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
TYOC029 is facing some CPU lockups. I'm going to send an email, apply a fix, and boot into BIOS quickly and apply a final fix and then reboot. I'll try to give you guys some further warning than just an immediate shutdown as it seems like it's recovering from the soft lockups and the server is still functional.
Funny enough, it seems like all the old templates work just fine on at least 3000 series CPU setups because we've already done it with some of our oldest servers PHXKVM1 and PHXKVM2 to Ryzen, a long time ago.
I think it just doesn't work with new installs because of new versions of libvirt and QEMU maybe.
**Los Angeles* -- I have IPMI set up now, setting up servers now.
NYC Metro - Most of the maintenance adding in new non-XPG SSDs was successful. Setting up more servers now.
The fix for this is complete, I'm not sure if it will work for you if you already migrated while it was broken. I'm also not sure if the fix was successful, I'll have to check with our developer and ensure he tested it.
No, because BF specials aren't eligible for migration with data for $5. Maybe we'll add that in, or maybe we already added in, I'm just going based off our old official policy for these.
@VirMach Just to update as I beleive you were manually going to fix this.
Migrated (without data) my "MIGRATION SPECIAL" in Amsterdam to Amsterdam 3 days ago.
New IP has been allocated but nothing else has happened.
No tickets or panic - thanks.
Yeah for most people it won't be worth it. If I was in the customers' shoes I'd wait but again we are after all VirMach so who knows how many other problems and unexpected delays we'll run into at this point, it might be a good option for anyone not willing to be indefinitely patient.
Looks like it's still broken then, I'll let our developer know.
Sorry, we originally had everything set up with Coinbase but Coinbase just fully ended their product and re-launched it as another product and in that time we switched to Coinpayments. We didn't like how Coinpayments did it either, it's not really how we set it up, it's all default settings on Coinpayments and a lot of this stuff they don't actually allow us to change.
I definitely don't like how Coinpayments just keeps the money hostage without automatically forwarding it to us. I'm sure they're going to have some lawsuit in the future for this.
It's a complicated process, usually if we contact their support it takes a lot of back and forth to actually get it processed still. That's why these get put up in batches and then we communicate it with them in bulk, but a lot of times it still just gets stuck in limbo. That's why it's really important to follow the initial instructions exactly and ensure you send it with high enough fee for it to arrive within that timeframe. I'll check again and see if they've since added any settings for us that we can tune further to avoid situations such as this one.
I do have these tickets in a queue right now. I do know it's taken much longer than it usually would as we're absolutely swamped with tickets.
@VirMach Please take a look at the network issues related to Node SJKVM10 on page 194 to avoid more tickets.
With how our existing contract work, we actually save money on Ryzen and for the non-special plans the way we have pricing and disk set up, we're actually barely overselling anything relative to E5 nodes so I do suspect we'll essentially be better than almost any other provider.
For example, if a normal node fills up with just NVMe2G packages, once it's "full" we expect somewhere around 60-90GB out of total 128GB RAM usage, and maybe 70% total CPU usage. This is based on existing data, and of course it may change as customers shift to using more CPU or RAM. Disk usage on most nodes is also relatively low which means you should be able to burst a lot to near full speeds once they calm down, and bandwidth usage is also very low compared to our total commits.
The catch is though, we will be overselling a little more, but based on benchmark scores, it should still be relatively good performance and no real CPU bottleneck. But we never know how it's actually going to turn out once everyone is migrated, especially if a lot of specials are re-sold and the new people using the VMs decide to use it for something that has higher CPU usage.
For Tokyo, to make the same pricing work when colocation costs like let's say 5x more than the "average" other location, we definitely do need to fill them further though so that location is probably always going to be less performance. I think this is still better than charging 2-3x higher price to maintain the same level of emptiness as other locations.
Create a ticket, don't mention migration. Do connection issue department.
Just make sure new ticket you opened is in outage department, priority.
Without data option for that doesn't work for specials. It only works for regular plans right now. I also don't know if it can work with Ryzen, I'm speaking with our developer.
It may be possible to take our new code from the new Ryzen migrate button and morph them together, and allow free without data with available, and if not available, do it for $3, and do with data for $5. Might also end up being a lot less confusing.
