This announcement will shed light on the general situation with XJ, explaining the details in-depth
regarding what exactly happened, what we've tried, and what we've eventually come to.
It all began on April 13th, when our hosting provider performed maintenance work. This resulted in us
getting logged out of the account and being unable to log back in manually. We tried the account retrieval
procedure, but the 2FA message was delayed to the point where the token was no longer valid by the time we
received it. Our email provider showed us proof that the issue was caused by kimsufi. We explained the
entire situation to kimsufi (we were fortunate to get in touch with them once), but they refuse to accept
that it's their fault and have shown no willingness to help resolve the situation. We thought that another
solution could be to disable the 2FA feature entirely from our account. However, reaching out to the support
line has become a huge problem because the only official support channel is via creating a ticket using an
account we don't have access to. We have tried emailing them, but they never respond. The support hotline
usually has a robot instructing us to create a ticket via an account. Seven days ago, we sent another email
requesting to disable 2FA from our account, but we have yet to receive a response. This is just a brief
overview of the story, as we have lost a lot of time due to communication issues. If I were to explain
everything in more detail, it would be quite lengthy. The situation required the involvement of multiple
parties, and most of the time, it significantly delayed the process.
Some words from 🇮🇷 PeKz regarding technical aspect and revealing new information:
Hello there, we're going to go into the actual problem in greater detail here.
After dealing with those massive DDOS attacks for two weeks, we thought we'd finally be able to enjoy
some peace and
quiet. But that was incorrect!
Technical explanation about old server:
After OVH went into maintenance mode for 1-2 days, I discovered that our nginx service had been
disrupted. The first thing I tried was restarting nginx's service, however systemctl reported "Failed to
start nginx:
Connection timed out."
So I believed there was something wrong with the nginx configuration, so I checked all I could, but
everything seemed OK, and yet nginx couldn't be restarted. So I ran another systemctl command (systemctl
status mariadb)
simply to see what would happen, and another timeout!
Every core linux command I tried failed, including package management, so I tried restarting the server
to see if that
would help. Even the "reboot" command did not work and returned a timeout error; as a result, I logged
off of the
server. After a few hours, I attempted to SSH into the server again, but discovered that it had been
totally shut down
and we were unable to restart it (ShoCk stated the reason above).
"Why won't you just restore backup on another server and point domain to new server?" you may wonder.
The problem is that the entire stack (from the e107 itself through php, mysql, nginx, and so on) is so
ancient that we
can't operate it on newer operating systems. They will not support older versions of the stacks I
specified. thus I
tried alternative solutions, such as using docker and getting old stack via the web (rather than the
real OS
repository), but I couldn't make it work, therefore we basically can't have old XJ and its data without
the old server
showing up...
What is our plan about future:
Right now, our only option is to start from scratch, and this time, there is a good chance that we will
go full
opensource on everything (except the database, of course:D). At this point, we have some of the new
design actually
coded in React thanks to our designers/front-end coders (I won't name anyone because some of them worked
on this before
my time as CTO, and I don't want to disrespect them by naming only some of them). I was considering
using Golang for the
backend, but since it would be entirely dependent on me, we decided to skip it in favor of Typescript
(we still need to
determine whether to use ExpressJS or NestJS).
That's all I have to say about our future plans, but I'll keep you updated.
Unfortunately, it seems that luck is not on our side this time, and there's very little room for luck left. We may attempt a few more options as we move forward with plan B, but it will certainly require some luck. So, plan B is kreedz.com, which has been an ongoing project for a while and is currently being worked on at an accelerated pace. We won't provide specific dates just yet, but it's definitely in the works.