Portal Home > Knowledgebase > Articles Database > Dinix offline
Dinix offline
| Posted by cesarmx, 11-24-2004, 09:08 PM |
| I am one vps of dinix and is offline... somebody knows something but?
|
| Posted by SoftWareRevue, 11-24-2004, 09:10 PM |
| Well, dinix isn't off line. And neither is my server there.
Maybe you should contact them.
|
| Posted by cesarmx, 11-24-2004, 09:13 PM |
| already was in contact with the enemy with them and they do not answer To my time is 5ta that this outside line and already takes but of 1 hour
|
| Posted by NeoGen, 11-24-2004, 09:44 PM |
| enemy??? who is enemy here?
|
| Posted by cesarmx, 11-24-2004, 09:50 PM |
| is no enemy. it can do Ping
if they want premiumwh.com
209.152.167.25
Offline
|
| Posted by mystro, 11-26-2004, 11:28 AM |
| I also have a VPS with them and have been unable to reach it for close to 24 hours.
I've opened a support ticket and attempted to contact them by phone only to be transferred to some voice mail box
First time this is happening with them so I'm willing to let it slide as long as it gets resolved ASAP :\
Anyone else in the loop?
|
| Posted by mdrussell, 11-26-2004, 12:04 PM |
| 24 hours is a long time to be down with no information. You are generous giving them that second chance.
|
| Posted by mystro, 11-26-2004, 01:03 PM |
| I'm not running anything mission critical on that VPS so I have the luxury of a little time, however I am getting rather annoyed at the same time, apparently they've erased my support ticket that I opened last night
Maybe I am being a little too generous
|
| Posted by cesarmx, 11-26-2004, 02:14 PM |
| Somebody knows because dinix no longer sells vps?
It has lowered his level on watch of dinix?
Apparently I am not the unico that has problems
|
| Posted by AniG, 11-27-2004, 04:34 PM |
| Nov 26-28
Dinix is not down. but yes their DNS has been timing out. The only response we (customers) have gotten from them in the last 48 hours is that there is a DNS delay/timeout problem due to UDP floods. No ETA of resolution at this time. Apparently datacenter engineers are still working on the problem. This is the first time I have seen such a tardy response to support from Dinix, even in times of crisis. Anyone else know anymore more about this please share.
Can anymore recommend a fully managed VPS host with great support and reliable network?
|
| Posted by Christine B., 11-27-2004, 06:04 PM |
| I have a VPS at Dinix and yesterday finally posted in their forum and opened a ticket once I could access dinix.com.
My mail volume has been way below normal (due to the holiday?) but also LATE. My server was rebooted yesterday and the restart messages came in one by one within the next 6 hours or so.
Customers who opened tickets were told of UDP floods. Several other users posted connectivity problems in the Dinix forum, and after 2 days of this Dinix still has NOT posted anything in their Network Status topic at their forum.
Customers are usually told to open a ticket, and personally, I'm quite tired of Dinix covering up outages. I realize it's not exactly an advertisement and just about all companies operate under the assumption that what customers don't know won't bother them.
Dinix never provided an explanation for another recent outage.
I had been rebooting my computer and resetting the modem on Thu - thinking it was a DSL problem because the sites then usually came up again. Of course now I realize that it has nothing to do with my DSL - sites including Dinix.com are simply not found one time and usually 30 seconds or a minute later the page loads fine.
I have to say that I'm very dissappointed with the Dinix handling of network problems and especially after the 2 hour network outage due to the backup router not working, I can only conclude that they're having problems they don't want their customers to know about.
In addition to the connectivity problems, they INCREASED the frustration for me by not telling me about the problems.
