Re: RMAN or Hot Backup
From: John Schaeffer <amerar_at_iwc.net>
Date: Fri, 27 Mar 2009 14:03:25 -0700 (PDT)
Message-ID: <c3ae9b8d-2a28-4ef2-8e9e-91c514b4d42f_at_f19g2000yqo.googlegroups.com>
On Mar 27, 2:43 pm, Michael Austin <maus..._at_firstdbasource.com> wrote:
> jgar the jorrible wrote:
> > On Mar 27, 8:01 am, John Schaeffer <ame..._at_iwc.net> wrote:
> >> On Mar 26, 9:30 pm, "Bob Jones" <em..._at_me.not> wrote:
>
> >>> "Mladen Gogala" <gogala.mla..._at_gmail.com> wrote in message
> >>>news:gqh502$q0h$2_at_solani.org...
> >>>> On Thu, 26 Mar 2009 17:24:28 -0500, Michael Austin wrote:
> >>>>> They are morons and should be relieved of their duties and COMPETENT
> >>>>> talent hired to replace them.
> >>>> I would disagree, and bitterly so. There are reasons for using hot
> >>>> backups, especially in this economy.
> >>> Hot backup and economy? That's ingenious.
> >> Well, I've used RMAN before, and I'm all for it. There is so much it
> >> does and takes care of for you which a Hot Backup cannot. But, these
> >> other nuts want to switch and I'm hoping to get a bunch of reasons why
> >> not to switch from all of you.......
>
> >> Thanks!
>
> > We might help more if you said their reasons to switch... you can see
> > from this thread that this grates on the fundamentals of dba work.
>
> > Michael Austin wrote:
>
> >> EMC has the BCV (Business Continuity Volume) where they use HotBackup
> >> Mode to "clone" a database by breaking a mirror and importing it into
> >> another instance. Again, On a very busy system, once or twice/week they
> >> would end up with corrupt undo-segments causing ORA-600 errors pointing
> >> to those corrupt segments. Basically, if you continually do a
> >> "controlled" crash like this, you are lucky more often than not if it
> >> works...
>
> > Did this happen with an alter db backup surrounding the break?
> > Yeesh. I've been wondering about what will happen putting the db into
>
> Yes it does/did... Again using the methodology is essentially initiating
> a controlled crash. - but, as I have experienced, depending on where the
> Database writer and logwriters are at the exact moment of the break, you
> can get corrupted UNDO segments and then you are hosed - and now you try
> again - maybe next time you will get lucky and it will work.
>
> > backup mode every 15 minutes, as one proposal coming down the pike
> > implies. I need Swiss pikes to defend my db.
>
> > jg
> > --
> > _at_home.com is bogus.
> >http://en.wikipedia.org/wiki/Pike_(cipher)
Date: Fri, 27 Mar 2009 14:03:25 -0700 (PDT)
Message-ID: <c3ae9b8d-2a28-4ef2-8e9e-91c514b4d42f_at_f19g2000yqo.googlegroups.com>
On Mar 27, 2:43 pm, Michael Austin <maus..._at_firstdbasource.com> wrote:
> jgar the jorrible wrote:
> > On Mar 27, 8:01 am, John Schaeffer <ame..._at_iwc.net> wrote:
> >> On Mar 26, 9:30 pm, "Bob Jones" <em..._at_me.not> wrote:
>
> >>> "Mladen Gogala" <gogala.mla..._at_gmail.com> wrote in message
> >>>news:gqh502$q0h$2_at_solani.org...
> >>>> On Thu, 26 Mar 2009 17:24:28 -0500, Michael Austin wrote:
> >>>>> They are morons and should be relieved of their duties and COMPETENT
> >>>>> talent hired to replace them.
> >>>> I would disagree, and bitterly so. There are reasons for using hot
> >>>> backups, especially in this economy.
> >>> Hot backup and economy? That's ingenious.
> >> Well, I've used RMAN before, and I'm all for it. There is so much it
> >> does and takes care of for you which a Hot Backup cannot. But, these
> >> other nuts want to switch and I'm hoping to get a bunch of reasons why
> >> not to switch from all of you.......
>
> >> Thanks!
>
> > We might help more if you said their reasons to switch... you can see
> > from this thread that this grates on the fundamentals of dba work.
>
> > Michael Austin wrote:
>
> >> EMC has the BCV (Business Continuity Volume) where they use HotBackup
> >> Mode to "clone" a database by breaking a mirror and importing it into
> >> another instance. Again, On a very busy system, once or twice/week they
> >> would end up with corrupt undo-segments causing ORA-600 errors pointing
> >> to those corrupt segments. Basically, if you continually do a
> >> "controlled" crash like this, you are lucky more often than not if it
> >> works...
>
> > Did this happen with an alter db backup surrounding the break?
> > Yeesh. I've been wondering about what will happen putting the db into
>
> Yes it does/did... Again using the methodology is essentially initiating
> a controlled crash. - but, as I have experienced, depending on where the
> Database writer and logwriters are at the exact moment of the break, you
> can get corrupted UNDO segments and then you are hosed - and now you try
> again - maybe next time you will get lucky and it will work.
>
> > backup mode every 15 minutes, as one proposal coming down the pike
> > implies. I need Swiss pikes to defend my db.
>
> > jg
> > --
> > _at_home.com is bogus.
> >http://en.wikipedia.org/wiki/Pike_(cipher)
Honestly, they have not provided any concrete reasons for doing so other than "We've done it before and it is easier to do'. Seems like they have provided no good reasons at all.........
So yes, it is DBA basics. I really have not done too many Hot Backups, which is why I was asking what others opinions and experience with both are. Received on Fri Mar 27 2009 - 16:03:25 CDT