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.
Question for VPS Providers regarding Meltdown/Spectre Patch
Hey Guys
I've plans to launch VPS hosting services soon (KVM Virtualisation) and until Meltdown/Spectre vulnerabilities were released I was planning to go for Intel XEON CPU's
But after a lot of reading it appears that applying patches for Meltdown/Spectre will lead to lot of performance degradation, so I wanted to know how exactly are you dealing with it ?
Are you adding more VPS nodes and moving some of the VPS to new nodes ? or you are just letting your clients use the same node ? I understand adding new node without charging anything extra client might not be possible for all VPS providers considering the costs involved for adding new VPS Nodes
Thanks
Comments
I don't know how to say this without sounding like a complete dick, so I am just going to say it anyway and prefix it with 'nothing personal'
The level of naivety that this post alone shows tells me and no doubt many others here that you are very far from ready or experienced enough to be offering anything hosting related on a commercial basis.
The last thing this market segment needs is another button masher host, i.e. you know what to click and how to run someone else's scripts but you really have no idea what they 'actually' do.
please seriously reconsider.
This is usually when you tell me about your years of experience that contradict your post that someone with years of experience would not be making or that you have experienced admins (that you decided not to direct this questions towards) but please.. go ahead and tell me why I am wrong.
Sincerely,
The people in the market segment that don't do this for part time pin money.
Is is pertinent to those who used to oversell their CPU way too much. Both Amazon, AWS, Rackspace and many others have already provisioned, and are keeping doing so, extra capacity to handle the load impact this may cause. For the small providers, I don't think they'll notice much. Now if you think of AWS, with their High-CPU instances, that'll cost them some clients ending up not upgrading, this is not a solution for them. They'd rather up the specs in a few months and keep the same price.
In the AWS example, this also applies to losing processing power they have in RDS, ES, etc etc. Customers won't pay more for the same service level / quality they already pay for, AWS is rather adapting. This is a real scenario.
Hosts pop up all the time. It's the nature of the business. Just go through LEB, I can't count the deadpools. Though it shouldn't stop anyone from trying if they're devoted and passionate. They may end up becoming the next Ramnode (unlikely but still).
You definitely sound like a DICK, what made you think that i'm the technical guy ? or what role do I have in the company ?
Even though I'm not the technical guy here,I'm still trying to do my homework before investing in business and won't be doing it for another few months, till I understand this business
Did I even mentioned anything about my experience? So what makes you think that I'll even talk about it ?
I would suggest to listen for resolution from AWS, Azure, Google, DO and many others since they are the first hand to take care of several months earlier than others.
Believe many of providers are also in listening/monitoring mode cuz patches aren't helping for real due to performance degrade like u said.
I'd like to take a brief moment to tell you what you just did, in the eyes of so many readers that it will be irrelevant that maybe 1 or 2 don't agree, everyone else will.
You just replied to one of the most respected and mature providers on LET, after he posted something that genuinely attempted to offer you solid advice, and called him a dick. He admitted his words might be taken that way, which should be your social cue to cut him a little slack in perception.
Personally I didn't take your post as implying any significant lack of knowledge, but that won't be relevant anymore. LET is brutal. You'll never live this down.
@jarland God forbid they decide to offer something here.
Ah makes sense, this guy is the new iteration of rectified.net. I'm sure you guys remember his other account: https://www.lowendtalk.com/profile/138275/rectified
..but that guy was smart enough to admin?
E: UPDATED-
This must be his wife, or his poodle. Or an open proxy.
There's nothing about the original post that implies inability to admin. There's some talk of performance loss around patches, he asked how other providers are handling it with their existing nodes. Is that not a good question?
That person was an admin working for SharkTech. To come here and ask about some software patches and how the long term implications are for something which is still very much incomplete and in the air does not make me think of a seasoned professional.
Put yourself in the shoes of an admin who has been recently thrust out of the role of actually managing the servers, shortly before all of this went down, and has probably been so busy that the wealth of conversation/information shared around the topic has simply been too much to keep up with in the short term.
I'd probably rent an hourly and test it myself?
It's not that. I've generally got a decent eye for these things. IP login / account creation time matters. When you suddenly never log into one account again, and another is immediately created after your last login using the same IP, it's the same person.
