@VirMach said:
I had a TLDR written for this but lost it. Basically I have an idea to let existing customers burn in new nodes we deploy so we find any issues we missed that may appear from "natural" usage. It's essentially a 14 day trial service of NVMe1G and it may otherwise go down or have issues, no support, guaranteed data loss. You can abuse it all you want though as long as you're not committing a crime (as in, high CPU or I/O is fine. But we can still suspend it if we want to, not that it matters.)
I got "You must have an active product/service to use this code" error when I applied the Promo code so it did not remove the setup fee. Was logged in to my account and I do have an active service
Me too!
Does this mean the node assignments were filled or just some issue with the site?
@AlwaysSkint said:
^Yup, I concur. I couldn't get netboot.xyz to play nicely, so booted Alama and flags are set.
Just FYI, you can use netboot.xyz from grub pretty easily. Just follow these steps:
apt update && apt install grub-imageboot
mkdir /boot/images
cd /boot/images
wget https://boot.netboot.xyz/ipxe/netboot.xyz.iso
update-grub2
(if you're not quick, edit GRUB_TIMEOUT in /etc/default/grub to 30 or higher, run update-grub2 after editing)
Then just reboot and select netboot from the boot menu (you'll need VNC access for this bit)
Now you do know that you don't really need that and I'll take it off your hands; find it a new home on the East Coast (or NL) where it won't get abused.
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
Now you do know that you don't really need that and I'll take it off your hands; find it a new home on the East Coast (or NL) where it won't get abused.
@Virmach Perhaps you can ask the mods to create a forum post specifically for reporting issues with the NVMe alpha/beta test. Not much point in adding to your Ticket count, eh?
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
@AlwaysSkint said:
Gateway is pingable but Solus HTML5 VNC doesn't attach to node and VMS doesn't boot up (to netboot.xyz ISO).
Same problem, I also noticed that client panel says AMSD030X but its RYZE.AMS-D001.VMS in Solus. I'm not sure if numbers should match but they usually do
It just means I can't type. (edit) or think.
I think it must the first one since that package has only 1GB of bandwidth. You cannot think anyone could proceed a normal testing with such a abnormal amount of data transfer
This is a glitch that ModulesGarden was asked to fix for probably years and kept saying it's fixed and didn't fix it. Maybe they did fix it by now but I'm pretty sure we're on the latest version already. I'll fix it now, order another one in a few minutes.
@VirMach said:
I had a TLDR written for this but lost it. Basically I have an idea to let existing customers burn in new nodes we deploy so we find any issues we missed that may appear from "natural" usage.
Ordered with Debian 11. The order process was fine. Was provisioned an 1CPU, 1GB RAM, 25GB disk, 1GB traffic on a AMS node. Was marked as offline. Went into the control panel and tried to boot it. Nothing. Re-installed with Debian 11. Still not coming up. Tried Debian 10. Not still offline. Tried VNC, unreachable. Tried remote echo request (PING) to make sure it wasn't being misreported by control panel, no response. Tried CentOS 8, still not up. Tried a couple power offs, boots, and what not.
@lemoncube said:
I have a tiny VPS on PHXZ004 idling all the time just triggered an email imforming me it was suspended due to bandwidth overuse. It was a BF flash deal plan and the bandwidth actually is as low as 125GB if I remember correctly so it could have fallen victim to some sort of attack and got it all burnt up ( or maybe the configuration messed up during migration or IP switching and I didn't pay attention ).
I don't think I'm gonna appeal for it. It's just kinda infuriating not be able to inspect how or when it ran out of bandwidth as either whcms or solusvm won't allow me to see stats and graph at all when service isn't active.
Yeah I agree with you there. We really need to consider using something else soon as our next big project but I'd rather just leave everything alone at this point. I feel like us and the customers alike have had a decade's worth of frustration from everything flying around this year.
But pretty much every case is what you described, it might have gotten hacked. The only way to be nearly 100% secure is to use SSH key and disable VNC most the time but I understand that's not convenient for most people.
