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.
★ VirMach ★ RYZEN ★ NVMe ★★ $8.88/YR- 384MB ★★ $21.85/YR- 2.5GB ★ Instant ★ Japan Pre-order ★ & More
This discussion has been closed.
Comments
So both TYOC039 and TYOC040 have the same problem?
Need to wait for the gen3 server is more likely?
Thanks. Forgot it was a domain. No wonder I can't find it. I was looking for a bare IP when I skimmed this thread.
MJJs, you will be in for a surprise in a few months' time.
Sorry, but I am one of the customers you're referring to, I purchased a VPS in Amsterdam, and I don't agree at all with you. It's clearly stated it's pre-order, it's clearly mentioned that delivery times are expectations.
For me, I'm on LET with the expectations of a stable VPS for a low price. That stability is much more important than an instant provisioning, and in cases that's important for me, I'm not applying for a VPS that is "pre-order" and stated clearly that the servers have yet to be shipped. We're living in 2022 - even though we are capable of doing lots of things right, I know shipping is a pain in the *ss these days. And I know that you almost always run into unexpected things when setting up an environment. So pre-order for me is "not needed very soon, but nice to have": if you want "need very soon", don't go for pre-order. Ever.
So, in my opinion, everybody complaining and pushing that they want their VPS is (again: personal opinion) not very fair. And pushing again and again is behaving like an a-hole. Because time spent on processing your requests is taking time away from solving the issues for everybody of us who are waiting for their VPS.
And compensation? It's already stated that while "pending" your time period of using the VPS is not started. So, compensation for what? For expectations you've created yourself, although @VirMach has been (compared to others I've dealt with in the past) very clear and open in communication?
Yes, I hope that shipments arrive soon, that the encountered problems get fixed, and that VPSes can be rolled out soon. But mainly because I think it's overly frustrating for @VirMach. And to everyone that keep complaining: realize that in my opinion this has been a heck of a deal for an amazing price - the more complaints there are, the less likely there will be a comparable offer again. Because it's not worth the hassle.
It's been quite some typing, but following this thread for the past couple of weeks, it's just something that had to come of my chest.
@VirMach:
Posted this locale change AMS>TKO on March 24, then posted the request again yesterday. Invoice #1398758
I understand what busy means. So if you got it, no hurries, no worries.
It's just sitting there in pending mode. With no reply, I guess it's just the same here.
I hadn't actually checked in WHMCS, I assumed I would of had more functionality through SolusVM. There looks to be templates available through WHMCS, and the Ubuntu 20.04 image at least seems to work, but none of the others that I have tested. There were no images available in solus.
However, I do not wish to use Ubuntu 20.04. I was hoping to try a custom ISO but its stuck on a message of 'We're working on this...' after adding the ISO
I can't receive the two-step authentication email. Can't log in to my account now. Can anyone tell me what to do?
ticket #552008
@VirMach
These aren't the test IPs, it's just from a post I made describing something else.
Both servers currently there, plus the 6 servers arriving tomorrow or Wednesday will have the same issue. At this time, I'm more confident that it's related to the motherboard's settings for communicating power levels for PCIe with the PCIe devices, and it not functioning properly on Linux.
These are some modified functions for PCIe 4 devices. I do believe the kernel was patched to ensure this was fixed for Gen3, but maybe not Gen4?
I've already applied a fix that has improved the situation. Now I'm looking for further kernel parameters to try to patch it manually for now. I've also contacted two of the NVMe SSD manufacturers, and the motherboard manufacturer, and I'm going to be doing some testing to see what the best course of action could be (for example, if Gen3 works on the motherboard, tested in our office, then Gen3 can be sent and NVMe can be replaced by DC hands.)
Man, I really need that Tokyo VPS now. AWS credit will expire in the next 36 hrs.
Yeh. Good luck with that... they are still ironing out issues...
You should switch to Vultr for hourly billing first
...or just top up a little bit more?
Kioxia seems like the domestic brand in Japan and I might be able to get it sent to the datacenter locally/quickly. Do you know if these are advertised to function on Linux? Or when you say you've had good luck with them, what OS?
The ones giving us trouble is XPG, not sure if I mentioned that already earlier. Also I'm not sure if I mention how much I hate them and how these were just purchased in relatively small quantity because we needed them quickly at the time and it's the only thing we could get without waiting a few weeks for that quantity.
kioxia has lower TBW than competition. why XPG and not sabrent rocket?
XPG wasn't really my top choice, it just happened for this small batch. I was trying to get Japan out ASAP at that time (but of course, everything got super delayed anyway so it was pointless.) This was so I could get Gen4 nodes out instead of using Gen3 and it hindsight since it's having problems, it was a bad idea.
