disaster recovery
Ben
I guess I'll just reference my post that was somehow moved.
I was wondering about the statement of better to use a current
controlfile rather than one that I recovered. Why is it better?
I believe I'm going to have to do a pitr, so why not just use the
controlfile and spfile that was included in the hotbackup that I'm
going to be using for recovery?
joel garry
I think it's just easier to let Oracle do if possible, but if you are
PITR then do what you need to. As always, it depends. A lot of calls
for help here indicate people are unclear on when to use which syntax
(like using backup controlfile and such), often they will make it more
complicated than necessary and blow it. The trick is to understand
how Oracle compares SCN's in datafile headers and controlfiles, in
order to decide what recovery is needed. The bit about disk versus
tape - if you still have an original controlfile on disk and the
archived logs that the recovered data files will need, you probably
want to use that one, letting Oracle recover as much as it can, and
rollback the rest. But it depends on why you want PITR - what
happened that you don't want to be committed? If it is just all your
disks got wiped including archived logs and you need to restore to the
past, just do that.
jg
--
@home.com is bogus.
Maybe it's just customers who are stable enough to use remote DBA
support that are slow to adopt...
http://www.computerworlduk.com/technology/business-intelligence/databases/news/index.cfm?RSS&newsid=9093
Ben
The reference to disk in the original post was a typo, the only disk
that the server can still find data on, is the root system disk. We
may be able to get back the filesystem that included the oracle home.
If we can, that would include our spfile and alert log from the day
before the outage. Other than that, I don't have anything left except
for backup pieces on tape.
The specific timeline of events was
Friday. full hot backup
Sunday: power outtage to disk cabinet, server's conception of disks
are scrambled.
Monday: we realize that the server is up but no data.
Frank van Bortel
I have to disagree with Joel on this one - *always* use the most
current controlfile - it has the most up-to-date state of
your database. OK - having said that:
Just restore your backups on a (different) machine.
Mount, alter the locations of the tablespaces, move
the datafiles around, recover (maybe until cancel, or
until sunday/seconds before crash) and open.
Sounds so simple, and sometimes it is.
May the force be with you.
FvB
Ben
wrote:
>
> - Show quoted text -
Oh but only if we had a machine to put it on.....
Ben
>
> - Show quoted text -
for the backup process we do a
backup database
when restoring the spfile if I tell it to
restore spfile from autobackup;
will it find that backup that was included from the hotbackup?
Ben
>
> - Show quoted text -
This is getting aggravating, I've tried almost all options other than
pl/sql route.
Here is how our backup was taken:
56> connect catalog *
57>
58> connect target *
59>
60> run {
61> # Hot database level 0 whole backup
62> allocate channel t1 type 'SBT_TAPE';
63> allocate channel t2 type 'SBT_TAPE';
64> backup
65> incremental level 0
66> skip inaccessible
67> tag jde_hot_level0
68> filesperset 10
69> # recommended format
70> format 'bk_%s_%p_%T'
71> (database);
72> }
now that we've lost the catalog, spfile, controlfiles, and everything
else. Is there a way to get the controlfile and spfile out of that
backup?
It seems to me that rman can't extract the spfile from tape unless
you've set controlfile autobackup on. Is there a way for rman to do
this?
sybr...@hccnet.nl
wrote:
>
>This is getting aggravating, I've tried almost all options other than
>pl/sql route.
>
>Here is how our backup was taken:
>
>56> connect catalog *
>57>
>58> connect target *
>59>
>60> run {
>61> # Hot database level 0 whole backup
>62> allocate channel t1 type 'SBT_TAPE';
>63> allocate channel t2 type 'SBT_TAPE';
>64> backup
>65> incremental level 0
>66> skip inaccessible
>67> tag jde_hot_level0
>68> filesperset 10
>69> # recommended format
>70> format 'bk_%s_%p_%T'
>71> (database);
>72> }
>
>now that we've lost the catalog, spfile, controlfiles, and everything
>else. Is there a way to get the controlfile and spfile out of that
>backup?
If you didn't have controlfile autobackup on, which you didn't
specify, shortly : NO.
Also it looks to me like the person who 'designed' the script likes to
gamble or is looking for a career move.
Not backing up the archivelogs and using 'skip inacessible' for a
database backup is not contributing to a consistent backup.
--
Sybrand Bakker
Senior Oracle DBA
Ben
> On Fri, 16 May 2008 13:02:36 -0700 (PDT), Ben <benal...@yahoo.com>
>
> - Show quoted text -
I figured sybrand the holy would weigh in on this, Good job, way to
go, your awesome!
Ben
>
> - Show quoted text -
just an fyi, it turned out that our sys admin had in-activated the
policy that had our development client in it, that was causing an
issue. And there might have been a hostnames file changed on the mml
server since we tested it out that was causing a problem. And there
are still issues with the veritas server that we have to work out. It
is actually attempting to restore now though.
joel garry
wrote:
See step 3 at http://download.oracle.com/docs/cd/B19306_01/backup.102/b14191/rcmtspit001.htm#i1008453
Hey, if RMAN does it... but I'm still not clear why a PITR would be
desired here anyways.
As far as simplicity, well, sometimes this stuff has too much
abstraction between what is going on and the very high level RMAN
commands, that can be very confusing.
To Ben: Frank is better at this stuff than me, Sybrand certainly
is.
jg
--
@home.com is bogus.
