Server down 3 times on first day

[littlelizard]

I'd be happy to get a trial VPS account. My sites are down again.

Dave,

We all hate downtime. It hurts your business, it hurts ours. I got involved here because I wanted everyone to see how much importance we put on delivering and over-delivering on our promises. We will get it resolved.

We are also considering offering existing customers a discount to trial VPS services (not intended as an up-sell but as an alternative). We're very open to feedback if anyone wants to comment on what they'd like to see.

Matt

[DaveTheShark]

Web server just timed out again.
Ran test loads from Amsterdam and New York and page load time is apalling.

Suspect network capacity issue as mentioned before. (No one ever replied to me)

Logged ticket with support, will keep you updated.

[DaveTheShark]

Support say
"As we previously stated, we didn't notice any problems with the server you are
hosted on. So we can also suspect the problems you saw were related with some
network congestions between your ISP on our server (e.g. somewhere at the Telia
level). Traceroute results will help us to determine whether it true or not."


Yes I suspect the same thing. I wish you'd do something about it though. No point paying extra for a business server if your network connection is at the end of an elastic band.

[Dennis A.]

Dave,

You actually did our technical staff job and identified the issue. Please accept our apologies for making you go through this.

However, I'd identify the primary issue here as lack of proper communication. You tried to contact our support in live chat (which is 1st level support, they hardly could diagnose the real problem themselves and gave you pretty much standardized reply), then you went to the forum, which is not an official support channel, and not closely monitored by support team. Forum is a client-to-client support channel indeed. Finally you submitted a ticket, but were referring to the issue as well-known to the entire support team, which resulted in confusion in early ticket posts.

Our internal support procedures assume that support ticket system serves as a primary channel for solving more or less advanced issues. Live chat / phone support team (1st level support indeed) is supposed to recommend clients submitting a ticket if the problem is non-trivial. Ticket system is actually the only support channel monitored by 2nd and 3rd level support teams. You could have saved weeks of your time and avoid all this ridiculous story by going to the ticket system directly.

Now back to the issue. Echoing the previous posts in a sense, but this is mostly to summarize. We don't have direct network links to Telia, nLayer and Mzima in Atlanta. We have our own floor space in GNAX facility and maintain our own network infrastructure, but external network connectivity is provided by GNAX, this is part of the deal. Your server's IP addresses belong to AS16626, business3 doesn't use AS3595 traffic. I've raised uplink saturation issue with GNAX, but honestly, their options in case of uplink saturation are pretty limited. They can try playing with route metrics and redistribute traffic, or they can try to identify possible traffic anomalies and block abnormal traffic sources, if any. I'd suppose there should be some abnormal traffic, their physical Telia throughput isn't that limited. We'll keep you updated in the ticket.

[DaveTheShark]

Thanks for the reply Dennis.

If I may respond.

Yes I intially contacted you via 1st line support. Despite my insistance that this was an on-going issue the response was very much "it's not happening at the moment we can't help you any more." At no point was I advised to raise a ticket. I wish I'd known this earlier. Don't forget your CEO has replied to this thread and could have taken further action.

I realise that the forum is not the place to get support, however, I felt I had nowhere else to go after being given the brush off by support. As a result of these posts we seem to be getting somewhere. I also congratulate you on allowing an open forum. It was one of the reasons I picked WBH in the first place.

Support have a further update since your post.

"We have received reply from our upstream provider. There were multiple large
bursts of incoming traffic over the last two days due to DDoS attacks which were
mitigated by now."


OK, makes sense. My point remains. In 2 years with a budget provider I have not experienced this level of network timeouts. There was a point yesterday, all be it brief, when it was impossible to connect to the web site.

[DaveTheShark]

Just experienced another period of slow response.

Load times from New York also very bad.

Looks like the same problem again.

[DaveTheShark]

Following my last post performance remained poor for about 5 hours.

Then all of a sudden we're back to normal. All of the resources hosted on my web host server are now being served less then 0.2 seconds.

Received this update from support.

"We have been experiencing some problems with network connectivity during a DDoS
attack, however this problem has been resolved and your site's load times are 1
to 2 seconds according to site-perf.com. We are trying to prevent such situation
from occurring in the future and are doing everything we can for this."

[JohnMC]

just wanted to say i found this thread interesting given my current issues and down times and i was prepared to read the messily 4 pages but the first page with alan b has prompted me to quit reading and just recommend you be less critical of peoples writing and less assuming in case a post was missed. Given this, my current frustrations with my hosting (elsewhere posted) and a very defensive response i found from an admin/support (play it safe) on someones thread i found, im loosing hope fast.