On Fri, May 13, 2011 at 10:21, Joerg Bruehe <joerg.bruehe@stripped> wrote:
> Hi everybody!
> Dotan Cohen wrote:
>> Is there a simple browser-based MySQL backup solution for
>> non-technical users. [[...]]
>> The main features needed are:
>> 1) Automatic scheduled off-site backups (via SSH or FTP)
> Off-site = good (for reliability purposes).
>> 2) Backup multiple databases and all their tables
> Definitely a "must have".
>> 3) Single-table recovery via GUI (the user simply chooses which
>> database and which table to recover)
> If your backup/recovery tool has this feature and your users ever go
> that route, you (your DBA / your authorities) must be aware that this
> will break any dependencies between that recovered table and all other,
> un-recovered ones.
> Assume a new entry is added to the "customer" table, then (at least) one
> "order" is entered for this customer.
> Before, during, or after that, some garbage change is done to the
> "customer" table, it is detected, and someone decides "let's recover the
> customer table from the last good backup".
> This will get rid of the garbage, but will also make the orders for new
> customer be pointing to nowhere.
> As soon as you have relationships crossing table boundaries, a
> single-table recovery is a very risky operation, and it will violate any
> "referential integrity" constraints involving that table.
Thanks Joerg for that insight. In fact, this is a very simple
installation with no joins but I will keep that in mind for the
future. Terrific point.