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
Here's the first 10 so far.
I never liked these things but I guess it can be quite enjoyable and easy when it involves your most utilized words! You just have to look for [a] all synonyms of maybe, [b] the systems and hardware you've been complaining about, [c] add in a bunch of random words I definitely didn't say but exist there to pretend like I finished it.
(edit) [d] randomly find a few more words and edit it in.
My 2.5 GB VM is still not provisioned and I've been waiting a while for that, so I'm not sure how accurate is that "100%". I'd guess I'm on "manual review", since it was my first order with VirMach?
I'm still silently waiting though, I just have no idea with the status of my service.
You might want to quote the rest of the message you quoted.
Ah, so I might have misunderstood originally.
Does it mean then that there's still ~200 VMs unprovisioned because they were thought to be, but were instead duplicates of already provisioned VMs?
I guess I can't read right now.
Correct, in terms of the estimation. So our estimates were off by a little bit. A lot of these also happened with the re-creation script to migrate so they were recent, and now it's taking up room that could have otherwise been used. Just updating for the reason for the last 200 taking longer.
Otherwise we would have been at 100% but we're not at 100%
Found most of them. Some of them like TYOC040 are impossible, unless you read his posts.
@VirMach: Understandable, thank you for the clarification. Time to wait a bit more then
A lot of the shitty SSD companies also rate TBW this way, make a drive that can only really do 400TBW and then rate it for 1600, but have a RMA system so bad that most people will just give up. Adata is known for switching around NAND/controllers so it wouldn't shock me if the same thing happened here. Their firmware has also never been too great, that might be what's causing some of the issues since it's the only thing they really directly control. Hope you guys end up buying more of the P31 to replace those adata drives, I really like my VM on that ssd.
tl;dr, wait.
This pot, Google translate to memorize
Yes, this 100% caused the issue. We do have other drives from Mushkin with similar but less severe issue and the kernel fix we had for them worked for both, but Mushkin is actually working with us to push out a firmware update. They've been very responsive, I've just been the one too busy to reply to their engineer's latest requests.
Yes we already have excess disks. We had excess Mushkin Gen4 as well as excess P31, Samsung, and Mushkin for the Gen3. Luckily I suck at keeping inventory so I definitely bought too many. We also bought about 50% more than we needed to start with just to ship out servers with 6TB instead of 4TB even though our average usage per node would be closer to 4TB so we're good there as well. This leaves most nodes with extra space should something go wrong. As in, they won't have a disk bottleneck.
ISSUE! How did I miss that one, it's all I've been having for a week.
I am surprised and happy that you enjoy playing it
There's only 2 more missing and I am very impressed by how fast you completed it 
For anyone interested, here's the solution (I made a small mistake but you get the idea):
Great job!
Yup that's why this is just for fun, not meant to be an actual English word search 
Thanks for playing it
Are you still processing the paid Ryzen preorder from the last year? Mine is still on the "complementary server" and not even changed after that.
My tokyo biennial didn't make either batch, still pending, but then I'm not unhappy or in any rush, if I was I wouldn't have bought a preorder. I ended up cancelling a yearly preorder to make the biennial to make sure I got on the newer Ryzen, I hope I didn't get bumped to the back as a problematic customer, I really try not to be at least
The entire pre -sale process of Tokyo: Start pre -sale − Order accessories ~ accessories issued ~ passing the Pacific ～ Pacific Wind Large ~ Customs Inspection Delay ~ Customs Detention ~ Show the delay of the document ~ Arrived in Tokyo ~ third -party configuration fault ~ Continue to buy accessories ~ Panel ~ Panel ~ Panel Bug ~ delay repair ~ Sunday delay ... ... Pired = refund, refund = boot，
Does this mean that there are still 200 people left to activate for the March Tokyo pre-sale
Or this mean that after activating the remaining 200 copies there are still 2300+ people left?
my 2560 is activated on TYOC038, but offline, please help me
Dedicated servers? No progress. In fact, we've regressed. We just couldn't do it. We needed the chassis for these Ryzen servers so although many of them were done, I had to gut them for the chassis. There's a huge chassis shortage it seems, at least the ones we're meant to be using because of the whole supply chain issue and increased shipping costs ensures that it's difficult to get them within budget either way. They also quite literally doubled in price.
