Maybe you can optimize this on your side and archive the files or pack them?
For me Storj is also more expensive than my actual backup solution, a simple FTP space.
Two incomparable services.
The question also is how resilient that FTP service is. Equipment might fail and data centers might burn down. We have seen this in the past and in some of these instances the data was irretrievably lost.
I think you meant to say â⌠80 simple FTP spaces⌠they I spread my data across⌠such that I only need files from 29 of them to restore.â
For me, itâs just a backup of data I already have in other places. The DC have 2 disks on RAID, so I have a little âbackup of the backupâ. I know isnât the same as Storj but I just want to say that Storj isnât for all the uses. I would like to use Storj for my backups as Iâm a SNO, but I have free transfer with the FTP, cheaper space, and easier to use.
it is when you need to drive you from A to B you will buy zaparojec - Search Images (bing.com)
it will drive you from A to B, but for some reason people buy something better.
I need Storj to be the same as S3, so optimization by packing a few together is not possible for me (and probably for most user as well)
Sorry but this is a bad example, I can only assume that no business will trust solely on Storj, as Storj themself state in their condition you canât hold them accountable for data loss.
So any sensible user will use Storj + some other backup at least.
Now, when this is on the table we can stop giving Storj the title of âmost reliableâ since there is no legal backup for this claim (and even if we had some, still no business will trust only Storj) so this claim is irrelevant.
Like it or not, Storj is just an object storage with a good (not better, unfortunately) performance on retrieval.
As such these are the interesting parameters:
A. storage coast (completely including the segment fee game)
B. BW coast
C. Retrieval time
D. Reliability (Not in the sense that we duplicate each file in 1000 places, but in the simple sense of would the file be there when I try to get it in x% of the time) 11 9âs annual durability is enough, and the industry standard.
Based on these parameters we can compare apple to apple all the object storage solutions.
In this comparison, Storj Storage is VERY expensive (and also the price is quite misleading, the segment price is thwarting the total per gb price)
obviously, you can dismiss me, saying we already have customers, we have so special technology, and only for that people will pay premiums and on and on.
In the end, people choose Storj for the price-performance ratio.
performance = 11 9âs (more of that is irrelevant), and under x ms latency in specific locations (based on the individual user need) thatâs it (yes, I know, itâs wired how simple it is ha?).
If the price doesnât go down SOON, people will notice their bills make no sense and are expensive in multiplication on comparable services (like Backblaze) and the bad reputation is something you canât fix that easilyâŚ
I think there are not many businesses that would account for data loss, and if there would be much larger price difference.
If you would compare the risk of data loss compared to other S3 solutions then Storj3 would have the lowest risk I suppose. When factoring the risk into the price, Storj is probably not expensive at all.
If you want to compare apples with apples you should compare all features, not just similar ones.
Being expensive doesnât give you bad reputation. Being unreliable does. There can only one cheapest offer, every one else would have bad reputation then.
Yeah⌠even Amazon avoids responsibility for data loss with their offering (claiming the same 11 9âs for durability in S3, and 99%-99.99% for availability). In fact, most software terms-of-use documents say something like âwe make no guarantees about how this software will operate: not even that it will do what our marketing says it will doâ.
Storj is priced competitively for customers that need the core S3 features. If they donât need S3 (or the durability/availability promises) then there are cheaper ways to store ones and zeroes on the Internet.
You get a bad reputation when you quat on pricing that you take only 0.004$/GB on storage and then put a little star that says âyou have extra segment costâ and the user that chooses you finds out after a month that he paying ~0.03$/GB stored.
This is my point, you can go and say we are the most durable solution (with infinites 9âs), the thing is itâs not matter, above a certain point.
as a costumer 11 9âs and 15 9âs are the same for me, in ether case Iâm not going to restore from backup any other day, in each of them, I still need some backup, so we can establish that after we get to the industry standard (currently is 11 9âs) itâs just become good enough and behind this is no reason for an extra cost (for most if not all reasonable users).
Now to the price, you keep compare yourself to AWS S3 - this is not the comerisom at all, you donât have all-around solutions like them, you donât have their reputation, and you surely donât have their performance.
Storj can be compared to Backblaze for example, similar solution (s3 compatible object storage) similar durability, hell, Backblaze is even more reputable than Storj.
And when we see the prices then, surprise, Backblaze storage is WAY cheaper than Storj (Almost X5 in some cases!).
Not to mention the fact that Storj slashed prices for node operators, so this is just more reason why Storj should lower the prices of the storage as well (the cost become smaller)
If backblaze works for them better â they should move to backblaze.
