As all of you have may have already noticed, our server hard drive crashed and we were unable to recover anything from it.  Adding fuel to the fire, we did not have any recent sql databases of the forum saved off-site.  This huge error on our end has costs us 5 months of posts, pictures, users, and so much more.  We are very sorry for the loss, it will not happen again.

Sigh. That sucks. I thought servers were usually on a RAID where if something went wrong you can get anything back from a redundant drive.

You guys should talk to Google.  They basically cache the internet.  I'm looking at a copy of the Cincinnati Streetcar thread with posts from yesterday:

 

http://webcache.googleusercontent.com/search?q=cache:e8nqZMjLUHwJ:www.urbanohio.com/forum2/index.php%3Ftopic%3D18957.20370+&cd=1&hl=en&ct=clnk&gl=us

 

 

I have no idea how that technology works, whether they literally grab the content of every page or somehow have the database in a usable form.  But it might be worth a shot.

Also check out to see if archive.Org had backed spidered any of the pages.... Couldn't restore the stats but could show what we posted.

OK everyone...breath deep. Also no worries richNcincy, stuff happens. I encourage everyone to make a list of the threads you feel we have lost since Feb. Whatever market you are in you know what is missing. Hopefully over the next couple of weeks we can get the missing threads back up and running again. So if you don't see a particular thread start it yourself or PM the person you know may have started it and remind them.

FWIW, since the site was resurrected, I get logged out every time I leave.

I have a new 4TB WD MY BOOK I would be willing to donate.

Whoa...time warp!

I found myself inexplicably more productive today with work than when UO is up and running.  Hmmmm.....

"we can rebuild it. we can make it better than it was..."

Aw come on, we all know it's just the mods' way of clearing the air without looking overbearing.  :evil:

You guys should talk to Google.  They basically cache the internet.  I'm looking at a copy of the Cincinnati Streetcar thread with posts from yesterday:

 

http://webcache.googleusercontent.com/search?q=cache:e8nqZMjLUHwJ:www.urbanohio.com/forum2/index.php%3Ftopic%3D18957.20370+&cd=1&hl=en&ct=clnk&gl=us

 

 

I have no idea how that technology works, whether they literally grab the content of every page or somehow have the database in a usable form.  But it might be worth a shot.

 

How did you get that? I'd like to use that to start rebuilding threads.

1) Go to google.com

2) Search for site:urbanohio.com cincinnati streetcar.  In the results, you will see a blue hyperlink that says Cincinnati Streetcar News - Urban Ohio.  Under that is a green hyperlink that says www.urbanohio.com › ... › Transportation › Mass Transit‎.

3) Click the green down arrow next to that green hyperlink.

4) Select Cached from the menu that appears.

 

This may not be helpful if you're rebuilding whole threads, but you could certainly find anything that you know was important with this method.  The data still exists somewhere out there, so Google (or some other app) may have a way to rebuild it using their cache.  But I've never tried anything like that before.

^Looks like you could conceivably view a whole thread by searching for the page number in brackets.  So, you could google:

 

site:urbanohio.com cincinnati streetcar [498]

 

and that would give you page 498 of the streetcar thread.  Then you could search again with [499], then with [500], etc.  Even for the streetcar thread, that's a ton of work, let alone the whole forum.  But it's a way to navigate through the cache of every thread.  Just start with the site URL, a search term that you know will find the thread, and page [1], since every thread has a page 1.

 

Nice to see the forum back up today.  I started to develop UrbanOhio DT's yesterday.

this link might be helpfull

 

 

https://code.google.com/p/warrick/

 

Warrick is a free utility for reconstructing (or recovering) a website when a back-up is not available. Warrick utilizes the Memento Framework (http://www.mementoweb.org) to discover archived versions of resources from web archives. The resources are gathered to provide a single collection of files.

 

Warrick is developed as part of a research project at Old Dominion University. Warrick is not affiliated with the web archives. Warrick is provided as is with no warranties or guarantees.

