Log in

View Full Version : Waheyy, it's back !


squidge
September 13th, 2003, 19:45
Well done everyone who helped get the board back so quickly to a September restoration

Woodmann
September 13th, 2003, 20:22
Ya'll ready for a board upgrade

squidge
September 13th, 2003, 20:30
You wouldn't dare

(Hmmm, is that actually how you spell that word? Seems kinda odd)

JMI
September 13th, 2003, 20:39
Oh wouldn't we??????

Regards,

Kayaker
September 13th, 2003, 22:06
Now that's a stable board, even with shakedowns and breakdowns and against all odds, not a single significant post or archive has been lost in 3+ years!! Lol, nice

Do your worst.

JMI
September 13th, 2003, 23:15
Well, Hopefully we can do better than our "worst."

I've already installed a running copy of 3.0.0 in a test location and am about to install another copy of the 2.xx version and try to upgrade to 3.0.0 and see what happens. I'd like to get the DDX modified version running again and upgrade from there and maybe that will work this time, now that I know how to accomplish a "restore" using linux.

Sing the mantra along with me: "backup, backup, backup, backup." Feel the energy?

Regards.

squidge
September 14th, 2003, 04:55
Hey JMI, ever heard of the phrase "If it aint broke, don't fix it" ?

naides
September 14th, 2003, 08:54
I am taking to the word, the advice of my spiritual guide, JMI.

'grasshopper' , Life is a series of backups. . .

but master, I replied, I have tried to teleport the whole board, and never have I got anything remotely working.

master said, directing his eyes to a different direction,

'treacherous technology is. Do not think tao is contained within this pages'

but master, I replied, it is not Tao what I pursue in here, is the amusement of its postings.

'In that case, little grasshopper, ask the lords of this place to make available on-line a copy of the boards' database. '


I meditated around the obscure message contained in my master's words and became one with the idea that zen is obscure, but simple, indeed.

dELTA
September 14th, 2003, 12:12
One reason that the lords haven't made available such a database copy might be that it would require a bunch of sensitive data to be edited out of the database dump before it was published (like password hashes, user contact information, private messages and other non-public data), which would first of all be quite time consuming to do, and second of all make the database unable to run standalone anyway.


dELTA

JMI
September 14th, 2003, 13:25
And the reason that a teleport doesn't work is that there are just too many objects associated with each page and, at least in my own experience, teleport stops functioning long before it has anything near to a workable copy of the board.

The new 3.0.0 version is even worse. I made a backup of the test board last night which, uncompressed, is over 87 meg., although a good portion of that is the word index and search data, which can be eliminated from the backup if desired. The new version also has a function to store attachments as files, rather than as part of the board's database. That should eventually reduce the backup also.

Part of the problem we were having was that our back-ups were created with modified code which the standard "mysql" restore instruction did not uncompress. This meant it wasn't a click and go to restore the backup on anything other than the modified version of 2.2.9 we had been running and there were some problems getting it running again, which eventually were solved. But the "public" board is running the 2.3.2 code and I haven't had time to attempt to incorporate our previous modifications and I'm not sure they would work with the new code of 3.0.0 anyway.

To work on the current public version, the backups needed to be "uncompressed" locally and uploaded, or someone , like me, needed to figure out the proper path for mysql instructions on a linux box. All my previous practice was from a windows command line. While downloads from the board seem to run at 150kb/s, uploads of the backup seem to only run about 15 kb/s. Makes getting an uncompressed backup "back up" to the board, rather slow. On top of that, real life was again intruding on time here and the work had to get done.

Regards,

Woodmann
September 14th, 2003, 18:04
Quote:
Hey JMI, ever heard of the phrase "If it aint broke, don't fix it" ?


Squidgy......... I broke it

Sadly, the new version of the back up script is still sucking ass big time.
It is sad that when you have a backup problem, they tell you not to use the "packaged" script.