SJKVM10 has been broken for some time. These are going to be migrated in the next batch of planned maintenance, and you're probably going to get several months of free service as an apology.
This has an issue with bonding and CC had a brain drain after Jon sold the company and Alex Vial left. A lot of these servers were set up by an actual networking engineer, and now it's difficult to try to communicate anything with the existing staff since they don't know how it works.
Maybe I'm wrong though, that's just how I remember our attempts of fixing the node went and we decided to just migrate to Ryzen as the fix instead. But then Ryzen migration kept getting delayed.
I'd rather have no cryptocurrency options at all then. Fishy companies like Coinpayments.net will turn away a number of good customers from you.
@VirMach Something wrong with TYOC30？The speed is so bad
Provider | Location (Link) | Send Speed | Recv Speed
| | |
Clouvider | London, UK (10G) | 143 Mbits/sec | 19.1 Mbits/sec
Online.net | Paris, FR (10G) | 68.1 Mbits/sec | 27.6 Mbits/sec
Hybula | The Netherlands (40G) | busy | 34.6 Mbits/sec
Clouvider | NYC, NY, US (10G) | 119 Mbits/sec | 27.2 Mbits/sec
Velocity Online | Tallahassee, FL, US (10G) | 10.9 Mbits/sec | 11.2 Mbits/sec
Clouvider | Los Angeles, CA, US (10G) | 115 Mbits/sec | 13.0 Mbits/sec
@jinwyp @56warmers In case you missed it here was VirMach's answer to your SJKVM10 slow network problem..
I checked @vincentwenad's VPS (TYOC040) for 2 hours and the problem has not been solved even after reinstalling the Almalinux 8 template. I'm a little tired, but VirMach will be even more tired because they handles a lot of tickets...
With @tototo help, we’ve tried the following actions. But none of them worked.
Checked boot order and ISO mount settings
Installed Almalinux template from SolusVM panel and waited more than 15 minutes
Installed Almalinux template from billing panel and waited more than 15 minutes
After enabling rescue mode and waiting for a while, I logged in via SSH and entered with showed password, but it was "Access denied"
Even after executing each of these one by one, VPS status is still offline, but SSH client shows login prompt and I cannot login
** Execute "Shutdown" from SolusVM panel
** Execute "Power Off" from SolusVM panel
** Execute "Shutdown" from billing panel
Remaining bandwidth is decreasing little by little
@VirMach
New ticket is opened on outage priority. #660458
@VirMach Japan moved to San Jose before, I want to move back to Japan now。#446486
I appreciate for u for your help.
Maybe we will wait for a while.
Anyway, it's so kind of u.
This is not how it works, Japan is disable for migration for now. Maybe it will be enable for later, maybe not. Maybe soon, maybe in a year.
Also that ticket probably lost you all migrations in future :-)
Recently, I start to think, over time ryzen has been "premium" change to "normal" product line. Ryzen are still excellent especially with the virmach price point, and will work for many years, just no rush.
what I can also offer is activations on San Jose instead, use ticket title "Switch Tokyo to San Jose" and if you do this you can reply back to the same ticket after a month and have a delayed free migration to Tokyo plus 1 month additional term as an apology. Just if you do this make sure you close all your other tickets. You should make this in the priority department.
To be fair to them, these cases are still the minority.
It's difficult to integrate anything else that allows as many options as Coinpayments right now. For example, Coinbase will not allow XRP and there's a good number of people that prefer using that due to the fast transaction time and small fee. That's the main reason we kept it and just also added Coinbase back in the new version of their module they released. It generally has less problems but it will be a little more limited.
I'll try to get to the Coinpayments queue today and get it resolved with them, and apologize for the wait.
I always appreciate people helping eachother out, especially when we're backlogged with a lot of tickets. In this case it sounds like it could just fully be broken in which case customer can use the outage report button in priority department and I'll ensure the LVM exists or re-create it if it doesn't.
I did offer some people special temporary migrations to San Jose and I'll still move these back for free but the problem is a lot of people are abusing this to pretend like they were part of it. I'll have to check if he was part of the offered migration.
Have you even read what you are poorly quoting?