A number of us were curious about the severe outage.

By my reckoning, it hit late on 16-OCT and did not recover until this past weekend.

At least one week.

Anything that can be shared?

Vbulletin database errors can be difficult to fix a 10 day outage is longer than most but other forums I'm on have had outages of several days and LifeCourse is a small operation and this forum is a minor part of. Neil Howe tweeted about the outage.

I shared with John X the fact that I used to co-manage a forum (not on VBulletin, on a different provider's network). It was a geopolitical forum that had a lot of subject matter that, shall we say, was not very flattering vis a vis the SCO countries. We were one of the first forums to be targeted by Russian and Chinese trolls and hackers.

Eventually, with the apparent help of an insider at the provider, they hit the provider with an attack that not only blew away most of the data in the production database, they also managed to severely damage the backup.

Originally Posted by XYMOX_4AD_84

A number of us were curious about the severe outage.

By my reckoning, it hit late on 16-OCT and did not recover until this past weekend.

At least one week.

Anything that can be shared?

Uh, I think in this case it was my EMET induced Firefox crash that did it.

EMET is a security tool I use to catch data induced data exceptions that overwrite executable machine code in specific programs / stack overwrites/DLL overwrites , etc. Vbulletin here stopped working right soon after my Firefox crashed during a post reply.
So my theory is that my Firefox session wrote to some database tables, then crashed before finishing updating all needed tables.
So... that's my theory. Below is an example of some code that can elicit an EMET exception

void fubar (char *retval);
{
str = "0123456789";
strncpy(retval,str,10); /* NO! we're returning a non null terminated string right here */
}

void fixed(char *retval)
{
char str[11]="0123456789"; /* ensure space for null termination */
strncpy(retval,str,11);
}


Something to share? Yeah, lousy programming. I have 2 Mafia Wars sessions going in a Firefox browser which run Java script and Flash along with another tab on TFT forum. Something manages to throw an EMET exception which causes EMET to terminate Firefox. Vbullentin has a bug where it can't handle a browser crash in mid post.

Originally Posted by Ragnarök_62

Uh, I think in this case it was my EMET induced Firefox crash that did it.

EMET is a security tool I use to catch data induced data exceptions that overwrite executable machine code in specific programs / stack overwrites/DLL overwrites , etc. Vbulletin here stopped working right soon after my Firefox crashed during a post reply.
So my theory is that my Firefox session wrote to some database tables, then crashed before finishing updating all needed tables.
So... that's my theory. Below is an example of some code that can elicit an EMET exception

void fubar (char *retval);
{
str = "0123456789";
strncpy(retval,str,10); /* NO! we're returning a non null terminated string right here */
}

void fixed(char *retval)
{
char str[11]="0123456789"; /* ensure space for null termination */
strncpy(retval,str,11);
}


Something to share? Yeah, lousy programming. I have 2 Mafia Wars sessions going in a Firefox browser which run Java script and Flash along with another tab on TFT forum. Something manages to throw an EMET exception which causes EMET to terminate Firefox. Vbullentin has a bug where it can't handle a browser crash in mid post.

If they cannot handle simple client behavior such as this, their architecture is unbelievably weak. I'd also say they are prone to be taken down by everyone and their uncle.