I'm not a techie, I don't monitor my sites, I just notice when they don't work as they are mostly reference sites used extensively by myself.
|
| Posted by AniG, 11-27-2004, 06:40 PM |
| Hi Christine,
I can understand your frustration. I am in the same boat as you. This is the first time ever that Dinix has been lax with its support. But there is some new info about Dedicated Denial of Service (DDoS/UDP flood) attacks that have been plaguing Dinix for the last 4 days! I agree with you completely that Dinix needs to find a better way to inform its customers to avoid causing them anxiety, but I would still give them the benefit of the doubt, considering I have never had any problems with them with being informed within 10/20 minutes of opening a ticket. I am waiting for a resolution to this problem as anxiously as you and a lot of unhappy customers at Dinix. You can check out this thread (Dinix forums) for more info:
http://www.dinix.com/forums/showthre...=2119#post2119
|
| Posted by aqi32, 11-27-2004, 08:06 PM |
| Dinix have always been brilliant with support, but.......
i have never ever seen anyone from any sort of management post anywhere.
and the vps's have been really slow for ages, with people complaining and trying to find out what's going on.
|
| Posted by WreckRman2, 11-27-2004, 08:35 PM |
| I've used Dinix for 10 months and they have been great. During the past 3-4 days I have had numerous complaints about DNS timeouts and numerous emails bouncing from unrouteable domains so today I opened a ticket. They replied with the following:
We are still having intermitant issues with UDP floods causing DNS to delay/timeout. Our datacenter engineers are working on the problem now. Although the main problem is resolved there is no ETA and we will keep your ticket updated. We apologize for any inconvenience.
Robert McCurdy
Support
www.Dinix.com
-----------------------
Then I come here to see if others are having the same problems and it appears that is the case.
|
| Posted by aqi32, 11-27-2004, 08:39 PM |
| yeah Rob has been sending out the same response to everyone
|
| Posted by Christine B., 11-27-2004, 09:12 PM |
| The response I got yesterday:
"We are still having intermitant issues with UDP floods causing DNS to delay/timeout. Our datacenter engineers are working on the problem now. Although the main problem is resolved there is no ETA and we will keep your ticket updated. We apologize for any inconvenience."
Copy/paste support for at least a day.
I suspect the people with a clue are out of town for the weekend and I doubt anyone is working on anything other than to paste that same response into ticket replies.
|
| Posted by bithost(NET), 11-28-2004, 03:19 AM |
| Hello all,
First if you go to the Dinix forums you know my level of access and information. The situation is very frustrating to me but I am doing everything I can to get a handle on the situation and keep everyone up to date. I get the same voicemail as you so I'm doing what I can!!!
The outage -- as best I can tell the UDP flood/DDoS is hitting only one VLAN. This means not all of the customers are being affected. For instance I have a dedicated server at Dinix and the response times/accessibility have been spot-on. Unfortunately it appears that there are VPSs/hostnode(s) (I don't know if it is just one VPS or multiple) in the VLAN that is targeted with the DDoS. This is why some VPSs are slow or timing out.
A UDP flood which hits certain servers, be they dedicated servers or VPS hostnodes, can overwhelm the server and hence make a reboot necessary. If you see notices in your e-mail or via your SSH session that the system has gone down for reboot -- that is why. A reboot will flush all the junk out of memory, flush out all the temporary sessions and connections, etc., so the server can resume operations at a normal level. (This is just a general explanation as to why a reboot could be necessary in this situation)
As for techs answering tickets -- I don't have access to the Support section of the desk so I have no idea there. All I can recommend is to please keep opening tickets. It is the only way I know of getting in touch with support. (If it helps any, I have a ticket open for 24 hours now with no response, so please don't feel as though I don't empathize. I do. If someone there would give me a blow-by-blow, I could pass it along to everyone.)
As far as sales, this is completely unrelated. If you search back in the WHT and DINIX forum archives, you'll find regular mention of planned growth. DINIX is built upon a very controlled and responsible business model where growth is controlled, so that the company does not become overwhelmed and reach a point where they do not have sufficient resources to support their customers. DINIX is in one of those stages again right now. The fact that VPSs are sold out for a bit is not a point of concern, it is a point of reassurance.
