We received some downtime about 6 hours ago. This was due to an attack. However, the downtime was only for 20 minutes before we bounced back.

For some odd reason a configuration anomaly slipped into our firewall configuration today that had to do with the throttling of our traffic (traffic shaping configurations). We only realized this once the traffic data did not match the general average.

The good news is that we’ve fixed this and things will be going through our pipes as usual =)

Our site has been up and down today because we’re receiving a DDoS attack. We’re working on resolving this.

We have just implemented weighted load-balancing on our servers. This is especially important for a setup such as ours because each server has different services attached to it and therefore different loads. Although we can design the setup so that an equal amount of load goes to different services, it’s simpler to weight the requests as they’re distributed around.

Through distributing our requests by using higher weights on servers that are less loaded by nature (vice versa for more loaded), we can fairly distribute these requests over.

Ultimately, it means that we are able to make better use of resources and therefore increase overall capacity to our site.

We’re pleased to announce that we have been receiving good uptime for the past few days thanks to the work of our server admins.

A huge effort has been put towards filtering attacks that has elapsed to GBs of requests sent to our pipes. And we are pleased with the results so far.

The other drain on our services is crawlers. These are programs that spider our site for posts/topics so that they can be automatically copied and posted on the operator’s own site. We consider these programs to be pests and we will ban the IPs and the accounts of users who use them.

There is a 10% to 20% impact on our servers because of these crawlers, if there is one thing to blame for increase in serverload, it would be them. This prevents genuine users from joining our site and contributing to our community. They also cause a financial impact. These crawlers do not load advertisements but most importantly, the advertisements are not seen by genuine users. This means lower revenue to contribute to our growing bandwidth and hardware costs.

Ultimately, we will do everything in our power to block and filter these pests. If you are one of them, do the right thing and stop using it

We have updated our blog script to get the latest DNS information off our DNS servers. So the next time we do change DNS, the information will update within two minutes with the new IP.

Please take note that this information is only available for the Main server IP and it is not live updating the Image server IP yet due to the nature of our static server setup.

So in the 7th installment of updates, we talked about fixing a bug that had caused 20 minutes of downtime. We’re pleased to say that this bug has been fixed because we didn’t receive any further downtime during the same period today.

However, we are still receiving attacks and we are working hard on setting up advanced filters to counter these attacks.

We realize we are down at the moment. We are working on debugging the issue.

Edit: Looks like we’re down because of an incoming attack. Jeez.

Edit 2: And… We’re back.

We had 20 minutes of downtime today just three hours ago. We have already implemented fixes for this and it should decrease the chance of downtime. Although we think that it will fix it, we can only confirm when it happens again tomorrow.

Fingers crossed

We acknowledge that there was an issue 5 hours ago (relative to the time of writing) with our site. We have already implemented fixes to mitigate that specific issue and we will be analyzing logs in the future to see if this issue arises again.