[mythtv-users] REMINDER: Back Up Your Database.

[Andrew C. (AFPup) Stadt]

On 31/12/13 12:00 AM, Paul Gardiner wrote:
> "Andrew C. (AFPup) Stadt" <acstadt at stadt.ca> wrote:
>> On 30/12/13 11:26 PM, tortise wrote:
>>> On 31/12/2013 4:38 p.m., Jay Ashworth wrote:
>>>> Every night.
>>>>
>>>> My personal recommendation, to avoid WAF crashes like the on another
>>>> poster
>>>> is impaled on, is to run the mythconverg_backup.pl script from cron
>>>> every day
>>>> at 4am, and then *copy the created backup file to every spinning
>>>> drive you
>>>> have installed on your machine*.  That way, if you ever have to
>>>> restorr, you
>>>> have 3, 4, (5, 6, 7... :-) chances to have a good copy, and you
>> don't
>>>> care
>>>> which drives might not be mounted at the time.
>>>>
>>>> Make it your New Year's Resolution.  Go check now: do you have a
>> current
>>>> backup of your DB?
>>>>
>>>> Cheers,
>>>> -- jra
>>> Great idea.
>>> {Smugly thinks to self I don't need to....}
>>> Hmmm, mine hasn't run for a good three months now....!!!
>>> Anyone else?!
>>> In anticipation of the advice, and avoiding risks of impalation best
>> I
>>> sort out why!
>>> Thanks indeed for the post jra
>>>
>> I do have to admit that at one point, mythconverg_backup stopped
>> working
>> for me.  I had transitioned from packages to git, and somehow an
>> outdated backup script got left in my path.  Of course I figured this
>> out when I needed the backup.  Luckily the fs that hosted the database
>> had been backed up... but..
>
> That could be the cause of my mythconverg_backup failure. I remember now, when I updated it from 0.24 to 0.25, I ran into many problems and ended up moving from packages to git before I could make it work. I may well have ended up using the 0.24 script to backup the 0.25 system. In your case did it fail silently and produce empty(ish) backups?
>
Actually that is exactly what happened.  I was checking to make sure 
that the backup files were there, and they were... I just failed to look 
at the file size, doh.  As I said, in my case I had a backup copy of the 
filesystem which had hosted the database, so not a huge issue.