Instability in web site very slow response from DASP - Need some way to get escalated support

A while back I was very upset at DiscountAsp. I've been with them since 2007. Then I moved my services off for a short time to Azure, what a mess that was. So I came back and now I'm looking for options again.

This morning, no code change, no additional users nothing different and my site is unstable because of this error:

An operation on a socket could not be performed because the system lacked sufficient buffer space or because a queue was full. Sounds bad. I would be worried about our code, but we haven't pushed a release in 2 months. Have not added new users to this service.

My last response from support was at 11:30am it's now 14:35 pst. Nothing, nada from dasp. My customers (at least they are my customers for now) they are calling. All I can say is that our hosting provider DiscountAsp is working on it.

We run our system under load on internal servers before pushing to dasp. We are not seeing this error. Has anyone run into something similar. The web site that is returning the error is a WCF web service running on .net 4.0 with integrated pipeline mode.

Any and all help is appreciated.

 

We are experiencing exactly the same problem since about 11am ET this morning. We opened a ticket about 5 hours ago but still nothing....very upsetting.

Our application which is hosted on DASP attempts to make an outbound API calls and we get this response sporadically..

"Unable to connect to the remote server
An operation on a socket could not be performed because the system lacked sufficient buffer space or because a queue was full [2607:f0d0:3:8::4]:80"

 

This is the first time we have seen this error, not the first time for a slow response. In general I really like DASP. One of the techs really helped us out with an OS Ticket setup issue, tools are good, support staff generally good, site is reliable.

I guess my main issue is that some people run applications on shared servers where it does not matter if it is down. That is not the case we us, we run a web hosted point of sale system. So you can imagine what happens when things go down. Luckily DASP has been generally great -- they don't go down. When they do go down though, FAGETTABOUTIT. We are all lumped in together and are probably looking at hours and hours of down time or intermittent issues. That said it's a shared host that offers good value. I really can't deal with the risk to my business reputation though. So once again I am thinking of leaving and may leave.

If I had access to higher priority support I would be willing to pay it and would be more comfortable staying.

So looks like they just reset the web server as my monitoring system went down when the site went down and now that we are up we have not had any errors for about 15 minutes. Fingers crossed.

Ticket logged at: 10:32
First non-automated response: 10:40
Likely resolution: 3:35pm (hopefully) --> 5 hours instability partially down

 

Michael D, when outbound API calls fail it's rarely a problem on the server here. We don't block or otherwise throttle any outgoing calls or ports, etc.

What we've found recently is that some APIs call on sites or applications that are hosted on round-robin DNS. It's a method used to load balance busy servers. It isn't used much anymore, but some big sites do still employ it as a load balancing measure.

Some of these sites use unusual methods to manage DNS, such as removing an IP from the round-robin group if there is a problem on that particular IP. When that happens, a call to that IP from your application hits a dead end and you get an error. The error persists due to DNS caching. But that cache should be cleared every five minutes.

Joseph Vandertol, I don't know if your application is making an API call to another server, you don't say, but if it is the same answer applies. Since you haven't posted again, I assume it was an unrelated issue that looks like it's been solved.

As far as support priority is concerned, there isn't any higher priority than a server-wide issue. But I just have to clarify here that there is never a case where a server-wide issue would go ignored for hours. Our monitoring wouldn't allow us to ignore that kind of problem even if we wanted to. But it can sometimes take a bit of time to troubleshoot more unusual problems. I understand that it can seem like nothing is happening from your end while that kind of troubleshooting is going on.

What would you like to happen in a case like that? It sounds like you would like more frequent updates from us via email. The support staff does update the outages forum here if it looks like something is going to be ongoing, but if you think there's a better way to handle it, we're open to suggestions.

 

Both the web site and the web service are hosted by DASP. Objectively it took 5 hours to resolve. The reason for the issue per Dasp support was an overloaded server. Should it take 5 hours to figure that out. DASP monitoring systems as you say should have identified the issue before I and unfortunately my customers did. I only know what happened and that was 5 hours to resolution.

I am a BIG fan of DASP, but all support requests are not equal. Until there is a way to identify the impact your customer (me) is experiencing in the support ticket priority things like this will continue to happen IMHO.

You guys have an outstanding service, great value. For small companies like me that run mission critical systems the time to resolution is kind of scary.

 

I also want to reiterate that Discount ASP is a great company. I realize we are running on shared servers and using an inexpensive service. I really believe that if you were to bifrucate support requests and offer a paid support option that Discount ASP would address an underserved market: companies like mine who cannot afford downtime, which are growing but who would like to delay the move to higher end services. Simply say these requests will get 2nd tier support and offer more frequent ticket updates. Limit the number of ticket so you dont get slammed. I would rather stay with Discount ASP, but I also want to sleep at night.

 

We've tried to avoid introducing any kind of "premium" support, if only because doing something like tiered support gives people the impression that your normal support is somehow lacking. I understand that right now you think it is lacking, but I'm sure you understand what I'm saying.

We try to provide the best service you can get as far as .NET hosting is concerned. But honestly, a "mission critical" application or site living on shared Windows hosting has to make some concessions to the platform. There is not always an immediate path to resolution (and there will always be at least some downtime for O/S updates and upgrades) - that's just the nature of shared hosting. We eliminate as much of that as we can, but it can't be eliminated completely.

Having said that...we are very close to making a new service available. "New" meaning technology quite different than the current DiscountASP.NET platform. A high-availability service that will never "go down" for O/S updates and upgrades, and one that is flexible enough to meet almost any site's needs (including the availability of reserved instances, where your site doesn't share resources with any other site).

If it sounds a lot like Azure, it is. It's built on the same platform (but we host the hardware and data - not Microsoft), and it's going to be less expensive (and way less confusing) than the same services on Azure. We've built the new service to accommodate people just like yourself, who might be outgrowing shared hosting or frustrated by its limitations, but don't really need (or want) the complexity and granularity of something like Azure.

I think you might like it. It's been more than a year in the making, but if you hang in there with us just a little longer, you'll see it very soon. Our existing customers will be the first to get an opportunity to move to the new platform, and we're planning on providing some incentive to ease the inconvenience of transition/migration. We'll let you know when it's ready to go.

But for everyone who is happy with DiscountASP.NET as it is, don't worry, we're not going anywhere!

 

Great to hear! Our experience with DiscountAsp has been great in general. While your service is a shared server environment it has been consistently reliable. If you bring that same quality to this new approach I think it will be a great service.

 

Joseph, we always bring the quality!

A lot of you will like this though, I'm pretty confident about that. Some pretty cool stuff happening.