of worldwide distribution).

the_svd_doctor• 6 years ago

I use it for personal backups with rclone. Works great.

atYevP9• 6 years ago

Yev from Backblaze here -> We're definitely more affordable and our integrations (https://www.backblaze.com/b2/integrations.html) make it easy to get your data to us. We even have partnerships with companies who can help transfer data from S3 into Backblaze B2!

mherrmann5• 6 years ago

Thank you Yev. I'm wondering about the bandwidth, especially internationally. Do you have any numbers on that? Say split by Europe/US/Other.

syedkarim• 6 years ago

How is Backblaze able to be so much cheaper than the other, larger competitors? I assume Amazon/Google/Microsoft has squeezed every last cent from suppliers and also has highly cost-optimized staffing costs.

Moeancurly• 6 years ago

Is there any consensus among Backblaze employees (or even just your personal opinion if applicable) for what brand/series of drives to use for home NAS devices?

I ask because the online favorite appears to be WD Reds, which you have phased out since 2018.

tjonsson• 6 years ago

I used them for both backup (B2 storage with restic on linux servers) and also for serving static content for my homepage, together with Cloudflare ( https://www.backblaze.com/blog/backblaze-and-cloudflare-part... ) Works like a charm

ronnier• 6 years ago

I have two ST12000VN0007 (VN) Seagate drives. The report shows the ST12000NM0007 (NM) has a 3.32% failure rate. I wonder how closely related the VN and NM models are.

war10252• 6 years ago

If you look, that drive model is also the most highly used by far. I think it's just a matter of the larger sample size / use time.

JohnJamesRambo• 6 years ago

Surely it doesn’t matter when you have 10,000s of drives? Aren’t you already at a large enough sample size? If it isn’t, what is the point of them publishing this every year? I don’t know the math of the matter though.

Kubuxu• 6 years ago

Looks like the Segate 0007 are 1y old on average, where the 0008 are 44 days old on average.

The 12TB HGST are 220 days old on average. The Segate 12TB failure rates seem high, quite unfortunate as I own 6 of them.

UI_at_80x242• 6 years ago

I live for these reports. Always insightful and professional. Thank-you SO MUCH for publishing this data.

atYevP9• 6 years ago

Yev here -> You're welcome! The conversation's always fun :D

donmcronald• 6 years ago

I barely had time to skim it, but I'm not sure I like how the ST12000NM0008 shows up in the table. I find it really hard to reason about what the real failure rate could end up being on those drives. For example, you've got about 45 days average on each drive, so the failure rate is multiplied by roughly 8 to extrapolate the annualized failure rate. Doesn't that over state the estimated rate of failure since drives will tend to fail more often at the start of their life?

I only guesstimated out of the table and didn't have time to look at the actual data, so it's possible I misread something.

alberth• 6 years ago

Why do people use Amazon S3 when Backblaze B2 is 1/4 the cost of S3 and also includes a CDN for free. You also get way faster access speeds with Backblaze vs Amazon since they tier their IO speeds.

https://www.backblaze.com/b2/cloud-storage.html

smeyer• 6 years ago

I think a big reason is that people are using the rest of the amazon ecosystem. If your costs aren't primarily storage, you might be willing to pay a premium to use something that integrates nicely with other services you're using. Here's an article[0] that does some other comparisons between providers and mentions things like upload speed and security features.

[0] https://www.cloudwards.net/azure-vs-amazon-s3-vs-google-vs-b...

3fe9a03ccd14ca5• 6 years ago

IAM access control, ease of use, reliability, speed.

mythz2• 6 years ago

Where's the info about the free CDN?

nielsole• 6 years ago

Check Cloudflare bandwidth alliance

koolba• 6 years ago

Paying for outbound bandwidth is a big one.

rvnx• 6 years ago

In the same way, why people use Backblaze when they can use Wasabi and not pay the bandwidth ? https://wasabi.com/cloud-storage-pricing/

kissgyorgy• 6 years ago

Because of vendor lock in. When you move a lot of data between S3 and EC2 it costs nothing (or very cheap). When you move data outside of AWS, there is extra cost, so it might not even be cheaper overall.

mceachen• 6 years ago

Unless you're a bootstrapped startup with just a couple people, paying the AWS bill is not something the engineer probably thinks about too much. Setting up a new billing account with another company is just enough friction to just use whatever AWS offers and call it a day.