Having "sufficient resources to support their customers" probably sounds ridiculous in light of the support responses/times/communication you have observed in the past few days. Please understand the circumstances of the last few days are completely separate and unrelated from the planned growth plateau that the company had reached in terms of sales -- the ongoing DDoS's have required the work of both DINIX technicians and the E^Deltacom NOC engineers, and are rather extraordinary in terms of "normal" support needs. Please resist the urge to tie these two things together.
The people with a clue most certainly are on-site, they are the E^Deltacom NOC engineers and they have been working on the UDP floods for four days straight. What is important to remember is a DDoS is a distributed denial of service attack which means the origin and type of data is constantly changing. It is not a matter of setting up a simple "if-then" rule and letting it go. This has been a hands-on issue for four days with people manually working on ever-changing floods from all over the place.
As for the techs working the desk, it may well be that what you have been told, copy/pasted, is the same status that it was 24 or 36 hours ago. It's a game of whack-a-mole. They think they have it stopped, then another flood pops up. They shut that one down, everything looks okay, then wham another one. Over and over and over again. In this case, it is still the same story. I know it is frustrating to be told the same thing over and over again, but truly there may not be anything new to report other than the cycle continues the same as it was before.
Anyway I hope this helps to address some of the concerns presented as well as explain in general terms how some of these things work.
Cheers
Bailey
|
| Posted by Christine B., 11-28-2004, 04:24 AM |
| I got a ticket update: "We have implemented a solution for the problems at our network. ..."
I wonder what that solution is. And what exactly was the problem?
What will happen tomorrow? More of the same?
If they were able to come up with a solution after several days, why did it take so long?
This is obviously NOT Bailey's fault, but it sure tells me something about Dinix management. I'd be insane to continue to rely on Dinix.
And since it's not a Dinix exclusive problem, it appears that the only solution is to have 2 VPS or servers on different networks.
|
| Posted by bithost(NET), 11-28-2004, 06:00 AM |
| Christine, that's exactly what I have done in building my hosting business. I have servers at 3 data centers (VPSs and dedicateds) and will continue to diversify as we grow. I've been at other data centers which were completely offline (as in, dead pipe) for 19+ hours straight. At that point my entire business was down.
By diversifying, not everyone is down when the data center pukes. This actually keeps our business prices down because we are not trying to juggle all customers' needs at the same time, as would be the case with a single point of failure (single data center). Our staff only has to juggle a portion of the userbase when a data center goes down.
Christine, at some point I am confident that your operations will be to the point when you will have multiple servers/VPSs, and network redundancy is a great idea. We actually shoot backups to off-network servers routinely so that if one data center goes down, we have fresh data to use to restore customer data elsewhere. VPSs work nicely for that, in fact. We also use a VPS to do our monitoring and notifications.
Outages are a fact of the hosting business, I just wish that they hadn't hit to this extent in this case.
Bailey
|
| Posted by Kaith Sutai-Rustaz, 11-28-2004, 12:10 PM |
| In the past 4 days my sites have been hit or miss, email delays and other glitches. My phone has melted with my customers asking what was going on, and me being limited to saying "network problem, techs are on it, more info as soon as I know.".
Criptic and terse, and widely spaced was the information coming in. It gets to be very annoying and frustrating when you don't know, can't do, and must wait.
The problem was the DDOS. The solution? I dunno. There are a few reliable filtering systems that seem to be somewhat effective against these things. They may have installed one, or managed to get enough data to improve their existing ones. I'd love to know myself, but moreso out of curiosity.
Hopefully not. If it does, I have to trust they will be better armed to fight it.
You have to understand exactly what a DDOS is.
"Distributed denial-of-service attacks are ones in which the hacker plants malicious code on numerous, scattered and usually unwitting, servers. Those servers, known as zombies then flood a single IP address with packets so it is driven offline, unable to handle the volume. "
How many computers are attacking? Alot. 10,000+ is often times not a large number here as various groups stage contests to see who can compromise the most systems, and make the biggest splash.
Why did it take so long? No 2 attacking systems have an identical profile, the attacks are designed to mimic legitimate traffic, and there are a lot of them. It takes time for both the techs and their hardware to learn enough to block the crap, while still allowing legit traffic through. No system is 100%.
