New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Wanted - Anywhere - 1vcpu - 1G ram - 50-60TB storage
unsafetypin
Member
in Requests
Wanted - Curious - 1vcpu - 1G ram - 50-60TB storage with ability to expand if needed
Any location?
Enough bandwidth for initial transfer, then incremental backups.
Curious what pricing on this would look like or if anyone on here could reasonably provide this. Not really interested in B2/S3 Glacier/others. Looking for block storage to store large backups. I looked at Hetzner but I dont really need a beefy box to store backups on. Last resort, I build my own box and throw it in the snow of antarctica.
Thanks
Comments
Buyvm slab can have upto 50TB max iirc. But good luck finding stock.
With Hetzner Auction you can get 4x10TB for 50 EUR or 4x16TB for 70 EUR. Thats 1.25 EUR/TB and 1.10 EUR/TB respectively.
I think you are going to have a very hard time finding anyone that can beat this kind of pricing.
Those specs with such high storage doesn't make sense at all.
You are better off using a dedi from Hetzner.
Yeah hetzner kills it with pricing on their auction boxes
I never saw a VPS with minimum spec (1vCPU + 1GB RAM) having more than 10TB, let alone 50++ TB. It doesn't make sense at all.
I don't disagree. I just don't need that much resource for a system used to store data that may or may not be retrieved and isn't running any applications. I'll have to look into some other options, likely making a small storage box and throwing it somewhere.
I made my own and colocate.
Yes, but you can buy Storage SLABS from BuyVM and attach it on their Dedicated KVM.
Sounds like the best move honestly. I've been looking at that. there's some awesome dudes in Dallas tx on here I was talking to a bit back about colo.
BuyVM is awesome, I have a VPS idling there as a backup for an application. But slabs are super rare to go in stock.
You said you weren’t interested in storage solutions but I can’t help to think that some really cheap cold storage could be interesting.
I just now made a choice between storing tons of data on my VPS and backing up to S3 compatible storage or storing it as a main storage on S3 compatible storage and then also backing it up to another S3 compatibe storage.
I went with the second path.
First of all they’re usually very safe and replicated (hence not even comparable), and, even more importantly, no scaling worries ever. Just upload, and that’s it.
Your ask of 1 GB ram indicates you really don’t want/need the compute power, so why not collect the benefits of not needing such?
I get that. The cost per TB is somewhat high and Wasabi is the only one that seems to be decent about degrees where B2 and S3 / S3 Glacier would be wildly expensive if I needed to do a full restore. You have a solid point there. $6x50 is $300 or so a month. For under that I could easily colo something and recoup the cost savings of the build.
At what budget are you wanting to stick within for this VPS?
Probably under $80-90/month. But I know I talked with you a while back about colocation so that's a pretty good option. Got some good ideas on options from everyone on this post.
We could offer you storage for backups:
https://netdynamics24.com/backup-services.php
contact us for details and discounts available.
Theres too much info missing.
Whats the usecase for that?
How often you will download this data and how much at once?
How do you want to connect to it?
How much data will you push every day?
You cant get any good recommendation without providing this info.
For example if your whole site is 10TB then your problem is transfer speed and you should avoid ANY VPS, because they share IO.
If you are constantly roating backups (let say these backups sit for 2-4weeks) then Wasabi is waste of momey, as they bill you for 90days per file so you'll end up paying 3-9x much.
Etc etc etc... like I said, write in detail what you want to do with it and why you want VPS in first place if you dont need compute at all. This is a lot of data, maybe you should think about failover with point-in-time snapshots instead of backups? In your price range you can get 1Gbit max transfer, its gonna be very long to restore even 1TB.
Sure, but cost per safety ain’t that high, considering it’s replicated.
Maybe there’s some other backup you could remove or make smaller and save money doing that, considering how replicated S3 is.
Hello sir if you like for pay anualy it s possible give you
48 TB Space (possible upgrade to 96 TB+)
1x e5-2450l
16 GB Ram
120 TB BW
1 GBps port speed
It s dedicated server
For 1150 euros/year , after the first year the price will be only 65 euro/month
B2 is replicated? I thought they are using only erasure coding.
Wasabi is 100% not replicated, they rotate data between servers thats why they have this minimum retention policy.
Don't know about Amazon offerings.
We've got several customers with 50+TB VMs with no problems, and we can do it in a bunch of locations: Los Angeles, Chicago, New York, Amsterdam, Stockholm - all top tier datacenters with top tier network mixes (no Psychz/M247/Cogent/HE, etc) and no DCs in the middle of nowhere.
Never used Wasabi.
BackBlaze shreds each file and can handle 3 drives failing until the ”tome” has been rebuilt (one ”tome” is 20 drives). (Edit: maybe that’s erasure encoding as you put it, not replication).
https://www.backblaze.com/blog/cloud-storage-durability/
I don’t think there’s been any case of lost data on B2 as of yet (apart from human/billing errors, of course. I’m considering only actual drive failures leading to data loss).