Also, most employees aren't really incentivized to reduce or minimize infrastructure expenses.

mritun• 6 years ago

There are usually many reason:

1. Scale - S3 is big - really really big! You don’t need to care if you store one KB or several petabytes.

2. Tiers: the default on S3 is several way replicated storage with 11 9s of durability with high availability. However you can select from cheaper options with the trade off you are happy with.

3. Cost: S3 has reduced prices several times, you can be reasonably sure your costs will go down over time on per unit basis.

overcast• 6 years ago

Here's one reason I would need something like S3. Sure, I hack all that together into something barely functional myself, but it's not worth it. Pretty handy. https://aws.amazon.com/solutions/serverless-image-handler/

bwilliams18• 6 years ago

I too use the serverless image handler but it's not perfect. The documentation is really crappy and over the summer they transitioned the whole system from thumbor to sharp and didn't provide great backwards compatibility.

meritt2• 6 years ago

Because the rest of my infrastructure currently runs on AWS and aws egress charges are far more expensive than the b2 savings.

rsync• 6 years ago

Genuinely curious ... do you not assign any value to having a backup outside of Amazon ?

AWS can certainly provide geographical diversity, but on the organizational abstraction layer, all eggs are in one basket, yes ?

Is having organizational redundancy something you assign zero value to, or something whose value conflicts with the egress costs so as to make it a difficult decision ?

Again, genuinely very interested ...

cheeze• 6 years ago

Last I checked, Backblaze still stores most data in 1 location, no?

So, durability of data (which to be fair doesn't matter for most s3 use cases), and interop with literally everything else in AWS

Intelligent data tiering

Actual access control

Pre signed URLs

chocolatkey2• 6 years ago

I've combined cloudflare workers with backblaze to implement etags, signed URLs, etc. Backblaze is part of CF's bandwidth alliance so your bandwidth fee is zero. This makes for a very low monthly cost

shiftpgdn2• 6 years ago

S3 is single zone unless you specifically request multi zone durability (most don't.)

brianwski2• 6 years ago

Disclaimer: I work at Backblaze.

Last I checked, Backblaze still stores most data in 1 location, no?

Backblaze now has multiple regions! One in Europe (Netherlands) and one is called "US-West". Quietly the US-West is actually three separate data centers, but your data will only really land in 1 datacenter somewhere in US-West based on a few internal factors.

To be absolutely clear, if you only upload and store and pay for 1 copy of your Backblaze B2 data, it is living in one region. To get a copy in two locations you have to pay twice as much and take some actions. So if this kind of redundancy is important to you for mission critical reasons Backblaze B2 would only be half as expensive as one copy in Amazon S3, not 1/4 as expensive.

In the one copy in one region in Backblaze B2, any file is "sharded" across 20 different servers in 20 different racks in 20 different locations inside that datacenter. This helps insulate against failures like if one rack loses power (like if a power strip goes bad or a circuit breaker blows). But if a meteor hits that 1 datacenter and wipes out all of the equipment in a 1 mile blast radius, you won't be getting that data back unless you have a backup somewhere else.

partiallypro2• 6 years ago

If you're with another cloud provider, you still have to pay egress fees to Backblaze. That cancels out the cost savings.

bithavoc• 6 years ago

B2 does not implement the S3 API. Also the B2 API is much slower than S3.

brianwski2• 6 years ago

Disclaimer: I work for Backblaze so I'm biased. :-)

the B2 API is much slower than S3.

This is "generally true" for 1 upload thread. We aren't even sure what Amazon is doing differently, but they can be a little faster in general for 1 thread (some people only see 20% faster, some see as high as 50% faster, might be latency to the datacenter and where you are located).

As long as you use multiple threads, I make the radical claim that B2 can be faster than Amazon S3. The B2 API is slightly better in that we don't go through any load balancers like S3 does, so there is no choke point. What this means is that in B2 40 threads are actually uploading to 40 separate servers in 40 separate "vaults" and none of the threads could possibly know the other threads are uploading and it does not "choke" through a load balancer. This was all designed originally so that 1 million individual laptops could upload backups all at the same time with no issues and no load balancers. And it works great every day.

