Dread suffered a crash that resulted in the loss of two weeks worth of data, including account changes, posts, comments, and even new account. HugBunter will have the forum up and running again in the next few days, according to a status update posted on the DreadAlert subreddit.
Attempts to visit Dread last week often resulted in a 502 error or short periods of complete accessibility. One of the Dread staff members, Paris, posted a post about the forum’s instability and issues handling Dread’s rapidly growing user-base.
In one comment on the post, Paris wrote:
We are currently sitting at about 150-175GHz of raw CPU power with the core and the front servers. The vast majority of the CPU usage is on the front servers handling the building and breakdown of all the trash circuits caused by the DDOS attack. In total we have like 50 to 60 CPU cores working constantly just to keep us up.
Its a lot. The server cost is very high. The infrastructure change isn’t really going to reduce the total cost but will make it more economical in the longer term but much higher value.
According to Dread’s creator, HugBunter, the server experienced a crash that resulted in the corruption of the data on the server. The most recent backup is two weeks old. HugBunter also explained that the crash had taken place at one of the least ideal times; he has been working on “new server infrastructure to handle a significantly higher load and provide complete uptime across the board, not only solving our recent 502 error issues, but also ensuring any future Tor DoS attacks would be futile in rendering the service unavailable.” Backups of Dread seemingly slipped his mind. He promised to automate backups in the future.
When Dread comes back it will bring some changes, including “includes bug fixes, optimizations and new features.” He is changing some of the rules as well, although the changes are not likely to have a significant impact on the majority of Dread users or their ability to use Dread. His complete message is available below and a signed version is available here.
Hi all, Been a while since I’ve necessarily needed to provide an official status update, but we’ve been down for almost 2 days now. This is fortunately not a DoS attack of any kind, however I do have some bad news to bring. I’ve warned for the need of a large update to the platform for a long time as it has became extremely unstable as it has significantly grown over time. Since I rushed the initial development there have been many factors that have caused issues for us behind the scenes on a regular basis, whilst appearing everything is running perfectly to the end user.
Yesterday, our core servers gave in and had a fatal crash which left the whole web server data including the database completely corrupt. We do usually have regular backups made, however this has really caught us off guard since we were in the process of building a new server infrastructure to handle a significantly higher load and provide complete uptime across the board, not only solving our recent 502 error issues, but also ensuring any future Tor DoS attacks would be futile in rendering the service unavailable.
At this point it appears that we have in fact lost around 2 weeks of data from the platform. So any posts, comments, new user accounts, etc within that time frame will now be gone, I can only apologize for this and reassure that this will not occur again as automated backups will be made on a very regular basis from here on out. We are finalizing the new server infrastructure right now as well as providing an update to the platform which includes bug fixes, optimizations and new features. There are also rule amendments that will be going live, which I will discuss at the time.
I cannot provide a specific eta right now, however we are hoping to be up and running within 48 hours at the very most.
Thank you for your patience and understanding.
Source: DreadAlert on Reddit