I think the big providers qualify new hardware configs for months before opening them up to non-beta users. 14 day burn in is for new installs of hw configs that have already been qualified. It sounds like the hw issues in this thread came from configs that were flaky to begin with, so 14 days wouldn't be nearly enough of a burn-in.
I certainly hope that all these emergencies resolve and things get near normal again. At that point it's worth doing some reflection to figure out what went wrong and how it can be prevented going forward. Postmortems are a usual procedure after any significant outage.
SJCZ005 is coming back, it was about to come back this morning and then I got dragged into something else for several hours.
We've definitely had way longer than months of testing for the configurations, but definitely not every single unit. They're all brand new parts and some just end up malfunctioning after people actually go on them. Based on our experience so far I think 2 weeks is pretty good. Most nodes didn't crash or have problems until we finally filled them, and the ones that did showed their problems relatively quickly. We do actually test everything though beforehand, it's just not the same. Like a node can do a stress test for several days and then once it gets the right combination of people actually using it or after a rough shipment it can still pass everything with flying colors and then go haywire once it has to actually boot up a bunch of VMs doing random things at random times.
@VirMach said:
I had a TLDR written for this but lost it. Basically I have an idea to let existing customers burn in new nodes we deploy so we find any issues we missed that may appear from "natural" usage. It's essentially a 14 day trial service of NVMe1G and it may otherwise go down or have issues, no support, guaranteed data loss. You can abuse it all you want though as long as you're not committing a crime (as in, high CPU or I/O is fine. But we can still suspend it if we want to, not that it matters.)
I got "You must have an active product/service to use this code" error when I applied the Promo code so it did not remove the setup fee. Was logged in to my account and I do have an active service
Me too!
Does this mean the node assignments were filled or just some issue with the site?
I have never done this before but my plan is setup VPN in one node, so others can connect to it and use that as the "external" IP address with the HA cluster setup
@VirMach said:
I had a TLDR written for this but lost it. Basically I have an idea to let existing customers burn in new nodes we deploy so we find any issues we missed that may appear from "natural" usage. It's essentially a 14 day trial service of NVMe1G and it may otherwise go down or have issues, no support, guaranteed data loss. You can abuse it all you want though as long as you're not committing a crime (as in, high CPU or I/O is fine. But we can still suspend it if we want to, not that it matters.)
I got "You must have an active product/service to use this code" error when I applied the Promo code so it did not remove the setup fee. Was logged in to my account and I do have an active service
Me too!
Does this mean the node assignments were filled or just some issue with the site?
Some old WHMCS bug IIRC.
I've tried switching browsers, adding "&promocode=IUNDERSTANDWHATIMORDERING" to the URL but still no joy.
Geekbench 5 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1385
Multi Core | 1334
To simulate real world client usage I have installed a VPN, a miner, H@H, and a small website. It is good that VirMach let us have so much swap RAM, I mean disk space.
@FrankZ said:
Well you could pretend you have HA.
(Always good to see you around.)
That's so nice of you, good to see you around too I have been focusing on work, relationship and family lately, dont really spend time on social much.
With the recession coming, I am also trying to plan my investment as all of them has been performing really badly, might want to start cutting my expenses too
Comments
Me too!
Does this mean the node assignments were filled or just some issue with the site?
Just FYI, you can use netboot.xyz from grub pretty easily. Just follow these steps:
Then just reboot and select netboot from the boot menu (you'll need VNC access for this bit)
Precise. Absolut.
Now you do know that you don't really need that and I'll take it off your hands; find it a new home on the East Coast (or NL) where it won't get abused.
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
Like ones with pending court cases about previous related companies that scammed their own userbase? Nah
I'll never give up my $25/year idler
--
So many years ago, way back in 2021
LES • About • Donate • Rules • Support
I have apple pie at home. I want it in the office.
There is strawberry jelly too.
You put strawberry jelly on apple pie?
@imok Can you PLEASE keep this inane stuff to Cest Pit and/or OGF? PLEASE!