Practically speaking, for most people in most applications, this means both Amazon S3 and Backblaze B2 are essentially free of any limitations. If you aren't using enough of your bandwidth, spawn a few more threads (on either platform) and soak your upload capacity. But in full disclosure, if your application is only single threaded, yes, B2 tends to be 20% slower for that 1 thread.

leokennis3• 6 years ago

I use B2 as “cold storage” of large-ish files. It’s incredible how low the monthly bills are.

bob1029• 6 years ago

I am strongly considering B2 as an option for a dropbox-style system. Something where I run 8-16TB of hot tier on my local LAN, with B2 serving as the slower mass storage tier behind it. It seems that the average B2 access latencies would be ~100-200ms, which is very tolerable for a cache miss on such a massive tier of storage. With this amount of space available you could have pre-fetch rules that do things like pull down entire directories as files within them are accessed.

generalpass2• 6 years ago

I have wondered about system downtime or time operating in a degraded state.

My understanding is other than mirrored, RAID configurations may take a long time to rebuild on the larger drives and this is a contributing factor to why the highest sales volume of drives has been 'stuck' at 4TB (thus the lower $/GB price).

linsomniac• 6 years ago

They don't use traditional RAID setups there. My understanding is they use a proprietary data encoding and distribution, which is more accepting of individual drive failures and reduces rebuild times. I believe I've heard they use something more like erasure coding rather than RAID-5.

ddorian43• 6 years ago

https://www.backblaze.com/blog/reed-solomon/

There are many open source libraries.

throwaway17_172• 6 years ago

Does anyone have any opinions and experience using backblaze as a personal only cloud storage and offsite backup for smaller amounts of data (under 30 TB)

clSTophEjUdRanu• 6 years ago

+1 I have the same question and would like to read replies.

dwohnitmok• 6 years ago

I use Backblaze's B2 service for both backup (via restic) and archival storage (via git-annex). I only maintain a distinction between the two in case I ever want to move to another service, and also because git-annex and restic have different strengths that make them more or less suitable for unchanging archives and often changing backups respectively. Between the two I have about 1 TB stored with them.

I have yet needed to do a full restore, but I do partial restores from time to time to double-check my backup procedures and every time it's done what I wanted. My monthly costs are usually a bit under $5.

Note I essentially never use B2's API directly, and only use it as a backend through wrappers others have written, so I have no real experience with how good its API is. One of the few times I did try the API, I remember at one point I think I was getting Java exceptions back in the error messages, which was mildly concerning from a hygiene perspective and made for rather terrible error messages, but no sensitive data was being emitted. I also think that's been fixed.

The bottom line is that B2 has worked fine for me and at a good price point.

newscracker3• 6 years ago

for smaller amounts of data (under 30 TB)

Did you mean to say 30 GB or 30 TB? Calling 30 TB as "smaller amount" seems weird to me in 2020, especially for personal data. Perhaps it would be the norm in a couple of decades. :)

FWIW, I have way under 1 TB of personal data to backup to different locations, and I consider that to be relatively large.

throwaway17_172• 6 years ago

I did mean 30 TB, I have approximately 12 TB of data currently between all of my storage for video, audio, books, and games. However, I have been avoiding doing a lot of conversions to digital media from my physical collections because I'm just unsure of running a full blown archival server at home. I would estimate if I converted my entire video library to 4k it would put me somewhere over 10 additional TB. My comic books/manga and graphic novels, upgraded to archival resolution would probably run over 10 TB as well. Then there is the soon to be required ripping of PS2/PS3/WIIU roms when those hardware units become less reliable for actual playing. So I think that 30 TB of storage would do for the time being for me, but I think I will eventually need more than that.

TL;DR I am a digital horder, so I've convinced myself I do in fact need 30+ TB of storage.

tshannon• 6 years ago

Yeah I use B2 with rclone (https://rclone.org/) and it works great.

sillyquiet• 6 years ago

No really related remarks about this handy study, but anybody else still in real awe about how spoiled we are with regards to the sizes and speeds of HDs nowadays? I mean the smallest capacity drive on their chart is 4 Terabytes.

newscracker3• 6 years ago

Since HDDs have, for the most part, been relegated to being external drives on laptops, I'm still looking forward to SSDs becoming way cheaper and reaching current HDD prices per GB. Internal storage on laptops has reduced or stayed the same while our datasets have grown exponentially over the years (with photos and videos). Since SSDs also perform much better when there's always a good amount of free space (for wear leveling and maintenance), it's all the more painful to live with lower capacity SSDs on laptops.

klodolph• 6 years ago

Sizes, yes, speeds, no. 600 MB/s of data transferred, and only for linear accesses.

zaat2• 6 years ago

Not feeling spoiled at all, not at all. Especially not with 2 to 3 percent of failure rate. The failure rate I experienced in my workstation makes me worry about not having raid 1 or 10. HDs for 9 TB in raid 10 are not that cheap.

