Used enterprises drives are cheaper and better than new consumer ones! Checkmate! 
Actually, power on/off and associated thermal stress is the most taxing thing for a drive. Running stready state is not. So if anything storj prevents your drives from sleeping, prolonging their life.
Do you already have such a load?
I am from Ukraine and have not noticed the increased load; the channels are idle.
I feel WD and HGST shall be avoided. Never had a Seagate fail in the last 10 years. Every single WD and HGST did.
Both you and I are wrong on this one. There are bad and good batches from any vendor. Not one vendor is inherently better or worse. And with raid drive failure is inconsequential, so buy whatever is cheaper.
Drives at this point are consumable commodities. Brands donāt matter. Itās race to the bottom on price, and gimmicks on consumer lines.
They are made to pass the warranty time line. Like 5+1 year for safety. Like every electric appliance; when the warranty is over, you must expect to buy a new one.
Amen! If you want availabilityā¦ use mirrored/parity configs. If you want recoverability then automate backups. No make/model canāt fail tomorrow: every one is doomed to die.
So buy lowest $/TB. Or $/TB/years-of-warranty if you want help with replacements. And for Storj prefer density to keep power/cooling costs down.
Completely off topic but SK Hynix has 300TB SSD prototype.
And yet there are still folks who think SSDs will never cost less $/TB than HDDs. I think itās inevitable: we have maybe 3 years left.
SSDs getting cheaper $/TB is purely manufacturing scale: no new science: they passed HDDs in density years ago. But HDDs continue to struggle for every gain in capacity: requiring essentially new engineering and materials for every advancement.
I canāt wait!
Yea i never understood, why donāt we have like only SSDās from like 5 years at least?
Whats the problem in making 16TB HDD, vs SSD?
are materials more expensive? geeeeez.
Problem is that when I look at 4TB SSDās, prices are about twice of what this graph suggests they are in 2024. And 4TB would be useless for Storj. Itāll likely be a lot longer before 20TB SSDās are going to be cheaper than HDDās and by that time there will be larger HDDās available too. I think they will coincide for a while longer.
And yet in 2023 the chart estimated $40/TB for SSDs: which we definately hit on budget models. Yes flash prices have seen a recent surgeā¦ but HDDs also raised two years ago when Chia drained the market of large drives. So yeah prices will surge and sag.
and whats better for āsolar flareā? they say its comming sooner or later like in 1859! ;>
I donāt see SSDs going cheaper than HDDs for as much time as HDD can still increase capacity and be a player in the market. The materials are cheaper. You store data on glass/plastic/aluminium, not on electronic cells.
Yes, there are SSDs bigger than HDDs and can scale up much faster, but the prices are way too prohibitive, and I donāt see them come down to qucikly because the big majority of users are pretty happy with 1-4TB SSDs and pay those huge prices conpared to an HDD. I have like 20 laptops and PCs, for business and personal use, and I donāt use more than 100-200GB on each. So Iām pretty happy with 1-2TB SSD and I donāt need more. Why buy 100TB SSD? Data centers buy the cheapest thing on the market. So 50-100TB SSD going cheaper than 100TB HDD? I donāt think so.
When data storage needs for AI and virtual life will be so huge that HDDs canāt keep up anymore, SSDs or crystal balls will be the only option, and nothing to compare too.
Magnetic recording is way more resilient compared to electrostatic. Free flying elementary particles can easily flip the bit, or shift he encoding; there is error correction of course but it is not allmighty. Most SSDs require constant partrol monitoring and refreshing cell data as it deteriorates; it gets worse with higher density and more complex symbol encoding constellations. SSD power-off Data retention is quite horrible. So for archival storage HDDs may be superior. Heck, tape is still used (and Iād argue itās way better than HDD due to its simplicity. There is just tape with data ā HDD has a lot of crap that can fail besides the data)
Someone please give me an answer and better stay on topic.
(sorry, but reading millions of posts of other peopleās off-topic thoughts is very tiring)
I donāt see any positive changes or increase in load yet, but they wrote about some fantastic numbers that look more like ācapacity extractionā rather than storing useful information.
Saltlake shows increased traffic from the start of May. I donāt think Storj is promising any particular rates of ingress for any particular duration: so weāll all have to watch the global dashboard to see whatās coming through.
But Iām excited like you are: probably lots of SNOs are looking forward to more data: even if itās synthetic!
Thank you for your answer, but alas, I donāt see a surge in data receipt.
Perhaps there are geographical restrictions?
The customers didnāt signed any agreement yet; Storj is only preparing, does tests and probably makes a report with their findings for customers. If they will be convinced, than will join and we will see the much awaited ingress. But this can take weeks or months. Itās a new technology, donāt expect a big company move over night itās PBs of data to something new and still in developement.
The bosses must be convinced and this takes time.
I have nodes in the UK, Portugal and Poland and have not seen any significant increase in node activity either.
Perhaps itās either not started yet or is limited in geographical scope?
There are some important changes in the upcoming storage node release:
I believe that there are definitely geographical restrictions, but they try not to comment on this, since STORJ is positioned as a decentralized storage, and real users need a separate continentā¦
If Iām right and there are geographical restrictions, please try to eliminate the stratification and take care of all operators.
