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
@ Maounique, thanks a lot of for clarification. We are certainly in very early stage, we may not have issues with any resource abuse cases. If we in that bracket, we are more than happy that we got a lot of paying customers, happy to pay more infrastructure.
Give me depth of Virtual Firewall/Router/Load Balancer option that cost about 4.x per month.
My understanding is that Virtual Router/Load Balancer/Firewall is again a VPS (technically), but managed by Prometeus and it is software based. Is that correct?
If it is software based, do you have HA for software based load balancer when the physical node for load balancer failed? How long do you think, this switch can happen?
If you don't mind, please list the tools and technologies used in load balancer? Do we have option to access load balancer instance via SSH?
One way, I am comfortable with having Premetus takes care of Load Balancer as block box, however like to understand HA related stuffs. The reason I prefer iwstack from ramnode is the load balancer part. We don't find hourly instances are not really use cases at present.
That is what it comes down to breaking down the monthly usage.
Every instance can use that bw, at the end of the month a calculation is done to weight total usage against how long instances were on and if in average was more than 1.5 an hour outgoing, the rest is billed.
Not really by us, but by iwstack master. It is totslly automated, nobody intervenes.
There is no specific node for it, if a node fails, all vms restart on another. Downtime can be some 10 minutes if your vm is last to restart.
The vr is totally automated, you can access it via the ui or api but cannot login to it.
It can do a lot of algorithms from round robin onwards.
@Maounique This is one of the most transparent set of answers about overselling/usage/resources I have ever seen in this forum! Kudos!
I do not believe in lying, tho I am known to exagerate at times. It is enough to tell the truth when people are used with low end setups and all kind of restrictions to make those work (barely).
I am honest and I believe uncle is overspending and, at times, wasting a lot of resources. We have servers with loads below 10%. At times even below 5. Is that CPU worth it? Is it worth it to put only 64/128 gb ram on openly oversold servers and have cpu used only 50% even with the minecraft people in full bloom? Probably not, but I am not him and he always gives the argument those can be used to the main business too when the idea was to reuse what is left from there, but it turned out the VPS for all business needs better gear overall than businesses paying thousands a month...
@Maounique, just curious, who is this "uncle"? Why this convention? I often read this on lowendtalk and lowendbox.
Uncle Salvatore Sciacco, the owner of Prometeus brand
He used to be active here when he had more time, and everybody loved him and called him affectionately uncle. Some events in his life as well as some key people in the business where he is CTO being recruited by amazon and other big companies, meant he has a lot less time and big contracts to deal with/make so he has no more time to spend here. Also he has been treated unfairly by some "competitors" (many people consider prometeus competition, even though prometeus does not due to huge fragmentation of the market on geography and quality not only on price) with threats and stuff because they tried to force him to kick me and didnt manage, so he is not so keen to come here or on other forums anymore. I do not care as I am used with bullies on all forums but he is taking those things to the heart, so better to avoid than treat.
Just curious about the virtual router. It is similar concept as virtual network in other cloud service providers, such as Windows Azure? In the virtual network, how the internal IP is assigned to the VMs, do you use DHCP or static IP?
If you are using DHCP, anyway to persist the IP address across reboot? What if I want to use name instead of IP, do I need to deploy my internal DNS service?
Also wonder whether you provide load balancing? If not, at least support VIP, so I can run keepalived by myself?
To me, the differentiating factor for the service provider is ultimately what network configuration they can provide, this significantly impact the operation cost.
I can remember him posting a few times, too bad because he was nice and always had his own opinion about things with a good set or arguments.
What I like about Prometeus brands is, you will never hear them say "Dude you are just paying $XXX, don't expect too much". Which I hear so frequent here on other hosts.
They do. Which is the true thing to say, too. You cannot expect 100% uptime, 100% full performance, 100% instant ticket responses, 100% great routes on a budget service on a non HA infrastructure. Economics doesn't work like that. Forget the economics, there will always someone overusing something and saturating a resource on the node. Because it's a shared environment by design. And budget services attracts people who don't know what they do mostly.
Not the same context I was refering to.