Amount

BEFORE YOU POST READ THIS TOPIC

12 Jan 2017 02:13 #52017 by fats
BEFORE YOU POST READ THIS TOPIC was created by fats
Hi all,

You will notice that the site has been down for a number of hours, this is due to a major FUBAR caused by the setting up of the email server, the problems have been extensive and it caused the corruption of the site's DataBase which has meant that I've had to use a 10 day old backup of the site.

So that means that all new users and new posts and stories from the last 10 days have unfortunately have gone.

I'm deep, deeply sorry for this.

I'll update you all on any other issues over the course of today.

Again my heartfelt apologies for the outage.

Fats
The following user(s) said Thank You: AlbertC, Raa, algae2k, Dru1076, Markiehoe

Please Log in or Create an account to join the conversation.

  • fats
  • fats's Avatar Topic Author
  • Away
  • Administrator
  • Administrator
More
12 Jan 2017 03:03 #52018 by Raa
Replied by Raa on topic BEFORE YOU POST READ THIS TOPIC
Thanks SWM team for getting back what you can get back!

Please Log in or Create an account to join the conversation.

More
12 Jan 2017 14:19 #52025 by brantley
Replied by brantley on topic BEFORE YOU POST READ THIS TOPIC
What did Apache have to do with this? While the site was down, they posted an ad on the URL.

--Brantley

Please Log in or Create an account to join the conversation.

More
12 Jan 2017 17:50 #52027 by njae
Replied by njae on topic BEFORE YOU POST READ THIS TOPIC

brantley wrote: What did Apache have to do with this? While the site was down, they posted an ad on the URL.

--Brantley


Most websites are based on XAMPP

X (either W(indows) or (L)inux
A(pache)
M(ySQL)
P(HP)
P(erl)

Obviously during the rebuild Apache showed its default page.

What surprises me is the lack of a backup. Back when I was member of the stuff those happened daily, so I am curious about whether the backup wasn't there due to time-constraints or if the ones available didn't work. I'm interested because I use the same backup software on a project of my own and would like to know if it has issues.

Please Log in or Create an account to join the conversation.

More
12 Jan 2017 19:56 #52032 by brantley
Replied by brantley on topic BEFORE YOU POST READ THIS TOPIC
If we're lucky, the Wayback Machine might have captured some of the missing material.

Don't have a problem with posts on my site; there's a Yahoo group for that. But I have multiple backups for story and picture files.

--Brantley

Please Log in or Create an account to join the conversation.

More
13 Jan 2017 10:11 #52038 by Woodclaw
Replied by Woodclaw on topic BEFORE YOU POST READ THIS TOPIC

brantley wrote: If we're lucky, the Wayback Machine might have captured some of the missing material.

Don't have a problem with posts on my site; there's a Yahoo group for that. But I have multiple backups for story and picture files.

--Brantley


Unfortunately I already checked and the last snapshot by wayback machine is from Novemeber 2016, way bfore the last site back-up.

Please Log in or Create an account to join the conversation.

More
13 Jan 2017 10:14 #52039 by Woodclaw
Replied by Woodclaw on topic BEFORE YOU POST READ THIS TOPIC

njae wrote:

brantley wrote: What did Apache have to do with this? While the site was down, they posted an ad on the URL.

--Brantley


Most websites are based on XAMPP

X (either W(indows) or (L)inux
A(pache)
M(ySQL)
P(HP)
P(erl)

Obviously during the rebuild Apache showed its default page.

What surprises me is the lack of a backup. Back when I was member of the stuff those happened daily, so I am curious about whether the backup wasn't there due to time-constraints or if the ones available didn't work. I'm interested because I use the same backup software on a project of my own and would like to know if it has issues.


I'll have to defer to Fats about this, but the problem is that the work to prep the install of the new email notification software screwed the DB badly, making it almost impossible to retrieve, so the first "safe" back-up was the one from the January 1st.

I know it sucks.

Please Log in or Create an account to join the conversation.

More
13 Jan 2017 12:30 #52041 by brantley
Replied by brantley on topic BEFORE YOU POST READ THIS TOPIC
Woodclaw wrote:

<<Unfortunately I already checked and the last snapshot by wayback machine is from November 2016, way bfore the last site back-up.>>

So no retrieving posts, unless the posters saved their remarks from e-mails or whatever. But I presume authors can resubmit story files.

--Brantley

Please Log in or Create an account to join the conversation.

More
13 Jan 2017 15:08 #52043 by Woodclaw
Replied by Woodclaw on topic BEFORE YOU POST READ THIS TOPIC

brantley wrote: Woodclaw wrote:

<<Unfortunately I already checked and the last snapshot by wayback machine is from November 2016, way bfore the last site back-up.>>

So no retrieving posts, unless the posters saved their remarks from e-mails or whatever. But I presume authors can resubmit story files.

