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.
VPSCorner - Low I/O, bandwith and extensive downtime
This discussion has been closed.
Comments
You're right, it's $1.75
@DennisdeWit Is this some kind of joke? ask them to show proof of it being a DoS. It's 0.6mbps. sigh
@Ishaq: I feel the same way. I just asked in the ticket if they shouldn't be heighten up the limit. Haven't had answers since 13:00 (17:35 right here, right now). But I might know the reason:
The lines in Moldova are not considered 'great', according to Simon. One node has to share a line of 5 Mbps. So, when I just use almost 1 Mbps, I'm eating up the fifth of full capacity, which is high. But this is just insane.
VPSCorner is busy having new hardware in an other DC in Moldova which can have an uploadspeed of 10 Gbps. It's still not ready, so they limit their customers. You can get on the waitinglist to be moved to the new node once it's ready. I did not get an ETA when I got this answer.
I got an email before I left them saying they were moving clients to Netherlands with 10Gbps. So is it Moldova, Netherlands or what?
They said they were moving clients to Netherlands but they will keep some moldavian node for those who want to stay there. No update since then.
Ah, right. Oh well, too late now anyway.
I'm on again. They told me in one single line they suspected a DoS.
"DoSing VPS 565 (): 24026 packets per second
That is alert we got, We have re-enabled VPS> "
I replied using the graph I posted on this topic.
I'm not sure what they really did, to be honest. They just turned off the VPS. I clicked the Boot-button in Solus, but nothing happened. It's weird, but it's good to be back now. Hope it will last some longer than last time, though. ;-)
I think that you packed your bags away from VPSCorner in proper time
Yesterday's notification:
Today's notification:
I am wondering if they have really so much issues with all their Moldavian nodes out of sudden or that's just a "smoke" before they inform us about new IPs on some other location because < put anything but truth here >.
Indeed @Spirit. I'm on the Alexis node, so I'm off, too. I asked about an ETA for the downtime since it's lasting 9 hours and we still have not received any updates.
That is not true. Most of vps providers allow to run such things, as long as you don't overuse cpu, exceed allowed bandwidth, stay on your traffic and use normal memory. I have a shoutcast streaming server in a 128MB memory vps with 1 TB of bandwidth of 100Mbps. I run a red5 server in another vps with normal use, with no problem at all.
Hnn, I mailed some providers before, like Torqhost, I believe. They told me they do not allow any of my software.
However, I e-mailed VPSCorner for the Alexis ETA. I received the following e-mail:
"The nodes are down due to a kernel issue. They will be sorted today (most likely) and credit applied accordingly. Sorry about this, we are working to fix it. Naturally we're trying to fix ASAP.
Kind Regards,
Luke"
@Jack: It is. I can't get any more info, as three nodes are down at the same time. Because Alexis has just 6 clients on it, I guess the node will have to wait.
I think Evo will be the first one to be fixed, as there are 20 clients waiting on it. I don't know what the exact problem is. But I can't think of that one kernel panic will cause 9 hours of downtime already. I don't have any details about what exactly happened. The only thing VPS kept emphasizing, is that they will credit their customers because of the downtime.
@Spirit Yeah I got that email even though my service is terminated.
@DennisdeWit of course not. All nodes didn't just die at once after months of running.
I have feeling that we will soon have alternative to either move to some other location or share one "big" mega slow node
Or no alternatives at all?
That's known here as "NordicVPS stunt"
@Spirit: I hope so not. If that happens, I will be gone soon. I've got moved into Alexis just because of the slow Evo-node. Sharing another slow node is not an option for me. And they may not, considering their own rules: "OUR VPS ARE NEVER OVERSOLD", is what their frontpage says.
Hnn, when logging in to SolusVM:
"Access to this virtual server has been restricted. The node is currently under maintenance. Please try again later."
Never had that error before. Even not when a node was fully down. Then I just could not log in at all. But restricted? Feel like @Spirit will be right soon.
No, I won't tarnish VPSCorners' reputation, at all. But Spirit and I are on the same (VPS)corner, right now. I asked @Jack's question about which errors they have.
I don't appreciate to be lied. They want to go out of Moldova? Ok. But all this wasting of our time with stories how they need to migrate clients from one node to other and day later all nodes down because kernel panics, +12 hours of downtime and counting with their response that it will be sorted today most likely (most likely?!) ... what a load of bull.
Thanks, @Spirit! Now I am not the one who's tarnishing a reputation. :-)). But I fully agree with you. It sucks to be offline for such a long time. I'm awaiting if they can refund me. Spoke with my co-owner of the webstation, we both want to move to MisterHost.