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
@Virmach Perhaps you can ask the mods to create a forum post specifically for reporting issues with the NVMe alpha/beta test. Not much point in adding to your Ticket count, eh?
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
This is a glitch that ModulesGarden was asked to fix for probably years and kept saying it's fixed and didn't fix it. Maybe they did fix it by now but I'm pretty sure we're on the latest version already. I'll fix it now, order another one in a few minutes.
I feel bad for grabbing one of the free instances, I'll probably just run archiveteam on it + yabs on cron or something.
Sounds like you should also order another one.
Yeah I agree with you there. We really need to consider using something else soon as our next big project but I'd rather just leave everything alone at this point. I feel like us and the customers alike have had a decade's worth of frustration from everything flying around this year.
But pretty much every case is what you described, it might have gotten hacked. The only way to be nearly 100% secure is to use SSH key and disable VNC most the time but I understand that's not convenient for most people.
Did you name this VM especially for me?
FYI: Still the discrepancy AMSD030X => RYZE.AMS-D001.VMS
That's fixed!
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
SJCZ005 is coming back, it was about to come back this morning and then I got dragged into something else for several hours.
We've definitely had way longer than months of testing for the configurations, but definitely not every single unit. They're all brand new parts and some just end up malfunctioning after people actually go on them. Based on our experience so far I think 2 weeks is pretty good. Most nodes didn't crash or have problems until we finally filled them, and the ones that did showed their problems relatively quickly. We do actually test everything though beforehand, it's just not the same. Like a node can do a stress test for several days and then once it gets the right combination of people actually using it or after a rough shipment it can still pass everything with flying colors and then go haywire once it has to actually boot up a bunch of VMs doing random things at random times.
I don't know, I'll stop rambling I guess.
Oh god my brain is actually melting, I could have sworn I fixed it but it's entirely possible I can no longer type 001 correctly.
If it makes you feel any better this node has just been sitting there idle for ~3 months now.
Some old WHMCS bug IIRC.
Now just add a 2TB HDD or therabouts and we've got the making of a sweet 15GB NVme/150GB HDD hybrid.
Then unleash on the Tokyo suckers punters! 
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
oh yeah just gonna triple YABS it
I bench YABS 24/7/365 unless it's a leap year.
I'll have more of this, thanks..
Mon 3 Oct 19:23:28 EDT 2022
Basic System Information:
Uptime : 0 days, 0 hours, 7 minutes
Processor : AMD Ryzen 9 5900X 12-Core Processor
CPU cores : 1 @ 3693.062 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 804.7 MiB
Swap : 256.0 MiB
Disk : 24.3 GiB
Distro : AlmaLinux 8.6 (Sky Tiger)
Kernel : 4.18.0-372.26.1.el8_6.x86_64
fio Disk Speed Tests (Mixed R/W 50/50):
iperf3 Network Speed Tests (IPv4):
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 939 Mbits/sec | 940 Mbits/sec
Online.net | Paris, FR (10G) | 936 Mbits/sec | 936 Mbits/sec
Hybula | The Netherlands (40G) | 941 Mbits/sec | 941 Mbits/sec
Uztelecom | Tashkent, UZ (10G) | 861 Mbits/sec | 459 Mbits/sec
Clouvider | NYC, NY, US (10G) | 876 Mbits/sec | 350 Mbits/sec
Clouvider | Dallas, TX, US (10G) | 716 Mbits/sec | 178 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 795 Mbits/sec | 241 Mbits/sec
Edit: Sorry, @yoursunny I haven't swotted up on IPv6, to enable a route but the vNIC appears to have one.
It wisnae me! A big boy done it and ran away.
NVMe2G for life! until death (the end is nigh)
I am going to try obtaining multiple of them and create a Proxmox HA cluster, let's see how well it does
This is acceptable use case for IO/CPU "abuse" right?