 

Please send any emails requesting help or reporting a successfully recovered page to [email protected].

I'm very sorry to read of our loss.

 

Rich, you said it would never happen again. What will be different in terms of making regular backups?

 

I have a Linux script (command line stuff) that I use to back up various PHP/MySQL based things, including one SMF board that I run. I'd be happy to provide it to you with customization instructions. You'd need to get a Linux shell (SSH) on the same server that SMF is running on. The script saves *everything* - the entire MySQL database, plus all of the files in the SMF installation. I have used the backup files to move installations to new systems.

 

What the script does is compress everything into one big backup file than you can then download to a desktop computer using FTP. If the board ever crashed again it could be rebuilt from the backup file (a "tarball") in a few minutes. It would be a 5 minute task for some admin to log into the command shell, run the script, and then use a program like FileZilla to download the backup file.

 

My SMF board has about 70K posts. It takes about 15 seconds to run and the resultant backup file is about 40 meg. It usually takes less than a minute to FTP to my PC for safekeeping.

What type of SSD failed? I have Intel and Samsung SSDs; I picked them for reliability.

SSDs fail too. Flash memory devices (on which SSDs are based) have a lifetime of so many cycles of writing and reading. SSDs are built with huge internal surpluses of flash memory, and the internal memory controller will detect internal failures and will move data around to fresher memory. In other words they're designed to fail internally in a soft way over their life span and to present a reliable interface to the outside world.

 

But at some point in normal frequent use, the internal redundancy of memory is used up and the entire unit fails. I have anecdotally heard of workstation users (single person computers) SSDs only lasting a year or two in daily use.

I'm very sorry to read of our loss.

 

Rich, you said it would never happen again. What will be different in terms of making regular backups?

 

I have a Linux script (command line stuff) that I use to back up various PHP/MySQL based things, including one SMF board that I run. I'd be happy to provide it to you with customization instructions. You'd need to get a Linux shell (SSH) on the same server that SMF is running on. The script saves *everything* - the entire MySQL database, plus all of the files in the SMF installation. I have used the backup files to move installations to new systems.

 

What the script does is compress everything into one big backup file than you can then download to a desktop computer using FTP. If the board ever crashed again it could be rebuilt from the backup file (a "tarball") in a few minutes. It would be a 5 minute task for some admin to log into the command shell, run the script, and then use a program like FileZilla to download the backup file.

 

My SMF board has about 70K posts. It takes about 15 seconds to run and the resultant backup file is about 40 meg. It usually takes less than a minute to FTP to my PC for safekeeping.

In short, yes. That is exactly what I need. I had to use SSH to get everything back up and running so that would be perfect.

Rich, if you want me to host an off-site backup, I can give you a place to upload it or you can drop a physical hard drive off at my place.

SSDs fail too. Flash memory devices (on which SSDs are based) have a lifetime of so many cycles of writing and reading. SSDs are built with huge internal surpluses of flash memory, and the internal memory controller will detect internal failures and will move data around to fresher memory. In other words they're designed to fail internally in a soft way over their life span and to present a reliable interface to the outside world.

 

But at some point in normal frequent use, the internal redundancy of memory is used up and the entire unit fails. I have anecdotally heard of workstation users (single person computers) SSDs only lasting a year or two in daily use.

 

Yeah, I know they fail. Which is why I picked those specific brands for their reputation of reliability. A lot of SSDs have a reputation for failing long before they are worn out from writes. Especially drives with Sandforce controllers.

I guess I picked a good week to not be connected.

Rich, if you want me to host an off-site backup, I can give you a place to upload it or you can drop a physical hard drive off at my place.

 

An off-site location would be nice to have.  I could FTP the backup relatively easy.  The forum database is about 550 Mb's right now.

Rich,

Just wanted to say thanks for all of the time that you commit to keeping this ship sailing.

 

I am a tech dinosaur now, so I'll leave the hard ware conversation to you guys.