Well, if the OP can Rectify his statements, perhaps we will know more..
Yes indeed been following the developments around this and it's alot to take in. Even big web hosting/cloud players are still learning new developments around this right now. Like that Intel Skylake/Kabylake are due for additional patches and updates for Spectre variant 2 https://www.phoronix.com/scan.php?page=news_item&px=RETPOLINE_UNDERFLOW
Busy times ahead !
VortexNode us mf'ers.
For sure. I mean, I'm a decent server admin. I have would feel perfectly fine saying "I plan to begin hosting KVM servers soon" if that were ahead for me, and I'd be able to handle the job more than just adequately (I'm not saying amazing, I'm just saying above the level of "I know how to keep the power cable plugged in"). Yet, I'm not fully informed on the depth of this issue and what many providers here are doing for it. Quite simply: I don't manage those servers right now today, I don't communicate with customers about it directly, I'm not involved directly in anyone's patching process, and I have a LOT of things on my plate (My team members at a certain job are well informed and could inform me at a moment's notice, but I don't need to fully understand every detail of it today, only enough to know that I'm well positioned with MXroute).
So if I decided to launch KVM hosting services soon, it might pay for me to catch up and ask a few providers for their opinions and plans about this, if they were willing to share and offer me some insight that doesn't require me to start at the beginning of what is now really deep threads everywhere.
If I can put myself in the position where it makes sense to be a decent admin, not be up to speed on this, be busy, and see the value in asking for some advice on it... hard for me to be too judging of someone else who I do happen to know has recently gone through some very significant and personal transitions.
I had no intention to offend anybody and I do understand he is a reputable provider, but judging someone just because of a question nobody else asked in the forum is not good.
I agree my partner owned rectified, and whatever happen between rectified and Sharktech I am not allowed to discuss due to the settlement agreement etc. that happened between them, all I can say is he did his best and settled the dispute amicably with Sharktech to avoid further downtime for clients. As far as I know most of his clients were happy till that disputed happened and he grew the company very quickly, without proper knowledge and experience one cannot do it, please rectify me if I am wrong.
I am using same IP as we work from same office, I am the investor and he will be working partner.
He could not reply to any threads related to dispute due to his agreements done.
Well if i was launching a new web hosting service probably pays to just wait and see how this spectre/meltdown mess plays out as who knows you may end up needing to replace the cpus anyway. That would potentially be a huge time waster when starting a new web hosting business and to inconvenience new clients who have no loyalty to your business would be detrimental to a new business's bottom line financially and reputation wise.
Also I take my words back calling you a dick, your advice is good but I am a semi-tech guy and collect as much information as possible, due to these new vulnerabilities our whole business plan need to be updated, hence I wa collecting more information on how is everyone else dealing with it
You seem alright to me, but one more piece of advice. I know this can also make me sound like a bit of a dick, but you have to understand just how often people try to deceive LET members. This is actually really really good advice:
If by any chance at all that is a lie and you are the exact same person, pretending to be two people, you should come out with it now and you might have a chance to save face. If people here find it out later on their own, they will make sure your business is done.
It's good advice because it's true, and because someone tries to do this to LET members every month or so. It never ends well for them. But the members here have a soft spot for genuine admission of mistakes.
Yes your advice is genuine, I mean if I say I am the same person does it makes any difference, because I am not hiding that he is my partner, even he will be using same account, so at times it will be same person at some other time it will be me or may be our future sales manager, after your first reply I asked him what to respond and he suggested me to tell the truth about rectified , hiding it is not the solution
Again both of us will be using same account for our business unless LET allows multiple accounts per company.
There is a possibility that we might even start with total 3 partners, still in discussions with third partner
Thumbs up, this is a good way to get started. We have strong noses at LET, transparency is king
Anyways as per the original thread topic, was wondering are there any CPU that are not affected with this vulnerability?
I read about AMD on some blogs
Can anybody throw more light on it ?
We are looking for CPU for VPS nodes as well as low end dedicated servers that we plan to sell on our own RACKS
might want to read
between these 2 threads, you'll have alot of reading to do and yes that's how folks are tackling this mess initially = lots of reading