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
@william edis is not hiding under a false separate brand.
That is not his father, and he is not a kid.
What comes to question now is how, with Jeremiah, this great sys admin everyone talks about (I'm not questioning you guys, everyone has incredible things to say about him), and with involvement from the datacenter, ChicagoVPS consistently received such mediocre points on performance?
Admittedly it looked as though they were slowly inching up in performance, but for the most part managed to completely skate past what most people around here consider to be standard requirements (high uptime, ability to burst CPU reasonably, enough disk I/O to act as a buffer between you and one instance of Minecraft).
I respect the guys for their marketing skills. They obviously did well. I just don't see how they maintained such mediocre performance while having such direct access to their equipment. Best of luck to everyone involved, I don't wish misfortunate on anyone, but I'm not going to sugar coat it...I'm aiming for a piece of this pie.
They confirmed with their own figures that they were putting ~110 vm's to a node.
I'm fairly sure I brought this up in a post somewhere (I know I did on IRC) but what happened was during their first sale they got lots of people that bought them as replacements for say a 512M or something smaller and never really ramped the RAM.
Once it proved the offer was pretty much a 'no stock limits anymore', they started bringing in the heavy (ab)users. We had the same problem on our $15/y's and such.
@rm_ showed that their CPU performance wasn't quite up to snuff but it's possible Jeremiah/etc didn't have a chance to clean up that HN yet.
What you can say is they changed the market since there's no less than 3 ther providers I can name that have straight copied his 2G offer.
Francisco
@justinb, so only a kid has a father? that makes no sense
Very true. Add 1 more to that sometime this week
That entire Christian philosophy is based on the idea that you can change someone's behavior or attitude by example. Threat him like you want to be treated and you can make that person see what is right and what is wrong.
That is from the old testament and it conflicts with the new teachings.
I am not going to church every Sunday and I don't like to claim that I know allot about the teachings of Christ, but I know this much: God is in the details and usually the way he does things is that he changes one minor detail in a persons's life, a detail that snowballs and that can have cataclysmic effects, and the purpose of this is to teach someone how they can better themselves.
Chris - Avante
Chris - ChicagoVPS
Chris - NixDot
Chris - KJMS(Closed)
and
All this talk about young folks running hosting companies makes me think of him ... and how he would look like running one ... just close your eyes, listed to the lyrics and picture him running his own Minecraft cluster:
110 x 2GB thats 200GB and more RAM vs. phys. 32GB. Thats over 600% overselling omg. May be 100% is fine but this is insane ...
this is totally irresponsible. I will not be surprised if they reboot the nodes on a daily basis on memory exhaustion.
Thanks to both of you for the insight. This should be interesting..
Also, I've needed to bring my VPS on LA1 back up twice now in Solus..
They probably have 32gb ram and are probably 28GB MAX
They do have 32 gigs of ram per node. However I can bet that they use less than 50% per node.
I use around 40% on filled up nodes myself
@Alex_LiquidHost i bet you don't put 100x 2GB RAM VPSes on a 32GB node.
it might be good before they decide to use the RAM like panel updates etc. Then the problem starts.
I don't think they said that they put 100 x 2GB on a 32GB node, just 110 VMs in total. That might as well be 50 128MBs.
Ok, let's make some statistics. To the people who had a VPS affected by this - what was your VPS plan? And on which node?
In a ticket that was published in a forum (wht I think) Chris talked about 90
http://i.imgur.com/WdXRWh.png
http://imgur.com/a/UBS5G
http://www.webhostingtalk.com/showthread.php?t=1125003
God node 31. I was on that node and it was just SLOW <-- Bold Text that, Underline, make it flash, and it still won't show my meaning of slow.
Using High IOPS SSDs as swap?
110 VPS per node (we do things other providers can't)
I don't think they are using high iops ssd's as swap.
Won't that wear the SSD quite fast and still perform quite a bit slower than actual RAM?
Of course it will be slower, but not nearly as slow as that swap hitting a mechanical disk
Wrong, slower. SSD writes are slower than a 4 disk raid10
But not faster than ssd raid10 boom!
And a raid10 ssd array will cost so much more. I know you are being sarcastic here, but it's like stupidity took over here. And nobody is going to display the truth, cuz those without understanding will cry foul.
Yes I was being sarcastic.
@miTgiB ...
Raid 0 array of SSD's sat as swap... beats any mechanical? Two 256GB Samsung 830's or something?
I was comparing apples with apples not with oranges. Although my original point was in half jest based on something that was sold to me a while ago from colocrossing.
Yes it would technically beat it but only by the smallest margins. Not enough to justify it.
Raid-10 , SSDs , cheap VPS with 2GB RAM ... get real please ...