EveryDNS News
Posted by davidu at 11:42pm PST on 03-01-2008
Latest Update:
We just had a mini-hackathon as a part of a larger overhaul and have released AAAA support for all users. Have a good time and enjoy your transition to IPv6. :-)
|
Posted by davidu at 7:42am PST on 02-10-2008
Our primary webserver in San Diego died this sometime yesterday making the site inaccessible. DNS was served from our servers around the globe during this time with no interruption. We have a backup webserver in San Francisco but a few things made transitioning to it nearly impossible. To add to the frustration, Level 3 has removed all on-site staff from their San Diego datacenter facility and I had no way to get someone down there quickly to fix it. We will be taking steps to resolve all these issues shortly. I sincerely apologize for this website outage. If you'd like to speed up the process for us deploying some new hardware, feel free to donate. :-)
|
Posted by davidu at 4:43am PST on 12-23-2006
I have enabled TXT record support for donators tonight. This is my way of beta testing it. Once I know it's solid, I'll open it up for everyone. Happy Holidays!
|
Posted by davidu at 11:05pm PST on 12-03-2006
EveryDNS suffered a large Denial of Service attack starting on Friday, December 1st. We have taken steps to mitigate the attack and service was restored within a few hours. There will be a more complete writeup posted on Monday, December 4th, 2006. This was our first major outage due to a DoS attack in quite some time. We are currently forming a plan to enhance our architecture to better deal with such a situation should it occur again. We take actions like this seriously and are working to find those responsible.
|
Posted by davidu at 12:52pm PST on 9-28-2006
Our webserver just got way overloaded and was pretty hard to access over the last couple hours. I've rebooted the machine and it's doing fine. I'll monitor the machine and we're going to ensure that it's stable. We will also ensure a failover webserver is properly configured in the future at a different location. There was no DNS outage during this problem as the web server does nothing besides serve webpages and do some email for us.
|
Posted by davidu at 9:19pm PST on 3-24-2006
All systems functioning normally. As a side note, our website can be accessed using a secure connection just by using HTTPS instead of HTTP.
|
Posted by davidu at 12:11pm PST on 3-23-2006
The DoS has been mostly mitigated. Hopefully things will be quiet for a bit... :-/
Summary: Throughout this attack DNS continued to be served although there were initial periods of latency across some of our nameservers. There were periods when we made our website inaccessible while we moved some services around or attempted to get a clearer picture of the attack. This was done to ensure continued operation of our DNS servers.
Here's an edited (for private data) chart of traffic sent to ns1 in San Diego over the last couple hours.
It also appears I am also being joe-jobbed with the EveryDNS URL being used in a bunch of spam sendouts. I assure you I have nothing to do with the spam. For those who don't know, a joe-job is when a spammer sends out spam with your email address or website in the spam to make it look like you're responsible.
If anyone has any information regarding those responsible for this annoyance I would like to know.
|
Posted by davidu at 10:49pm PST on 3-22-2006
The DoS has not stopped, unfortunately. I have taken steps to mitigate the damage it is causing by shifting services and servers. DNS resolution (while potentially a bit latent) should be fully operational. Please don't hesitate to email support or me directly with any issues you are seeing.
For those looking for a more technical description of our problem here is what I'll say publicly: It's roughly 36-40mbps consistently across many of our publicly facing nameservers. It isn't the bandwidth that is hurting us but the high packet-per-second rate and the spoofed nature of the UDP packets. Those of you familiar with DNS will immediately understand why this is non-trivial to filter. I along with the rest of the EveryDNS team will evaluate long-term strategies to ensure this doesn't recur.
It's going to be an expensive month and this not what I enjoy diverting our resources and money towards. (Less money for beer!)
If anyone has any information regarding those responsible for this annoyance I would like to know.
|
Posted by miked at 6:41pm PST on 3-22-2006
We are currently experiencing a denial of service attack against our nameservers. Measures are being taken to mitigate the attack and we expect to be fully recovered soon.
|
Posted by davidu at 5:06am PST on 12-28-2005
We have surpassed 78,000 users. Let's see the same reliable service continue in 2006! Happy Holidays.
Thanks!
davidu
|
Posted by davidu at 11:46pm PST on 04-07-2005
I'm currently investigating some slowness in our master DB server. My hunch is that a query that used to be fast is now slow. Since it's never been optimized there should be plenty of room to fix it. Expect it to be resolved shortly.
Update: I have slowed DNS updates to once every five minutes while I repair this. I will move than back to the old value when this issue is resolved. For most cases, this won't affect anyone.
Update: A few rewritten queries later and we're rockin'.
Thanks!
davidu
|
Posted by davidu at 5:42pm PST on 12-17-2004
Well, it had to happen. I made a mistake and took our website down for about an hour today. Of course, every nameserver was still updating, parked and webhop'ing zones were still online. Just the actual www.everydns.net page was down. We've very sorry and in addition, I feel sheepish it happened. Like most mistakes, this one was fully human error on the part of this human. :) Have a happy holiday season!
|
|
|