"Humanity is f*cked up" - Jay
Obligatory YABS
I have never done this before but my plan is setup VPN in one node, so others can connect to it and use that as the "external" IP address with the HA cluster setup
"Humanity is f*cked up" - Jay
@FAT32 Please install Alma8 Linux template first to get AES-NI, VM-x/AMD-V enabled. Then install Proxmox .
LES • About • Donate • Rules • Support
I've tried switching browsers, adding "&promocode=IUNDERSTANDWHATIMORDERING" to the URL but still no joy.
So is everyone else just paying the setup fee?
Just realise all of them are in the same host node, probably defeat the purpose of HA
"Humanity is f*cked up" - Jay
Well you could pretend you have HA.
(Always good to see you around.)
YABS with IPv6 for @yoursunny
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
# Yet-Another-Bench-Script #
# v2022-08-20 #
# https://github.com/masonr/yet-another-bench-script #
# ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## #
Mon Oct 3 19:32:56 CDT 2022
Basic System Information:
---------------------------------
Uptime : 0 days, 0 hours, 5 minutes
Processor : AMD Ryzen 9 5900X 12-Core Processor
CPU cores : 1 @ 3693.062 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 964.7 MiB
Swap : 4564.0 MiB
Disk : 24.3 GiB
Distro : AlmaLinux 8.6 (Sky Tiger)
Kernel : 4.18.0-372.26.1.el8_6.x86_64
fio Disk Speed Tests (Mixed R/W 50/50):
---------------------------------
Block Size | 4k (IOPS) | 64k (IOPS)
------ | --- ---- | ---- ----
Read | 377.29 MB/s (94.3k) | 353.85 MB/s (5.5k)
Write | 378.29 MB/s (94.5k) | 355.71 MB/s (5.5k)
Total | 755.59 MB/s (188.8k) | 709.57 MB/s (11.0k)
| |
Block Size | 512k (IOPS) | 1m (IOPS)
------ | --- ---- | ---- ----
Read | 791.81 MB/s (1.5k) | 2.13 GB/s (2.0k)
Write | 833.88 MB/s (1.6k) | 2.28 GB/s (2.2k)
Total | 1.62 GB/s (3.1k) | 4.41 GB/s (4.3k)
iperf3 Network Speed Tests (IPv4):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 940 Mbits/sec | 940 Mbits/sec
Online.net | Paris, FR (10G) | 937 Mbits/sec | 929 Mbits/sec
Hybula | The Netherlands (40G) | 941 Mbits/sec | 941 Mbits/sec
Uztelecom | Tashkent, UZ (10G) | 822 Mbits/sec | 417 Mbits/sec
Clouvider | NYC, NY, US (10G) | 876 Mbits/sec | 378 Mbits/sec
Clouvider | Dallas, TX, US (10G) | 813 Mbits/sec | 190 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 767 Mbits/sec | 219 Mbits/sec
iperf3 Network Speed Tests (IPv6):
---------------------------------
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 743 Mbits/sec | 306 Mbits/sec
Online.net | Paris, FR (10G) | busy | 518 Mbits/sec
Hybula | The Netherlands (40G) | 831 Mbits/sec | 833 Mbits/sec
Uztelecom | Tashkent, UZ (10G) | 752 Mbits/sec | 77.1 Mbits/sec
Clouvider | NYC, NY, US (10G) | 574 Mbits/sec | 87.9 Mbits/sec
Clouvider | Dallas, TX, US (10G) | 311 Mbits/sec | 113 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 593 Mbits/sec | 85.1 Mbits/sec
Geekbench 5 Benchmark Test:
---------------------------------
Test | Value
|
Single Core | 1385
Multi Core | 1334
To simulate real world client usage I have installed a VPN, a miner, H@H, and a small website. It is good that VirMach let us have so much swap RAM, I mean disk space.
LES • About • Donate • Rules • Support
That's so nice of you, good to see you around too
I have been focusing on work, relationship and family lately, dont really spend time on social much.
With the recession coming, I am also trying to plan my investment as all of them has been performing really badly, might want to start cutting my expenses too
"Humanity is f*cked up" - Jay