All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Any information on QuickPod? 483GB of ram for $14.4/Month
Grentenville
Member
Yes, you read that right. I thought the deals here were absolutely insane, but no, I was wrong. We got places with even BETTER deals
I have never heard of this provider before. I’ve never seen a single mention of it anywhere. It just... appeared. Like an online shopping site that suddenly knows your darkest purchasing desires.
Naturally I had to check it out, I saw a masterpiece, a glorious "pod" , I knew I had to get one immediately.
24 cores of a Xeon Gold 6326
483GB of pure RAM
30GB of storage
Networks speeds of 732.59Mbps up and 7571.76Mbps down
And for the low low price of $0.02/h, I had to try it!
Topped my account with the minimum amount of credit, a single, mighty buck, and I couldn't be happier.
Entering the pod and I was impressed, Ubuntu is heavy! My pod was using a total of 206GB of memory, I was deploying a small nation's worth of memory-hungry services!
total used free shared buff/cache available
Mem: 483Gi 206Gi 15Gi 188Mi 266Gi 277Gi
Swap: 127Gi 38Gi 89Gi
Even a mere YABS couldn't run properly as to how heavy modern operating systems are getting, I could only use ~4-6 cores! (YABS Link )
This isn't a post recommending the provider or something like this, I just found it "interesting" and wanted to take a look, also I'm curious if you guys know anything about this provider.
If this post violates the laws of LET, may the mods smite it down.
Comments
debian thx
There's another website like this one that is a "marketplace" service that's advertised here recently. They're just a broken basically and a third party company is the one providing the resources. Does it say who the actual end-provider is or are you only able to see the network provider via IP info?
Either way, $144/m (0.02/h) for 483GB RAM isn't bad 🤣
I thought op was trolling... maybe an error in amount of ram.
It is true, but the offer is occupied =P
Yeah, there also is vast.ai and salad, that I know of, that offers services in a similar way.
And no, there isn't any mention about the end-provider, I'm only able to get via the IP (Mine in particular being AS397423, tier.net)
What... did they find some secret technology that can display MB as GB in Linux?
That's impressive for $14/month!
One question though. The initial description had 30GB of storage, YABS showed 100 GB and the screenshot from @Nanja showed 2,570 GB. What disk space is expected?
oh sorry, I was looking at the offers with Xeon Gold 6326 like OP mentioned.
They have 3 offers, only one had 483GB ram, while the others had 40GB ram.
This might not be the plan that op has, they probably hid it or removed it or something else.
And that may be my machine! At least the host and mach (However I don't know what this means) match.
There are also other interesting offers, I was able to find this one just now, with 1TB of ram for the same price ($0.02/h), however, I can't start any other pod to check it out as my balance is below $1:
https://imgur.com/Ir6DjGI
Just saw they had a supposed 30 something+ epyc cores and 1000GB of ram in Germany for $0.02/hr but it vanished in front of my eyes after about 30 seconds, i hope it was one of you that bought it because something weird is up with this site/company
edit: it was 256 cores as per the screenshot from @Grentenville https://i.imgur.com/Ir6DjGI.png
yeah this shit is a scam
The storage that appears on the page may be the max that you can get/the machine has, when creating a pod you select how much storage you want, and you are billed for it separately from the hardware, on my case it was $0.007/h for 30GB
Here is the output to a df -h on my pod:
Looks like those are containers rather than KVM virtualizations, something with OverlayFS.
Try to use the whole resources for an X period of time
So I made some tests with stress-ng:
stress-ng --matrix 24 -t 30m --times --vm 4 --vm-bytes 100% --vm-keepThis morning I was able to allocate ~28GB per VM process (~112GB total)
Now I can't allocate more than 60GB total
Final stats of the 30 minutes run:
Are there any specific tests or ways to stress/use the resources that you'd like for me to try?
Probably you pay them with your data