But the bigger issue is that the warranty terms for HDs nowadays is down to 2 or 3 years, so this investment is short living. It also tell you something about the manufacturers reliability estimation of their products.

AnIdiotOnTheNet2• 6 years ago

Can't say I agree with that sentiment. The fact that I can quite reasonably have a 30TB usable RAID5 NAS array makes me feel pretty spoiled. Then again, I'm old enough that my first HDD was 10MB.

Siecje• 6 years ago

I have about 10 TB of video files. I use BackBlaze for Windows but I would like the files to be available on other computers and my phone in my local network.

What can I use to do this and still keep offsite backups?

mrguyorama2• 6 years ago

I think their more premium plans offer sharing

sandoooo• 6 years ago

Dropbox?

robertoandred• 6 years ago

I love Backblaze, but their log package in my Library folder has grown to something like 10 gigs. Wish there was a way around that.

newscracker3• 6 years ago

Slightly off topic: is anyone using B2 (which seems cheaper if you have more than one computer for a certain amount of data) for personal data backups with strong client side encryption across multiple platforms (Linux, Mac, Windows)? If so, how do you handle it?

w33ble• 6 years ago

I do this, though not from Windows, just Mac and Linux. I use restic, which has B2 support smd handles all the encryption. it also does diffing for backups. There's a Windows build, so I assume it would work for you there as well.

You can view and download builds at https://github.com/restic/restic/releases/

I don't automate this though, I just use it for occasional backups. Not sure what the automation story around restic is.

hashhar• 6 years ago

Yes. I use restic same as the sibling comment.

Have >8TB of data from multiple machines with a lot of deduplication (source is somewhere around 10 to 12TB).

S3raph3• 6 years ago

I sync all my device files to a local Freenas server which runs duplicacy in a jail and sync's it every night to backblaze B2. I looked at duplicity, restic, attic, borg and in the end settled for duplicacy. Pay attention to the duplicacy license, for somebody it could be a problem.

orhanhh• 6 years ago

I use Arq on two macs and it works very well with B2.

zippergz• 6 years ago

I use and like Arq also, but the OP asked for something that covers Linux, which I believe Arq does not.

seanlane• 6 years ago

Similar to the other two sibling comments, been using restic to sync to B2 over the past 6-7 months. Stored amount has been 450-475 GB, and total costs tend to be about $2.50-$2.75 per month.

ksec2• 6 years ago

Looking at those Data,

It seems they will soon reach 1000 PB / 1EB.

The top 5 Annualised hard drive failure rate are all from Seagate. All Drive from Hitachi and Toshiba has AFR lowered than 1%.

So basically dont buy Seagate.

arminiusreturns• 6 years ago

It's pretty much been this way for a few years, with only a few model lines of Seagate being the outlier. As always, thanks to the BackBlaze team for publishing these numbers.

LeifCarrotson2• 6 years ago

My math says they're already over 1000 PB/1 EB:

1,089,318 = 4 * 2852 + 4 * 12746 + 8 * 1000 + 12 * 1560 + 12 * 10859 + 4 * 19211 + 6 * 886 + 8 * 9809 + 8 * 14447 + 10 * 1200 + 12 * 37004 + 12 * 7215 + 4 * 99 + 14 * 3619

Don't think I made a typo there, but please check my work. Even counting as 1024 TB = 1 PB and 1024 PB = 1 EB, that leaves 1,048,576 TB = 1 EB and they're over that threshold.

The February 5, 2018 "500 Petabytes and Counting" blog post should soon be eclipsed by a 1 EB post - though it appears they're counting actual data stored, not capacity. Nonetheless, with some redundancy, extra capacity, and overhead, we'll likely see that number soon.

AdamGibbins• 6 years ago

So basically dont buy Seagate.

Or do, because they're cheaper than the competition and modern systems can handle failures.

antonyh3• 6 years ago

True with enough redundancy it's fine, and if they have special terms with SG such as free replacements and heavy discounts then it's little wonder they use so many.

Myself though, for SoHo use, I'm willing to pay more for less stress because I don't have the sheer volume of devices, and the time to replace is time spent doing something useful instead of shuffling HDDs and rebuilding RAID arrays. A 5% saving on a handful of drives is not worth it, but a 40% saving on thousands makes them competitive.

ksec2• 6 years ago

Every time I look at HDD price it seems Seagate is always a little more expensive. But the difference is less than 5% to the point one may argue they are all priced similarly.

