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.
Comments
It'll just be an experimental offshoot, a few servers that are sold with guaranteed hardware/performance bump or just something crazy. I've really been wanting to do a massive NVMe storage VPS for a while and everything was lining up for it before all the hardware shortages.
Picture something like:
8GB RAM @ 3200MHz
1TB NVMe (RAID10 @ 256Gbps Controller)
2 Dedicated vCORE @ 4GHz
@ $30-40/MO.
Forgot to post this a day or two ago:
Update on the servers and the heatsinks... Amazon did something really weird which they haven't over the last decade. For some reason they thought it was suspicious I was buying so many NVMe heatsinks from so many different brands/sellers I guess and completely nuked the account and then brought it back, but all the orders got cancelled. I re-ordered them but no chance that they'll actually make it here Sunday anymore, it's looking to be pushed into Monday/Tuesday.
And an actual update to the update:
Everything's here. Doing a lot of testing today then ordering the best one for all of them and still on track other than the few days' delay. Actually potentially able to ship more than the quantities I mentioned above since more parts are coming in.
Fucking Amazon, lol.
DO NOT STEAL HEATSINKS DESIGNS, WE MUST PROTECT OUR SELLERS.
Update --
Testing heatsinks is taking too much time and I really don't want to keep delaying things. At least for Los Angeles, I'm going to slap in a few different heatsinks and it'll just become part of the beta test. Since it's in Los Angeles and we're here as well, it'll be pretty easy to go in and swap them out.
Going to finalize and take a few of them to the datacenter by tomorrow, fingers crossed.
GOOD LUCK!
Going to Psychz once traffic dies down, wish me luck. That was some pretty bad traffic... somehow I've been stuck in it for an entire extra day!
Sounds like the time, back in the last century, I decided it was a good idea to drive to Las Vegas for New Years. Left Hollywood Hills at 10 am on the 31st and ended up ringing in the New Year around Barstow.
I Should have listened to Wild Man when he shouted as we drove away
"You need to be in Vegas like a fish needs a bicycle".
Wise words indeed.
And I thought it was bad coming back after New Years, only 11 hours.
Still stuck in traffic, had to fix a few servers on the way and I'm in the middle of scheduling maintenance for others and ensuring we're ready for some migrations Monday. It's difficult to drive while I'm on my work laptop, I'm sure I'll make it tonight sometime.
Hello. How is everybody doing?
16GB RAM @ 3200MHz
300GB NVMe (RAID10 @ 256Gbps Controller)
3 Dedicated vCORE @ 4GHz
@ $30/m. Enter VIRMACHCLUB to get 50% lifetime discount. Final price $15/m.
Hey guys, about to post our first sale in a while. Wish us luck!
We had Amsterdam's power go out today, and dealing with other outages/problems. I did confirm with Psychz we'll be dropping the servers off tonight and they're configuring networking as well. I'm going to post the sale and monitor it for a while and keep an eye out for tickets and resolve the remaining outages, and then hopefully drop off the servers before traffic starts back up (6AM~)
Happy 4th of July.
Good luck!
The above post ^^ was the 13,000th post in this thread.
Hello! I'm fine, I hope the rest is fine too.
I need one more fictional prize then.
Phoenix and Frankfurt are coming back, unless everything goes catastrophically. Just updating everyone and letting you know we're about to pay for colocation space in these two locations.
Great news on Frankfurt returning. Thank you.
I wanted to know if it would be an advantage to you (as well as me) if we were to combine multiple smaller VPSes in the same/near location into one larger VPS to reduce IPs.
example:
In LAX - a 2 core 2GB, 10GB @ $8.79 + a 2 core, 2GB, 50GB @ $21
into 3 core, 4GB, 60GB @ $29.79
In Amsterdam - a 2 core 2GB, 15GB @ $10 + a 1 core, 386MB, 17GB @ $3
into a 2 core 2.5GB, 30GB @ $13
In NJ - a piscataway 2 core, 1.4GB, 25GB @ $13.84 + a chicago 1 core, 512mb, 5GB @ $4.70
into a NYC/NJ 2 core, 1.9/2 GB, 30GB @ $18.54
I can give back some of my IPs too... and VPSes too tbh
We do get a lot of these requests. For many reasons we say no, but for the reason you just outlined I think if we could actually build an automated system for this, it would be worth it.
It would just probably be a lot of coding, a lot that could go wrong, and I'm sure someone would figure out a way to buy specific services from others and combine all their services into one giant service and get upset they can't burst to 88 cores on a Ryzen 3900X. It's definitely something we've considered though, especially with those tiny specials that have no disk space.
The likelihood of this happening is still realistically low though based on what we already have on our plate, we'd most likely rather eat the IPv4 costs.
Give back a VPS, and take three in Buffalo.
Migrations today took way way way longer than we anticipated, and I slept longer than 2 hours last night by accident. I can't have this meme of us never delivering continue though, I'm going to the datacenter tonight, it might be past midnight unless literally more servers start going offline for no reason. It still counts as the weekend (ish.)
How to get infinite VPS:
I'll give back one in LAX for one three times larger in Buffalo for the same price
/j
Would be neat if VirMach started a cloud, and our VPS could be liquidated into resource pools.
If this ever happen, send me a message, I will give a call to Lisa, so they can make a special "3900X" for you.
Something like Lunanode stuff.
CloudMachines.
RYZEN!?
We've had the domain for a sub-brand for a while now. We might actually do something with Fleio, I still haven't had time to fully demo their panel but we've considered OpenStack for Windows since we can do Hyper-V and Hyper-V supposedly runs a little better, although I'm not looking forward to running Windows on the actual node.
I've angered the gods. For some reason they really don't want the R word being mentioned. Every time I do, everything that could possibly go wrong goes wrong.
Since telling you guys that I was ready to drop these off over the weekend, the following has happened and kept me very, very, very busy:
@VirMach Congrat for writing this post while LET is being DDOSed
Hi
I just had to pause the attack for a few seconds to copy/paste the post. Only kidding, not trying to be GVH.
Spent most of today dealing with LAKVM24 which was starting to look very bad. Luckily, we were able to get some function restored and we're now migrating data to another more stable (fingers crossed) server.
I'm pretty happy about that, we still have one other fragment of mess less over from the hectic week but this actually helps a lot and I think I should be able to focus on Los Angeles Ryzen now, barring anything crazier from happening. I actually had some time to diagnose the storage node as well. I believe all the RAM on it got fried alongside the previous motherboard. I'm ordering new sticks, and at this point we have 3 motherboards for that so hopefully one of them will work (and we do have around 6 x 16TB extra disks.) I also went ahead and ordered some E5-2690v4's to use instead of the v3's and we'll just keep the v3's in case of emergency or for the next storage node. We also have two RAID controllers and pretty much double everything else for it in case of emergency outside of the chassis melting down. I'm going to see if we can incorporate some NVMe LVM caching into the RAID 10 HDD build.
Full specs for anyone interested:
Dual E5-2690v4 (28 cores, 56 threads) @2.6GHz
256GB+ 2400MHz DDR4 Memory
36 x 16TB Seagate Exos X16 HDDs
MegaRAID 9460 RAID Controller (RAID 10)
Highpoint SSD7202 (?*) w/ 2x 4TB NVMe SSD
*Maybe, haven't decided or tested yet.
🤞