Well, we're beyond low I/O and bandwith issue here, aren't we? And their vague explanation why are all their Moldovan nodes down while we know that they want to go out of Moldova doesn't make any better
They promised to keep node(s) for customers that wanted to stay in MD, bad if they wont honor that.
They use Raid 1...
To clarify (from VPSCorner),
I've been speaking to Simon but only vaguely as he is working on fixing the issue. Essentially we're having problems with 3x nodes in what I gather is a kernel issue from an update/change made to all 3.
To clarify as well, we are NOT exiting Moldova at all and are staying in Moldova as requested at the time.
We've had a few hiccups as of late but as always are working on fixing any issues.
I do appologize to all clients and we will be issuing credit.
Kind Regards,
VPSCorner.
@VPSCorner
What about the Netherlands move that never happened?
A secondary comment to clarify further is that the node move was to fill slots that were available. We don't oversell at all, keep in mind nodes are based in Moldova so speeds are going to be slower and as we tell customers frequently, we prioritize actual traffic and slow down speed tests as they waste bandwidth and reduce speeds. We also have various rules in place to automatically suspend when a certain rate of packets is exceeded indicating DDoS etc and this has a low false positive rate.
I'm sure customers who think back to before this will praise us for our attitude and the way we deal with things as well as what we permit.
We decided to first look into merging the lower occupied node and then move to NL after that billing period so we could time things correctly. We still intend to do this but like any business need to time things correctly so that it works out better for not only us but for customers.
We had lots of negative response to NL so we decided to delay it for that reason as well as the above which is specifically why we aren't leaving Moldova.
@VPSCorner: Maybe a good idea would be to start communicating better with Simon. As customers, we want to know what's happening. Before we know, you're down for a day, and we still not know what is happening exactly.
Besides: why haven't you communicate with your customers about the kernel update. You just updated it, it failed and know you're talking about a failing kernel? No wonder.
We do communicate very frequently but today due to the issues are very hectic. We're very aware of how things are perceived and for the sake of being open and honest ill be blunt and say we are not bullshitting and the issues are very real and very serious. We're customers ourselves in other sectors and know this will damage our reputation and pocket.
Simon has told me that it was a vzkernel release that had a panic essentially under DDOS. We Deal with DDOS rather effectively, especially when we know who it is but in this instance it has caused other more serious problems.
We throttle benchmarks as we constantly get it and it slows things down when it's frequent by multiple users.
To finish things up we also had 2 rogue VPS using an intensive spider, we at the time took their word it was legitimate traffic but when we received lots of i/o abuse we discovered it to be impacting quite a bit.
We're normally flexible and deal with things relatively smoothly but unfortunately this has been quite a big problem.
In all honesty we've never had any issues like this in the past and are quite lenient, we often challenge our isp and other third parties on behalf of clients and have in the past taken extreme measures to protect even $5 clients.
We're not in this for the money, we're in it long term and rely on our reputation as our commodity of choice.
I hope this clears thongs up. Rest assured were working on it and we aren't happy either about the issues and time.
All affected clients WILL receive FULL credit.
Thanks
@VPSCorner good luck, looking forward to keep using you in .md
At this point in time, we don't expect it to take much longer. It's currently out of our hands and with the ISP as the issues we're dealing with aren't fixable via KVM etc. They're taking their time and spoke to us last night about it and will apparently take a further look into it this morning. I imagine I should have an update for about lunch time in the UK.
As you can imagine, we're in damage control mode and have taken quite a hit when it comes to both reputation (as unavoidable as this is) and financially but we intend to repair, patch and exceed our previous status.
As one of the admins, I personally had a VPS on one of the affected nodes with multiple sites that made money and are suffering due to being down (in different niches). I also had some freelance clients on my VPS from website work I do as a side project which are unfortunately affected as well.
Naturally I will deal with the above issues, I wanted to mention them to highlight the severity of this situation and so that customers know we aren't sitting around twiddling our thumbs.
All affected customers as mentioned previously (not all nodes are down, 3 are) will be provided credit and as mentioned, we're quite clearly in severe damage control mode so I'm sure we will pack plenty of value and surprises in over the coming months to patch the gaping hole in not only our pocket and reputation but our pride. With a bit luck we escape with only a few scrapes and at the end of it, we're hoping we can play our cards right and as good an analogy as it is, gain a lollypop in the process from those noticing what we're doing.