 

Somewhere a "devil wears prada" admin is wondering what happened to all of the hits to his/her site.

This is tragic.

Rich,

Just wanted to say thanks for all of the time that you commit to keeping this ship sailing.

 

Seconded.

Yes- thanks, Rich!  I'm not a tech guy, so I really have no idea about the types of work and challenges you deal with to make the site work.  Best of luck!

1) Go to google.com

2) Search for site:urbanohio.com cincinnati streetcar.  In the results, you will see a blue hyperlink that says Cincinnati Streetcar News - Urban Ohio.  Under that is a green hyperlink that says www.urbanohio.com › ... › Transportation › Mass Transit‎.

3) Click the green down arrow next to that green hyperlink.

4) Select Cached from the menu that appears.

 

This may not be helpful if you're rebuilding whole threads, but you could certainly find anything that you know was important with this method.  The data still exists somewhere out there, so Google (or some other app) may have a way to rebuild it using their cache.  But I've never tried anything like that before.

 

Thanks. Some of the most recent updates shown in cache memory are a couple of weeks old. In those cases, it sounds like those pages are gone for good.

Unfortunately, achive.org didn't seem to pick up the individual threads.

You probably already know of this problem, but just in case you didnt, the forum stops working a lot throughout the day. It also keeps logging me off.

 

Sorry  :|

^ When logging in, are you changing the drop-down to the right of the password from "1 Hour" to "Forever"?

I have tried every option when logging in and I get logged out every time I leave.

Safari on Mac

You probably already know of this problem, but just in case you didnt, the forum stops working a lot throughout the day. It also keeps logging me off.

 

Sorry  :|

 

I'm working on the current issues we're having.  I had to do the same when we first moved to the new server and this is pretty much like starting over.

Too bad the crash didn't solve the logging in problem.  Though I'm on right now obviously.

One thing it did solve for me is my inability to get on the site at all when at work. Regardless of internet browser used it would always come up as saying 'cannot connect to server.' It would occasionally do that at home for me as well and that appears to have stopped as well.

^ If you use chrome at work (Chrome Usually installs without Administrator privileges) you can load a previously cached copy typically.  But that's all lost now.  But strangely the site is working for me more often. 

If you are having issues staying logged in or are getting the unable to verify..... error, try to delete the urbanohio cookies you have stored and clearing your cache.

If you are having issues staying logged in or are getting the unable to verify..... error, try to delete the urbanohio cookies you have stored and clearing your cache.

 

CDM slacking again?  Lord!

I'm having a weird issue where my company's websense filter is blocking the site intermittently, not because of content, but because of 'network errors'. Any idea what's driving that?

I'm having a weird issue where my company's websense filter is blocking the site intermittently, not because of content, but because of 'network errors'. Any idea what's driving that?

 

It's probably the reverse DNS issues we're having.  I'm working on that issue too. 

The rDNS issues we were having have been resolved.  Let me know if you continue to have the network errors. Thanks

Has anybody checked with the NSA?

Has anybody checked with the NSA?

 

ROFLMAO!

The rDNS issues we were having have been resolved.  Let me know if you continue to have the network errors. Thanks

 

Looks like everything is working fine now. Thanks!

The rDNS issues we were having have been resolved.  Let me know if you continue to have the network errors. Thanks

 

Everything back to normal from my employer's network.

Has anybody checked with the NSA?

 

Yes. The thing to do is put in a request for a backup copy of the board.

Oddly, when I go to the main page of the board I see a "recent forum topics" listing but it doesn't indicate new posts since my last visit as it did in the past.  (I'm still logged in when I get here though.) 

Oddly, when I go to the main page of the board I see a "recent forum topics" listing but it doesn't indicate new posts since my last visit as it did in the past.  (I'm still logged in when I get here though.)

I had to remove them due to errors we were getting. I will attempt to add them back though.

It's not a big deal.  It still tells me which ones are new when I get to the actual forum pages.  Thanks for all the hard work!