A better communications system needs to be in place yes. Given their size/price level. But to expect that others are going to be better...well, let me assure you there are worse out there...much worse.
Putting this another way. What do you want?
A tech to take 20 minutes to answer your "what it the problem" or 20 minutes to fix said problem?
This is also about priorities. DINIX has a thousand, couple thousand? customers. 5 minutes each is a LOT of time to not work on the problem. They can spend 5-10 minutes personalizing each response, or they can do cut-n-paster 'update everyone as fast as they can so they at least know we saw their cry for help and can get back to fixing the big problem' replies that only take a minute each.
I'll take the 'cut-n-paste' for the moment, and will then wait on the indepth reply later.
It is also about priorities.
Network issues come first, because if the network is funky, everyone suffers.
Dedicates, Colos are next due to the number of sites usually contained on them, with VPN also in that mix.
Individual domain problems are last.
This is battlefield triage. DINIX doesn't (like the great majority of hosts) have a huge tech staff. They have to prioritize and time-manage.
Yes. The more segments you have, the less chance of a complete business failure. It takes time to build that far though. Right now, we use DINIX for our servers, and Rack9 for some high-intensity sites. We've looked at others but so far the cost/benefits ratio hasn't been favorable to us. We're also looking at coloing our own DNS servers (1 on each coast) to add better redundancy.
|
| Posted by bithost(NET), 11-28-2004, 06:58 PM |
| Hello all,
This just in from Eric, the Dinix Support Desk Manager:
I should mention that I also received an e-mail from Alex. He confirmed that the setting up of filters and network management was as I had suggested -- the E^Deltacom engineers were manually adding rules as each flood came in. It was very tedious and labor-intensive.
The trouble is, DNS queries run on UDP. Hence rules put in to block UDP attacks also affect connectivity, lookups, mail, etc. The filters which affected DNS queries were removed at about midnight, and service has been smooth since.
I hope this helps to answer the qeustions put forth in this thread.
I also wanted to thank Kaith for the explanation, it's excellent. Kaith and I have been through a DDoS or two in our years hosting at the same d.c.'s ... They're very frustrating and I completely empathize with anyone going through one.
Bailey
Last edited by bithost(NET); 11-28-2004 at 07:01 PM.
|
| Posted by Christine B., 11-30-2004, 05:06 AM |
| Is it normal to be unable to download files over 100 mb? I've tried 3 times now and FTP keeps shutting down.
And sometimes my sites don't load.
Am I losing more mail?
Is this more of the same?
|
| Posted by cesarmx, 11-30-2004, 10:40 AM |
| Network SLA
**After 24 seconds, Dinix will refund the customer 5% of the monthly fee for each 30 minutes of downtime **
does not want to compensate in lost time says that:
Service failures are not covered by the SLA. Our SLA covers Network outages. The network within our control has not gone down this month
Best Regards,
Douglas Kuntz
Server Support Tech
To somebody there are refund the time to him outside line?
|
| Posted by Christine B., 11-30-2004, 07:12 PM |
| Service failures are not covered by the SLA. Our SLA covers Network outages. The network within our control has not gone down this month
Best Regards,
Douglas Kuntz
Server Support Tech"
Hahaha! That is just hilarious. I can hardly believe that's for real.
BTW - last night's outage was SCHEDULED - but Dinix didn't bother to notify the customers.
I believe that something is SERIOUSLY wrong with Dinix.
If I was working at Dinix, I'd be looking for a new job.
Is it cashflow problems? Management problems?
I don't know.
There's no point to arguing over refunds - I vote with my money and my mouth, well, actually with my keyboard. Dinix won't be missing my $75/month, they're sold out as it is.
If enough people don't put up with crap - some hosts will take notice and start treating their customers with respect.
I don't think I've EVER had a host not notify me of scheduled downtime - been doing this since 1995.
|
Add to Favourites 
Print this Article
Also Read
