Due to the server failure backups from May 16th had to be brought up as most of us have noticed. Digitalpoint usually has its pages indexed pretty fast. Maybe it would it be possible to grab content from different online caches and restore threads created between May 16th and June 1th? I don't know much about XenForo but this is something that would be possible and with DPs server power (judging by a thread containing the server specifications) and with some PHP it would definitely be doable.
Excellent question and one I'm sure many of us are now asking. I am *very* interested in @digitalpoints response to this, as it would seem that it ought to be able to be retrieved?
Threads are just a tip of the iceberg, there is no way to restore lost data if you don't have every piece of the puzzle IMO.
The cache contains all the information that we're currently seeing right here. The username of the user that had posted the message including the content; exists in these caches. These could be grabbed and restored, after all, the threads and the posts are the largest losses of the rollback. I understand that it would be a whole lot of posts to grab but what about focusing on the original post of all the threads instead? That would have the discussions going on the threads again.
That might get the whole thing back to some semblance of what we lost, I wonder how much work that would involve?
The problem is, that there is also a lot of meta information that Google cannot see. I think that it might be possible to create a static archive of created threads, but that's it. There is no way to import archived data back in the database without affecting the forum's stability. If you take a look at a database structure of any sophisticated forum, you'll know what I mean.
Please elaborate about the meta information. I was thinking about having the script fetch the username (or perhaps the user ID) together with the original post. Then a new thread would be generated with all necessary information (including the meta info) that's usually generated when a thread is manually created by a user. The thread is created with the user's ID and contains the post fetched earlier from the cache. Sure, the search engines might detect it as duplicate content as the date was changed. This could be solved by fetching the date and adding that to the post.
I had thought about it... but like Devtard pointed out, there's a LOT more to it internally than just the post table... in fact, I actually have an "in-tact" post table (but that was the only table) without needing to go to a backup. But then you start dealing with posts on threads that are missing as well as all the related data on user accounts related to those posts. You also would more or less need to stop new threads/posts from happening while the post recovery was running, which would first have to be coded and then run (which would probably mean the site being down for many days while posts were restored).
Would it be a chance to at least have the original post of the threads back? That shouldn't cause too much downtime either.
Just adding the original post should really speed that process up. In my opinion recovering the original posts (if possible) is definitely worth doing.
Think of the headaches such an attempt to rebuild all threads that were started in the past. Suppose I sold and transferred a site to lucky bidder 1. Then thread was reposted at Lucky bidder two thought he bought site. In my mind there is a time to let sleeping dogs lie.
I'll give it a shot. The user could simply close the auction right after the recovery. Or in the worst case scenario recovery of sales threads would be avoided.