We are mainly going to be using Samsung 970 Plus, SK Hynix Gold P31, WD SN750, Mushkin Pilot-E which we specifically got because they use the good SM2262EN controller and it was the only company at the time where I could verify in bulk that all of them would be this controller and not replaced with a cheaper alternative (all manufacturers were switching out controllers without notice around this time due to supply chain issues.) Then the rest are XPG, Inline Platinum, and a few other random brand 4TB I could get my hands on (at the time these were hard to get at an affordable price and my vendors were all sold out.) We did get a few Sabrent Rockets but that was a very long time ago, I think I misplaced most of them. The affordable ones for these at the time were QLC and I wanted to avoid that, the rest were super expensive because they were popular at the time. And then let's see... we have super minimal quantities of a lot of other drives as well, Kingston, PNY, Corsair, Intel, etc. Actually, we even have a few U.2 enterprise ones that are going to go up, I don't remember the brand.
I basically bought every single SKU SSD available and tested them all at some point in time because I wanted my own data, so some nodes will end up using those.
We've had great luck with their enterprise NVMe, not sure about their consumer facing stuff. Out of all the drives you listed, the Gold P31 is probably the best.
https://billing.virmach.com/serverstatus.php
"Affecting System - LAKVM26
03/15/2022 20:52 Last Updated 03/16/2022 22:42
Dear VirMach Customers,
We are currently investigating an unexpected outage with LAKVM26. We will update you as soon as we hear back from our datacenter.
Update 10:27PM 03/15/2022 - Currently it appears that the RAID controller or failed drives may be the issue. When rebooted the machine appears to boot into a MEGARAID drive missing screen. IPMI is not currently working so we are having a KVM attached so we can attempt to resolve IPMI and then the RAID array.
Update 2:20AM 03/16/2022 - It appears that 4 out of the 8 drives have failed in this machine. We have forwarded this issue to a systems administrator so we can find a solution to minimize data loss as much as possble.
Update 10:41 03/16/2022 - Unfortunately we were unable to repair the failed RAID array. We are currently evaluating the next best course of action for this.
Thanks,
VirMach Support"
No fix to date!
@VirMach
Actually, this reply moved me，I was moved by the boss's desire to kill in Asia!
So many pre-orders prove that the price of such a machine configuration in Asia is simply earth-shattering.
However, with so many complaints, they ignored this point, and even touted the Green Cloud. I bought 20 units in the Green Cloud pre-sale, ranging from $25, $35, $22, $55, and experienced in various regions. In fact, the cloud is also facing various problems of deploying AMD in Asia, and the waiting time is also very long. Why do they complain less?
Because they don't try to talk about technology with a group of people who don't understand server deployment, they just tell one startup time node after another. During this period, except for bullshit, appetite, and hype, they don't talk about any specific details of server deployment, so that some people even say vir The technology is not as good as Green Cloud, I don't know whose technology is better, but, VIRMACH has been the butcher of oversold prices for so many years, without technology? how can that be possible.
What I want to say is to do your own thing and let professional bullshit people come to the forum to bullshit. If you set the price of Asian routes so low, you must be prepared to face challenges. How can it be easy to be a leader?
By the way, I have 2 orders, if you think this advice or understanding is valuable, I think maybe one of them can be deployed in advance, thank you!
Order ID: 658382
Lol, such a fairly nice message. And then spoiled by the "please deploy mine in advance".
dude，Life needs a sense of humor
Hey bro, I just wanted to ask when will Tokyo be ready to start? I purchased it on the 12th and supposedly some people have already opened it, is it possible to give an exact date?
I remember Biennial pre-orders in Tokyo will boot first, am I wrong?
https://lowendtalk.com/discussion/comment/3399238/#Comment_3399238
Probably not earlier than next week if VirMach can't find the issue for those gen4 magic problems.
I think I changed my mind on TBW.
It's not really indicative of anything. Yes, theoretically, and I would probably originally argue this, if they advertise higher, then they clearly are confident in their product and it will last longer. But the more I looked into it, there's plenty of shoddy ones that have 5 year and really high TBW. Except they just make the entire RMA process hell. Some will ignore you, some want you to send it directly to another country and take weeks to months, and reject it for whatever reason, and most importantly, a lot of them count on you just not cashing in the warranty so they'll advertise whatever.
In the end TBW doesn't mean anything other than when the warranty runs out.
MTBF is the one that means something, it's the mean time to failure, but pretty much all of them are the same there.
(edit) I must say though that of course different specific controllers combined with different specific DRAM and NAND/V-NAND will definitely have different lifespans but the only way to truly know is to look at the advertised life of those components specifically, which is difficult to find and kept behind closed doors in a lot of cases.
Only solution at this point is data wipe and re-install if you are in a hurry, give me the ticket ID for that request if that's what you want. We're still trying to avoid that, and that sometimes takes weeks for shipping things around and coordinating with the datacenter. This situation happens once every few years, and this is one of them.
The RAID controller failed in conjunction with at least one drive, and caused a chain reaction.
then kioxia should be a pretty good choice as i recall apart from TBW their reviewed performance was quite decent.
i personally got a SN750 SE due to the assumed ease of RMA if ever required for a cheap dramless Gen4.
It's a pity that you don't do marketing. Tell me about your experience and work progress every day. It's very good. We'll watch your performance here.
It seems that after so many days, the RYZEN thing has not reached any progress, everything is back to the original position, guess I will see you guys in May!Hopy it will be ready then!
Thanks for your reply, your order has been canceled, and the refund will be ready within 3 business days.