--Brantley


Absolutely. Right now I'm waiting for Willow to resub the first two installment of "Fall of Superior Girl"

Please Log in or Create an account to join the conversation.

More
14 Jan 2017 17:47 #52066 by brantley
Replied by brantley on topic BEFORE YOU POST READ THIS TOPIC
Something else I noticed, even before the latest decbacle: posts made during the same 24-hour period are tagged "today" and "yesterday." No big deal, compared to losing posts and files, but I'm still curious as to why.

--Brantley

Please Log in or Create an account to join the conversation.

More
14 Jan 2017 18:40 #52067 by Woodclaw
Replied by Woodclaw on topic BEFORE YOU POST READ THIS TOPIC

brantley wrote: Something else I noticed, even before the latest decbacle: posts made during the same 24-hour period are tagged "today" and "yesterday." No big deal, compared to losing posts and files, but I'm still curious as to why.

--Brantley


Maybe the individual timezones setting get screwed up. This should be easy to fix in the individual profiles.

Please Log in or Create an account to join the conversation.

More
15 Jan 2017 01:04 #52072 by fats
Replied by fats on topic BEFORE YOU POST READ THIS TOPIC
Hi all,

Here is what happened the other day. I tried to install a working email server to enable the site to be able to send emails to people, you know the ones that you get when you receive a notification about an update to a thread, or when you change your password. The install failed in the most dramatic way possible.

What it did was overwrite the settings for the web server rendering it inoperable, I spent 5 hours checking, rechecking and restoring the settings that I had, that's why some people saw an Apache screen and other a Nginx screen when they browsed the site.

Apache and Nginx are not the problems, so please don't be confused by that.

So my next action was to download the data and Database of the site, then reassemble the parts and create a new backup and use that backup to restore the site once I've reinstalled the operating system (centos 7) and re-initialise the server software. The files are fine they downloaded without any problems, the Database, on the other hand, seems to download without any problem, but when I came to import the SQL file using big-dump staggered SQL importer, a number of errors appeared. I edited the SQL file to remove the first few errors, it soon came clear that there was more than a few errors and it would take a lot more time to sort.

So I re-downloaded the Database to see if there was an error in the download, but a check of that file showed that there were the same errors. I then had to make a decision as to what I should do, so do I edit out the errors and restore that version of the site or do I use the last back-up and use that version of the site.

I decided to spend some time editing the downloaded Database and see how many errors there is and see if I can restore that version. After an hour and 40 errors fixed, I came to the conclusion that not knowing how many errors there were and not knowing how long it would take to fix the errors I took the reluctant decision to use the 10 days old backup and take it from there.

So I reinstalled the operating system and installed a working email system (xeams) and restored the web server and restored the backup and got the site live.

The email server while free needs a serial from the company that makes it, I’m just waiting for that to arrive, once I have that I can reactivate email notifications and get new members registered.

The issue of the backup came in the is thread, I want to address some points raised, I normally do a 7 day backup routine, every Sunday I back up the site, download the backup and then uploaded it to my dropbox account for safe keeping, the start of the year has been very interesting for me, I suffered a serious chest infection (being asthmatic does not help) that laid me up for about a week, I contacted it on the second and it took until the 9th for me to get over it, I’m still not fully cleared and I’m not on flight duties at the moment, so normally I would have done a new backup on the 7th but missed that due to the infection, one problem with the back-ups is that they take about 15 mins to create and they are large, at the moment the current backup is 3.4GB in size and will get bigger as the site gets bigger. That’s not a problem in terms of space to store the backup just the time it takes to download from the server and upload to my dropbox. One thing that I will do from now on is up the period between backups to once every 3 days, that way if there is a problem with the server and a back need to be used to restore material, at most all we will lose is 3 days of content, not as bad as losing 10 days which I think you will agree is a much better solution.

So in closing I would like to state for the record that the only person that will take the blame for what happened is me, not the host, not any of the software that we use, not the admins, just me, the blame lies fully on my shoulders and my shoulders only.

I hope that you all take this lengthy explanation as a closure on the matter and that we all get back to doing what we all do best.

Is post great Ubergirl material.

Fats
The following user(s) said Thank You: Woodclaw, Lulu, TwiceOnThursdays, Monty, LustMonster

Please Log in or Create an account to join the conversation.

  • fats
  • fats's Avatar Topic Author
  • Away
  • Administrator
  • Administrator
More
15 Jan 2017 06:50 - 15 Jan 2017 15:22 #52080 by Monty
Replied by Monty on topic BEFORE YOU POST READ THIS TOPIC
Thankyou for all your hard work Fats! I didn't know and alot of us won't realise how much work goes into the running of the site and ensuring the maintenance of the site. Kudos to yourself and the admin staff!
Last edit: 15 Jan 2017 15:22 by Monty.

Please Log in or Create an account to join the conversation.

More
Time to create page: 0.086 seconds