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
Local storage is definitely not planned yet. We are trying to further expand CEPH and make it more stable. This offers better availability and flexibility.
It's definitely worth considering when you can... Maybe even just having a few nodes which have a small amount of local NVMe with no CEPH storage, for people who require performance but don't need reliability or much storage space.
I will gladly take your idea on board. But I can't promise anything.
@PrepaidHost is refund of 9EUR credit deposit option or not ?
why do you need local disk?
@PrepaidHost why do you just respond to praises and not complaints?
Hello,
we will probably be editing the posts on LowEndTalk later today. However, we are currently still dealing with a high volume of support. I apologize for this.
My idling server, which has nothing running, has been using almost 100% CPU on I/O wait for most of today...
@Cybr what Service are you using for this kind of monitoring?
Hello,
that seems very strange to me. Can you please write a ticket with the info on the LowEndTalk post? I would like to check that out!
@PrepaidHost welcome at LET. you have come to the right place to have your product hardcore tested.
for using ceph make sure you have plenty OSDs and free space. under no circumstance do any overcomitting. also put proper IO limits on each VM... otherwise you'll find out the hard way, that this isn't just another raid ;-)
To be fair, the I/O Wait improved, it was up to 66%.
Thank you @Falzo for your contribution, je genau. OSDs and memory are sufficient, but with this workload it turns out that there are problems. More OSDs have already been ordered, but so far we are getting to grips with it. The rush on the offer was even stronger than on our Black Friday offer, which is why we are a little bit over the top, as we didn't expect it.
lowendtalk - highspeedorders
@PrepaidHost
Can't access dashboard, can't even access your website, and server is so slow it's unusable.
Garbage.
Our customers as well as we reach the homepage and the administration of the servers. Otherwise the hut would already be on fire. On our status page we report openly about problems and what our plan is to fix them. If you have any problems, it would be great if you could create a ticket. That way I can take a look at it. Thanks!
Well one thing is for sure.. that you wont have current issues next month because I guess 50% of orders wont be renewing.
For me your website is also not working.
Bad gateway
Ray ID: 852a9e970c9c2c6b
Sorry. How can you create a ticket if the site is not accessible?
The main dashboard is always online. There you can always initial tickets.
The service for service management however is indeed bugged. Couldn’t reach it several times myself. Most of the time waiting for a few hours solved the issue.
The support will answer quite fast though, there is also a discord server where you can talk to the staff/owner directly.
TLDR: wait an hour and 90% of the problems are solved, if it didn’t, open a support ticket. The Main Dashboard is separate from the server control panel.
PS: The VNC not being accessible is also normal, just wait some time, had it 4 times at least since I bought it last week xd, eventually it will work again
https://prepaid-host.com/de/dashboard ? Offline
F, first time the main dashboard is also down xd.
Yea my VPS has died 2 hours ago, seems to be that CEPH was unhappy
edit: yea was unhappy, however the VPS is back again.
...and it took
tmux
about 5s to resume the session. CPU looks fine, but I don't think many programs can burn the CPU with such I/O performance.Wow, this looks so bad. Hope they will find some solution to fix this.
Hello everyone. New hardware was installed yesterday. It would be great if you could check our status page. We try everything to satisfy you. Sorry for the inconvenience.
https://status.prepaid-host.com/
Hitting the lowend market is bit of tough as I understand, but I really appreciate your pricing and efforts. Hope everything get back to normal soon.
I'm unhappy with what happened after node was filled.
CPU from this offer - EPYC 7443 2.9GHz is on par with EPYC 7702P that has just 2GHz base clock in terms of singlecore. It should be WAY faster.
NVMe has 2.3K IOPS. This is performance of cheap microSD card and is very noticeable when doing anything, even updating OS.
I'm not even sure if I should consider "300GB NVMe" as advantage, cause ZFS array on just HDDs performs way better.
Only good thing I see in this offer is 26GB of RAM, which I didn't test for ballooning or killing RAM-hungry apps (like SSDNodes does).
I got what I paid for. These specs are just too good for this price so it's way oversold.
Good thing I charged just 20euro balance.
I gave that a few days, CEPH cluster issues were supposed to be fixed, its 1AM in Germany (so its not even saturated by others) so I don't have any hopes it will improve.
If at least IOPS will somehow improve I'll use that as failover VPS, as disk is quite big. CPU performance wont be that important then.
According to the status page its fixed.
This is a YABS from BeroHost, local NVMe.
The difference on CPU is marginal, however I/O is night an day.
This is my prepaidHost
What do you expect? Its a reseller from a reseller (https://prohosting24.de/). This is some Proxmox/Ceph cluster built together with Proxmox and does not work with high utilization. No reasonable provider will be able to offer these prices without overbooking, regardless of whether it is a Ceph cluster (which is naturally always slower than local storage) or local storage. At some point, the air is out.