Competing on price in the race to the bottom is a horrible strategy. Being the cheapest on the block is a dubious goal.
Users that benefit from features storj offers understand that storj today is already heavily underpriced. The same Backblaze is at least 3x more expensive if you want multi region storage, and a complete non-starter if you want performance from anywhere but their datacenter vicinity.
What bad reputation are you referring to?
It isnât. I would not touch b2 with a rotten stick.
Storj never lost customer data and never returned garbage through the API. Backblaze did, on multiple occasions.
This clearly is not the case. Why would you say that? You know, storing all data on a single HDD connected to raspberry pi is even cheaper than Backblaze and just as relevant
If there is a little star then everything is fine. If you donât read the fine print then itâs on you.
This is the case.
To start, there is no way to get 1GB storage at 0.004$ as advertised since the maximum segment size is 64MB and each costs 0.0000088$, so the MINIMUM you will pay is:
0.0000088 * 16 + 0.004 = 0.0041408, so the honest price equate should be start at: 0.0041408.
Now letâs decompile the âstart atâ.
the above price is ONLY for the perfect user i.e. a user for all files is exactly 64MB, lets take another user, letâs say our user wants to store a bunch of photos.
average JPG size is 10kb, i.e. this user will have 100_000 segments (in a gb), now letâs calculate his price:
0.0000088 * 100000 + 0.004 = 0.884$
How much the same will cost on Backblaze? 0.006$
ratio? X147! (in this example Storj is even much more expensive than AWS S3)
No, this is misleading, and this tends to backfire
At the end of the month, people will take a look at their monthly bill and will start to wonder how the hell they pay so much if the price quoted is 0.004$ and a little star. I checked in the document Understanding Storj Pricing Structure - Storj Docs "For most users and most usage patterns, we do not expect a Per Segment Fee to be charged. "âŚ
wellâŚ
I fail to see your point.
Itâs honest price. Everything is disclosed. Cost per TB, cost per segment, ingress costs (0), API costs (0), egress costs.
No. This is only for the 64G segment size. People may have different usecase, and can calculate their actual cost themselves.
Total cost is a sum of parts. Whatâs so hard to understand?
Lol. Are you from the past?
Thatâs the point, this is done on purpose.
The pricing is set to shift the usecase to benefit from the platform the most. Storing small files involves huge overhead, and is not desirable on any platform.
Application developer that utilizes storj as a backend may choose to clump your 10kb low resolution jpegs together into a single segment for cost and performance reasons.
This example is invalid. Itâs like saying âI cannot attach the cultivator to my Ferrari â itâs a shit overpriced carâ, when you actually needed a tractor.
Storing trillions of microscopic files in the bucket is horrendous idea regardless of the storage provider. But if you have to do it for some legacy reasons â this will be better suited for backblaze: you wonât benefit from storj highly concurrent distributed network because all of your performance will be eaten away by access latency, regardless of the provider.
It may surprise you, but amazon has similar overhead costs on some storage tiers, for the same reasons â to drive intended usage. Itâs just buried in the fine print.
Furthermore, your comparison with B2 is invalid because the storage classes are entirely different, B2 cannot offer what storj offers: they only have two separate datacenters, so you simply canât accomplish georedundancy; for same level of performance you would need to pay for CDN, there is no end-to-end encryption either. (Their offering has only one benefit â itâs somewhat cheap. Thatâs it. I would not use it)
This is why Iâm stating that storj is heavily underpriced. Heavily. I donât understand how can you claim otherwise in good faith.
The key here is to pick tools that are right for the job. Iâm not sure why do you want to ignore storj benefits, misrepresent the comparison, and pretend that nobody can read and understand value proposition and/or pricing.
Itâs like @meir went to a fancy steakhouse, but only needed a cheap burger.
He didnât want the valet service out front, or the well-decorated venue, or attentive waitstaff, or world-class chef, or prime cuts of meat. And although he chose a service that didnât meet his needs⌠heâs now complaining that the prices should be lower for everyone⌠âbecause itâs still just beefâ.
If someone doesnât need S3, or georedundancy, or SOC2 or 11 9âs⌠they wonât use Storj. Bunny or Petabox would love their business! Weâve all seen the same math and understand that even with the current payout changes Storj will just barely be in the black (after bleeding for years). I donât see where else they can cut. Does anyone?
I hope Select helps improve their finances. And that they keep looking for higher-value services to offer. Because the lowest-price market is crowded, and cutthroat, and not where they should be competing.