I dont want to save a few dollars for potentially 4x the chance of failure and hassle.

And even if we ignore the outlier to 2%+ from two models, Seagate is still on average 2-3x more likely to fail.

LeifCarrotson2• 6 years ago

Interesting how the numbers carry over year-to-year in

https://www.backblaze.com/blog/wp-content/uploads/2020/02/Bl...

Some models are dwindling. Some are being tested. Others (like the Seagate and HGST 12 TB) are increasing. Only thing that's really perplexing is why they keep buying more and more of the high-failure-rate Seagate 12 TB drives. It must be more than 3% cheaper to buy (and service!) a Seagate with a 3% chance of failure than to buy an equivalent HGST with a 0.4% chance of failure. I guess when you have 120,000 drives, easy hot-swap enclosures, and software to handle it all that makes good sense! But as an individual consumer, even with a Backblaze backup, it's definitely worth my time to spend a bit more on a drive that's far more reliable than to save a few dollars on a Seagate.

alanfranz• 6 years ago

At least in Europe, HGST is much more expensive than Seagate. Almost double the price, usually.

leokennis3• 6 years ago

Very anecdotal evidence, but 3 of the 3 Seagate drives I ever used (all external 2,5” USB 3 HDD’s, in Seagate’s own enclosures) failed within 2 years, under very modest workloads (just used to store video files for my tv to play).

Meanwhile all WD’s have been rock solid.

hinkley• 6 years ago

If I make a hard drive, and sales are crappy, in part because BackBlaze told the world how shitty they are, I'm going to have to drop the prices to move product.

I suppose there's a movie plot in there where BackBlaze negs their favorite drive so they can buy them cheaper.

sixothree• 6 years ago

Only thing that's really perplexing is why they keep buying more and more of the high-failure-rate Seagate 12 TB drives.

I am guessing they RMA the drives and get replacements.

universenz• 6 years ago

Your comment just sparked an interesting question in my mind: If a drive has failed, until now I always imagined the drive was just trashed. But now that you mention they are probably RMA'ing them, do you think that BackBlaze send the RMA drives through a magnetic tunnel of some sort before they ship the drives back to the manufacturer? Because otherwise, how do they ensure potentially unencrypted customer files are not accessed during the repair/refurbishment process?

ahofmann• 6 years ago

They once wrote somewhere that they have contracts with Seagate that allow them to get the drives much cheaper if they buy certain quantities.

magnat• 6 years ago

Does anyone remember what is their definition of "drive failure"? Is it SMART "failure imminent" report, single uncorrectable read error or complete data loss for a whole disk? I recall reading about it in one of their previous report, but can't find it again.

EDIT: nevermind, found it.

"Backblaze counts a drive as failed when it is removed from a Storage Pod and replaced because it has 1) totally stopped working, or 2) because it has shown evidence of failing soon.

A drive is considered to have stopped working when the drive appears physically dead (e.g. won’t power up), doesn’t respond to console commands or the RAID system tells us that the drive can’t be read or written."

https://www.backblaze.com/blog/hard-drive-smart-stats/

jmnicolas• 6 years ago

Any particular reason they don't use Western Digital drives ?

volkl48• 6 years ago

I will point out that HGST is owned by Western Digital and all their products are being rebranded to WD.

forgot-my-pw• 6 years ago

Other than HGST is owned by Western Digital, they also said:

There were no Western Digital branded drives in the data center in 2019, but as WDC rebrands the newer large-capacity HGST drives, we’ll adjust our numbers accordingly.

gesman2• 6 years ago

So what does these mean:

smart_177_raw

smart_177_normalized

smart_233_raw

smart_235_normalized

???

thenewwazoo• 6 years ago

They're S.M.A.R.T. attributes: http://www.cropel.com/library/smart-attribute-list.aspx

gesman2• 6 years ago

Thank you!

S3raph3• 6 years ago

I'm a very happy customer, but please do something about your mobile app (android) it's really horrible.

leokennis3• 6 years ago

I agree the mobile app (on iOS in my case) is at best an afterthought, and most likely not even a high ranking afterthought.

However, out of curiosity...what would you imagine a better Backblaze mobile app would do?

