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
Around 3/4 of the new 7950X's are up and running. The guys are fixing the last 1/4 (looks to be NIC firmware issue, thank god).
I just need to setup the VXLAN bridge and final software installs. I expect to be rolling this week. Was supposed to be last week but the NIC issues caused delays.
Francisco
Hello, it will then be possible to purchase DDOS protection together with the VPS.
Can't wait，hehe
I am looking forward to the Swiss region, haha.
Hi, may I know if Frantech WHMCS license can be hosted at Cloudzy?
BuyVM always out of stock, at least 1-2GB plan.
Thanks!
Not yet.
Migrations started last night
Francisco
is buyvm down now anyone?
We’ve been doing the migrations and I’ve had to reboot nodes to get them matching software wise, sorry.
I’ve been mass booting when I do the reboots though then kick off a migrations
Francisco
My LUX VM has been down for 1 hour and 49 minutes. The status is showing as UNKNOWN.
I would click start in stallion. Worst case ticket and I’ll sort it asap.
Francisco
When can I buy it?
CH should be sometime this coming week.
We're making good progress on migrations. Slowed down a bit since we have to do OS upgrades at the same time, but good!
Francisco
hi, I just 1 day ago I login to my VPS, I can login without problems, but I realize that there is no connection ... I have a website of my company, and this without connection. how long will it take? it is still under maintenance?
That doesn’t sound right. The migrations shouldn’t disrupt network for most users. PM me your ip.
Francisco
PM, sent.
@Francisco
Have all the migrations from LUX->SUI completed? If I mtr my server's public IP, I still see packets going to core2.lux1.he.net and then another 17ms is lost on "???" before it ends up at the server.
Is there any chance that this forwarding from LUX->SUI is rate-limiting / dropping UDP packets? (assuming that's what's currently happening)
The reason I ask is that I only use my server as a haproxy front end to my dedis, with the connection to the backend carried over wireguard. For the last 2.5 days, it's been marked as failing on hetrixtools. I've done a bit of rudimentary digging, and any repsonses that fit in a single packet complete fine, but anything that requires multiple packets, it seems that some of them get dropped and so the haproxy returns "504 Gateway timed out" to the client, long after the client transmitted the full response back to haproxy.
I can verify that from the buyvm server,
wget -O- http://server-over-wg-private-net/small.txtworks fine andwget -O- http://server-over-wg-private-net/larger.txtfails, so it's not an haproxy issue, just transport over wireguard where multiple packets are sent in quick succession. It wasn't a problem before, it's not a problem on any other servers I have (and I've had this setup for several years now), and I've rebooted both sides just to be sure.I can see a steady stream of TCP between the two over wireguard because of the various monitoring systems I have in place, so it's not trivial to correlate both sides exactly which packets are going missing. I can do if required, but if this is just a temporary solution I'll just wait.
Looks like it's true.
When will you send me back to the "ghetto" where I belong?)
Things still somewhat route through lux, it’s split st the moment while we wait on prefix list updates.
Looks like an improvement to me?
Francisco
Absolutely
Besides, I never had any complaints about the previous behavior of the server, given the specifics of its use;)
Well, Hannan did right by everyone on this.
Francisco
Even if there was a differentiated approach to users with the goal of separating “abusers” or “those similar to abusers” into a separate segment (which, of course, I cannot prove), I approve of this.
Hi, I would like to know how long it will take for the migration to complete, as I don't have any storage blocks and it has been taking more than 24 hours.It is still not online.
With storage slab attached, can confirm 24+ hours offline. Unfortunate, sure, especially with the short notice. From what I understand, this is a physical relocation, though - as in, loading the servers into trucks, driving them from Luxembourg to Switzerland across at least two borders, then installing them in a data center there. Lots of potential for delays, hoping for the best.
Your VM got classified incorrectly then. I grabbed everyone that was a non slab and moved it but maybe the script was wrong.
Was some delays in the move but the guys are working quickly.
Francisco
Since the move my non slab lux vm has over 100ms extra latency on it from UK, even a ping from it is horrific
Slab vm still offline.
We will have HE online soon
Slabs have been a slow move but they’re at least in Switzerland now so it’ll be up today.
Francisco
Great stuff awesome work as always! appreciate the fact it's physical move for slabs so the logistics of it must be a nightmare. Assume you have a coffee machine right next to you
I wasn't involved in the pickup/transport, I simply worked with the timelines they gave. Obviously that all went out the window due to a bunch of delays, but they're at least in the DC now with the gear and actively racking.
I won't be going to bed any time soon :P
Francisco
What is the current ETA for restore slab VMs?
Nothing concrete yet. There was a misunderstanding on which nodes we needed in CH and so they have to do another trip.
If you don’t need your OS drive (or you have a backup or snapshot in stallion) we can remake your vm and you can probably attach the slab.
There is a few slab nodes needing some fixes (basically all cabling related) so we hope to get that fixed in the next few hours.
Francisco