I will see if I can make a public version of the board available once a month. I dont think you will like the size of the file

Woodmann

JMI
September 14th, 2003, 18:22
I'm not sure a "public version" of the Board is a practical solution at the moment. Assuming we can remove the private data, i.e. private messages, user identification information, etc., and still present a useable database, what would we have? To be useful, the data needs to be searchable, i.e. one needs to have the correct version of vBull to actually "run" the database at home. The newer versions require a customer number to install (which might bring another project for our intrepid reversers).

Assuming the database was still in version 2.xx format, that would not be as much of a problem, but otherwise the "database" is really just a flat file system with a bunch of post and thread numbers connected through the database program. In the backup, unless properly sorted beforehand, the entries aren't necessarily near eachother and it can become a not too trivial exercise to put them in order, unless you have a mySQL database management program of some flavor.

In short, the data, without the board, it not a very workable proposition. Also, if one saves space by removing the word and search indexes, it kind of defeats the purpose of having the database in the first place. Unless one can recreate the indexes, searching would be very difficult. Unless one has setup one's own board and has access to the AdminCP, you can't re-index the counters to make searching work.

There might be some other way to do it, but at the moment, I don't know what that might be.

Regards.

JMI

squidge
September 15th, 2003, 05:54
I'd say we need some custom Teleport-like software, written specially for the board and rip the html pages directly. With such a html database, I'm sure it could be a lot easier to search.

Now, I think I could develope such an application, as long as the admins don't mind the hammer the board would take with such a dump. Also, I'm not sure how the database would be updated with new threads, but I'm sure we could sort something out.

Mfriend
September 15th, 2003, 07:02
Just to gain the raw data, I have used this one:


#!/bin/sh
export i=1
while true; do wget "http://www.woodmann.net/forum/printthread.php?s=&threadid=$i&perpage=1000"; i=`expr $i + 1`; done


of course there are a lot of "wrong" threads - but those can be easily filtered when you start making a browsable html "offline" board, where you need to parse every html page anyway. ( Currently we have around 5000 threads to "check" )

I will write this parser somewhen, if nobody writes it first, I only was glad to see the board up and sucked every thread as fast as I could ( don't worry, "only" 36 Mb traffic ). Of course there aren't any attachements - because they are not thread specific but message specific. You could download them too by this way ( you need to check about 25 000+ Urls ) but the problem is: I don't see a way to reassign the attachements to threads or even single messages ( that's way I haven't downloaded them ).


Another way would be - if the board admins make a db dump by some clever sql statement (e.g. only threadtitle+threadid, messagetitle+messageid, message, author+authorid, attachementifany) available - it would be a lot easier to parse the raw data and even have the attachements in the right places.


-mfriend

SiNTAX
September 15th, 2003, 18:03
Quote:
Originally posted by JMI
The newer versions require a customer number to install (which might bring another project for our intrepid reversers).


Just curious... how many seconds does it take to bypass?!

Kayaker
September 16th, 2003, 00:05
Quote:
Originally posted by Woodmann

Sadly, the new version of the back up script is still sucking ass big time.


Hiya, I just stumbled on this, any chance we could make use of it?

MySQL Backup Watcher
Program MySQL Backup Watcher is intended for making database backups of the MySQL server. Archiving can be done on the local computer as well as using the remote server in the
Internet. The program has the built – in and easily customized scheduler, thanks to which it is now possible not only to carry out "manual" backup but also to customize the scheduler so that
the program would automatically carry out the backup of the specified databases daily, weekly,
monthly or once on the established time.


hxxp://www.afreego.com/Categories/Business_and_Productivity/Database_Management/007916.php

Kayaker

Woodmann
September 16th, 2003, 18:55
Howdy,

I think a scheduled backup is a good idea.

Woodmann

[yAtEs]
September 17th, 2003, 08:26
backing up is for wussies eh woodmann

Woodmann
September 17th, 2003, 18:22
I hate you [yAtEs]