S3raph3• 6 years ago

For sure it is/should not be high priority, but releasing such an app in 2020 for sure does not reflect the great skills of the backblaze team. At least show me some basic stats, account settings and invoices. You can only download files from your buckets and that's it.. really?

exabrial• 6 years ago

I still can't believe BackBlaze gives this data away for free. Seems like something they should be selling to other cloud providers

Mirioron• 6 years ago

Maybe they consider this report to be an ad for their services? The name recognition this report gives them is probably quite valuable.

icelancer• 6 years ago

It also pressures HDD companies to make better products and appear higher on these lists, which is good for Backblaze.

kevinchen3• 6 years ago

I’m sure Amazon, Google, Facebook, etc collect their own stats on drive failure. It would be almost negligent if they were just guessing in the dark every time they buy drives.

Main difference is probably Backblaze is small enough to publish these stats without hurting their supplier relationships. (pure speculation)

nolok• 6 years ago

To quote them : Transparency breeds trust. We’re in the business of asking customers to trust us with their data. It seems reasonable to demonstrate why we’re worthy of your trust.

donatj• 6 years ago

I have made all my hard drive purchasing decisions based almost entirely on these reports for the last couple years and have not been disappointed with the results.

voldacar• 6 years ago

Seagate always seems to have much higher faliure rates compared to HGST/WD/Toshiba etc

Does anyone here know the exact reason why? I assume there are enough people on this site who have worked for them or a competitor :)

rasz• 6 years ago

30% profit margin, why change anything?

anonsivalley6523• 6 years ago

Ah yes, the reliable BackBlaze folks. That they've out-Googled Google in a niche using mostly commodity infrastructure and kept their business alive for so long is a testament to their ingenuity (I wonder how their operating costs compare with AWS Glacier which has a theoretical advantage of unpowered disks.). And the releasing of this proprietary operational business data is a testament to their coolness factor.

It's a timely article as I'm looking at HC530's (WUH721414ALE6L4 / WUH721414ALN6L4 (wiredzone carries it)) for a home FreeNAS box:

- any relatively-modern enterprise 4U 3.5" storage box with Xeon 4 cores or so

- quieter, high-volume fan mod

- RAM: 64-128 GiB, beyond that isn't useful unless deduping

- NIC: X710-T4L 4x 10GbE copper NIC

- ZIL: mirrored pair of high-endurance, write-intensive, reliable SSD like Optane 900p/905p 280-480GB

- L2ARC: striped pair of read-intensive/larger SSDs like the Gigabyte Aorus Gen4 1 TB

This will fit nicely as my home NAS for a water-cooled dual EPYC virtualized server/workstation build underway. I managed to get a single water block with (3) G1/4 connections that will cool both CPUs and the VRM chokes/converters.

If anyone has better suggestions, please chime in.

tbrock• 6 years ago

Why does someone need something like this? I ran a home symbology NAS at one point but it wasn’t worth the trouble. Let others run and maintain those hot, loud, power hungry disks.

anonsivalley6523• 6 years ago

Then you already made the mistakes of:

- conflating trouble for you with trouble for me, which it clearly isn't

- not owning your own data

- paying more to store it

- paying to access it

- ability to keep things that aren't worth storing on paid clouds but aren't all that much when kept on cheap drives

Furthermore, there are additional network costs such as AWS network charges AND home ISP data limits.

And there are other uses, such as:

- backing-up VMs

- backing-up computers

- caching package and source code repos

- backing-up CCTV footage

- and whatever else comes along

zaarn• 6 years ago

Speed. Streaming files (movies, tv shows, youtube channels, linux ISO's, large file collections) is a lot faster; I can reliably hit 100MB/s on my home connection, my DSL caps out at 20MB/s if I do nothing else.

kevinchen3• 6 years ago

I’m a Backblaze customer of many years & respect their team a lot. But seriously, “out-Googling Google” because they have cheaper storage is a meme that needs to die.

GCP and AWS both store full copies of your data in multiple locations by default (Availability Zones in AWS-speak). So it’s not an apples to apples comparison. The reduced redundancy is priced in, for people who can tolerate it.

anonsivalley6523• 6 years ago

Woah, chillax the dramatic rhetoric, your majesty. ;-P

The original scrappy Google was founded on commodity hardware held together by LEGO. The point was to not do as enterprise with redundant everything, which was wasteful for web-serving use-cases that were solved with better high-availability in software. These days, if you're a giant company like FAANG, you can easily afford to go to Quanta and say: give me 10k racks worth of compute nodes to this specification. If you're starting out and broke, you gotta use what's on the shelf, cobble together a custom solution optimized for the purpose and/or kit out a test lab with a mis-mash of used servers from eBay.

Hard Drive Stats for 2019

Related Stories