I might be very wrong though.
AWS S3, as I understand it (100 % on top of my head, so this might be wrong as well) replicates each bucket across 3 avalibility zones. (Or, erase encodes..)
I think AWS S3 never lost a file as well (apart from human/billig issues, of course, considering only the drive replication).
My opinion is that if you’re backing up to S3, including but not limited to AWS, B2 or R2 (not counting every S3 compatible service, surely some are not secure or even replicated one bit), it’s fine as the ONLY off-shore backup, apart from say at-provider backup.
For example, if you’re got a web server at Hetzner, I’d consider their daily backups + any S3 mentioned above to be close to 100 % safe.
Issue is rather; how do you know if an S3 bucket has failed? Or if your card expired? Of if an invoice somehow isn’t getting paid? Especially if your email decide to reject the reminders or put them in junk.
Clearly, there has to be manual ”checking in” on the S3 of choice.
From the B2 link:
At the end of the day, the technical answer is “11 nines.” That’s 99.999999999%. Conceptually, if you store 1 million objects in B2 for 10 million years, you would expect to lose 1 file. There’s a higher likelihood of an asteroid destroying Earth within a million years, but that is something we’ll get to at the end of the post.
Edit:
Found this online:
Backblaze's main reliability system is erasure coding[4] of shards of files, which is not RAID[1][2]. RAID mirrors disks or volumes on block or filesystem layers, not files in userspace.
Don’t know if it’s true.
Anyway, is this method of splitting files across different drives less secure than full replication?
I know RAID isn’t backup; but are these 20 drives connected in a way where the ”tome” can become corrupt like in RAID or are these 20 drives on different systems?
I guess I’m just not very knowledgable on this really. How secure would you judge the S3 services to be?
Edit 2:
I know losing files is possible: https://www.quora.com/Does-Amazon-S3-use-SSD-If-not-then-why-not-Is-cost-the-issue
This was a bug, and obviously there’ll be bugs, what I’m wondering is if a file had ever been lost simply of drives failing faster than new one’s have taken over, as if, imagine no bugs, no billing issues, no humans involved at all.
It seems ANY solution even if it’s reolicated across all continenta in 10 drives in any continent still have a failure rate of more than 99.99999999 or whatever they claimed simply because of the humans managing it.
If I use AWS & Backblaze, the highest risk is probably me accidently removing the buckets, or my code somehow removing it’s content.
If you are planning this, please contact @yoursunny. He offers state of art South pole colocation, unmetered snow bandwidth, network with YOURSUNNY-AS(snow blend). I colo my server with him, the DC is highly secured with Bear's guarding the entrance and equipped with snow racks and cables. It's also my vacation spot when I want to visit DC to change some stuff.
Vouch for sunny, just sometimes I faced latency issues between North pole.
@emgh So Backblaze splits chunks of data across servers. Data itself isnt replicated, lost chunks can be rebuilt like in RAID-5.
Althrough this architecture can be durable, it doesn't fixes issues that replication solves - disasters like fire, fiber cuts etc.
So it is just redundant, not replicated.
Replicated at this price would be absolutely amazing.
I guess not even S3 with it’s ”replication across zones” isn’t even replication, but I can’t really explain why not.
Still, combining any of the above with a production envirement running in a whole other continent with backups in place, equals replication I guess.
Why not just use S3 + RClone?
As I suggested, but S3 is more expensive - so we’re discussing if it’s that much more reliable to justify it being that much more expensive.
Well, you could use any S3 compatible provider I assume.
Storj is $4 a TB for example.
Again, I was proposing S3 and I stand by it in the sense that it's great, but you can't forget that's close to 4x what a dedi would cost that would include bandwidth (at least to a degree).
If it's like a backup of a backup then it might not make sense, although if S3 can justify getting rid of other backups (or making them "less complete") maybe it can.
We can do any VM size since our storage infra is SAN
So we can do a VM with 100TB if needed
our VM pricing is 2$/TB for large deployments (ram and cpu are free) (BW equals your total storage per month)
The VM storage is raidz2
if you are looking for storage server you can check our orangutan dedicated servers with 3x 14TB disks for 95$/m
But if you need usable 50TB thats not enough as if you go with any raid you will lose 33% of that so you need to add more disks or you can go with
https://clients.servarica.com/store/black-friday-2022/orangutan-dedicated-server-non-raid-2
After having learnt a lot more: S3 Glacier Deep Archive
If you very very rarely need it, it's much safer than RAID, costs $0.00099 per GB .
Expensive to retrieve, but great for the backup of the backup of the backup.
Hetzner is the cheapest option, if you just need a place to store your data. You should go with them.