The philosophy was unfortunately it's better to have one customer on an E3 and migrate that late and upset the customer versus upset 100-200 people on a node it could become instead.
I will still have some extras, but they'll be the ones that have rails that are difficult to purchase. I've found some universal rails that work with them though but this means we have to move all the motherboards and drives from the ones we're using for Ryzen, and move them into these other chassis.
Just unlucky with the way we ended up automating these, it created the smaller plans first.
There's barely any of these left compared to the total quantity activated though so I'm hoping they get activated soon™
2300~ activated but probably closed to 2100~ because 200~ of them are duplicates we need to clean up. And 200 left.
ADATA is shit,I am sure.
All services outside of maybe 5 of them have been recreated for SJCZ003 and SJCZ007 issue. Most TYOC038 have been recreated. Some still need to be created if they never created properly in the first place on TYOC038. And SEAZ001 is still having issues, I'm trying to look at that again next.
Atlanta server!! yahoooooo!!
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Yet-Another-Bench-Script
v2022-02-18
https://github.com/masonr/yet-another-bench-script
## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ## ##
Sun Apr 24 01:08:46 EDT 2022
Basic System Information:
Processor : AMD Ryzen 9 3950X 16-Core Processor
CPU cores : 2 @ 3493.436 MHz
AES-NI : ✔ Enabled
VM-x/AMD-V : ✔ Enabled
RAM : 2.3 GiB
Swap : 256.0 MiB
Disk : 49.0 GiB
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) | 850 Mbits/sec | 271 Mbits/sec
Online.net | Paris, FR (10G) | 776 Mbits/sec | 263 Mbits/sec
WorldStream | The Netherlands (10G) | 865 Mbits/sec | 297 Mbits/sec
WebHorizon | Singapore (400M) | busy | busy
Clouvider | NYC, NY, US (10G) | 933 Mbits/sec | 935 Mbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 926 Mbits/sec | 805 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 841 Mbits/sec | 477 Mbits/sec
Iveloz Telecom | Sao Paulo, BR (2G) | busy | busy
I'm going to do an inventory of all servers out in the wild with an XPG drive and [A] move everyone off the disks ASAP in an emergency migration, and [B] remove them from the active pool until we can get the XPG drive out and incinerated.
SEAZ001 has scheduled maintenance right now with DC hands to remove the drive and place it on a brand new server so I can try to move as much data off it as I can before it fully breaks (if that hasn't already happened.)
TYOC038 will be after these two, since the people still facing issues never had a functional service to begin with (unless I made an error) then we're considering this less urgent. You'll of course have your due date reset.
Are they still working? Sounds like you need recycling at a reasonable price. Let me know, I knew a branch of PC players who can accept used hardware's.
You can blame their firmware or implementation or things like that but not the NAND. The NANDs are all sourced from the same few brands and they are not stupid enough to sell "bad" NANDs. Also, what really gives them the nice benchmark speeds is by using the unused space as SLC cache so having a mix of good and bad NANDs wouldn't work out anyway.
When your SEAZ001 VPS had issue do you also see "Could not read from boot disk"? That's what I am seeing right now after I saw Virmach's update.
@VirMach
I have tried to replace ISO (mount/unmount) and install template for SJCZ002 and mine still does not work, I'll open a ticket on this issue.
Also, DALZ003 is acting weird. The service is online and I can connect to VNC, however it doesn't look like the IP is routed (both inbound and outbound). On the service via VNC, everything is "No route to host". Used Ubuntu 20.04 and AlamaLinux (both installed via template) and also tried to reconfigure network.
just leave DALZ003 i tried to migrate to same location and change ip, now i have 2 ips and still broken, looks like auto script that created VPS did not connect ip4 to VPS