Replying to my own message: ----- Original Message -----
> We are shortly to go live with a new set of databases (InnoDB). > > The data in each database is identical in fields and types, the only > difference being in the relevance of the data - they are soccer > leagues, and each database represents information on the leagues/teams > for a single season. There is a separate controlling database which > directs web requests to the right year. > > Each league is identified within every table by a 3-4 character code. > This all works pretty well, and was deemed a better scenario than > having a database for each league, with a numeric field to identify the > year. There are something like 160 leagues involved, but only 6 years > (and rising). So we have 6 databases to take care of, not 160. > > Having explained the basic setup, we come to a problem: what if one of > the administrators accidentally deletes a set of league information > (maybe fixture information, for instance) six hours after the last > backup? > > We could, of course, restore from the backup, but that would compromise > every other league administrator's efforts since backup. Ideally, we > would restore only the data relevant to that particular league (with > the proper 3-4 letter code). > > Apart from opening up the latest mysqldump file and extracting the data > from there (assuming the dork who deleted it knows more or less what > has gone), and re-inserting table by table, having removed the data > from other leagues, I can't think of another way to do it. > > We're currently using 4.1.3/Apache/WinNT, with hopes of upgrading to > 4.1.7 before going live, though I doubt if the MySQL sub-version is > relevant. > > I'd appreciate any help or guidance or advice on suitable tools. > It would seem that one way to get what I need (and I have tried this) is to restore a 'usable' backup into a dummy database, and run through the tables of the one which needs the partial restore (after backing it up first, naturally) deleting records for that league (using the 3-4 letter code), then insert ... select from the dummy. There are fifteen tables involved, and in the trial I carried out using this method, the actual delete/insert...select sequence took about 20 seconds on my beat-up, memory-starved machine. Took me longer doing the backup/transfer to dummy. So I have found *a* method for doing what I need, while quite a few of you have been away enjoying (hopefully) your Thanksgiving holiday. It would be useful if anyone could point me at a simpler alternative, though. Cheers Terry -- MySQL General Mailing List For list archives: http://lists.mysql.com/mysql To unsubscribe: http://lists.mysql.com/[EMAIL PROTECTED]