Thought I could skip the Playstation 2 generation and go right to the
newer ones, until someone wanted Dance Dance Revolution for Mother's
day. One version for 3, one version for X-box, one version for Wii,
lots for P2. It _is_ the app.<sigh>
Ben
>
> Hey, if RMAN does it... but I'm still not clear why a PITR would be
> desired here anyways.
>
> As far as simplicity, well, sometimes this stuff has too much
> abstraction between what is going on and the very high level RMAN
> commands, that can be very confusing.
>
> To Ben: Frank is better at this stuff than me, Sybrand certainly
> is.
>
> jg
> --
> @home.com is bogus.
> Thought I could skip the Playstation 2 generation and go right to the
> newer ones, until someone wanted Dance Dance Revolution for Mother's
> day. One version for 3, one version for X-box, one version for Wii,
>
> - Show quoted text -
Sybrand is too rude to be better at anything for my taste. I'd rather
shake hands with a tasmanian devil than have to put up with his
condecsending comments.
Frank van Bortel
>
> Sybrand is too rude to be better at anything for my taste. I'd rather
> shake hands with a tasmanian devil than have to put up with his
> condecsending comments.
1) Tasmanian devils do not have hands...
2) Sybrand is right - if you have no controlfile(s) in your backup,
(and your online ones are lost, so I understand) you are skrewed...
Maybe Sybrands language is not always as political correct as you
wish, but hey, he's not in politics, now is he?
Neither am I, btw, and -apart from being countrymen- I am not
associated with Sybrand in any way.
Sorry for the delay(s), I'm less online than I want, lately.
How are the attempts to get your datafiles back online
going on? Any chance you might recover some (all?!?) files
from disk?
That would make your life real easy... Just mount, recover,
and you're probably done (assuming your development was
not executing heavy jobs during black out).
You could try to restore the spfile from the backup piece
(if you know where rman stored it).
Nevertheless, without control file... Did you try with
using backup controlfile? That would signal RMAN that
you are using an out of date controlfile, and that the
control file(s) should be rewritten according to data
from the datafiles.
FvB
Ben
wrote:
> Ben wrote:
>
> > Sybrand is too rude to be better at anything for my taste. I'd rather
> > shake hands with a tasmanian devil than have to put up with his
> > condecsending comments.
>
> 1) Tasmanian devils do not have hands...
> 2) Sybrand is right - if you have no controlfile(s) in your backup,
> (and your online ones are lost, so I understand) you are skrewed...
> Maybe Sybrands language is not always as political correct as you
> wish, but hey, he's not in politics, now is he?
> Neither am I, btw, and -apart from being countrymen- I am not
> associated with Sybrand in any way.
He's right about a lot of things, I just don't like him. I'm not
necessarily politically correct myself, you don't have to be if your
not an asshat along with it.
> Sorry for the delay(s), I'm less online than I want, lately.
>
> How are the attempts to get your datafiles back online
> going on? Any chance you might recover some (all?!?) files
> from disk?
well, my attempts at recovering the controlfile and spfile from backup
pieces via pl/sql would probably have went better if it weren't for
problems with the tape library and veritas. I did however get to use a
auto snap of the controlfile on disk and got the database mounted and
ready for use with rman & controlfile. But alas veritas still has
problems and is not being very cooperative in getting me my datafiles
back. It's looking like we might have to fall back on an old copy of
the database and then import some crucial data from exports that were
made the night before the failure.
> That would make your life real easy... Just mount, recover,
> and you're probably done (assuming your development was
> not executing heavy jobs during black out).
The files weren't accessible after the failure to mount. The entire
file system had to be re-created and we had an older (6 months) copy
of the database on some disks that didn't get wiped out. Those will
probably be our fall back.
>
> You could try to restore the spfile from the backup piece
> (if you know where rman stored it).
> Nevertheless, without control file... Did you try with
> using backup controlfile? That would signal RMAN that
> you are using an out of date controlfile, and that the
> control file(s) should be rewritten according to data
> from the datafiles.
>
> FvB
Thanks for all the info once again.
DA Morgan
> shake hands with a tasmanian devil than have to put up with his
> condecsending comments.
Many of us disagree. I suspect Sybrand speaks a minimum of four
languages of which English is not the first and likely not the second.
How nuanced are your statements in your third or fourth language?
But give credit where credit is due ... as others have already pointed
out ... he's correct. And for correct answers the proper response is
"Thank you."
--
Daniel A. Morgan
Oracle Ace Director & Instructor
University of Washington
damo...@x.washington.edu (replace x with u to respond)
Puget Sound Oracle Users Group
www.psoug.org
DA Morgan
> necessarily politically correct myself, you don't have to be if your
> not an asshat along with it.
Au contraire. No person on the planet looks in the mirror and sees
themselves as other see them. A bit of humility and tolerance might
be an appropriate starting place. No doubt a quick look at any of
us would produce some opinions that vastly differ from our own.
Shakespeare
"DA Morgan" <damo...@psoug.org> schreef in bericht
news:12112354...@bubbleator.drizzle.com...
>
> Many of us disagree. I suspect S. speaks a minimum of four
> languages of which English is not the first and likely not the second.
> How nuanced are your statements in your third or fourth language?
>
It's a pitty you can't read Dutch (S.'s first language). It's even MORE rude
than his English.
Shakespeare
Ben
> "DA Morgan" <damor...@psoug.org> schreef in berichtnews:12112354...@bubbleator.drizzle.com...
>
>
>
> > Many of us disagree. I suspect S. speaks a minimum of four
> > languages of which English is not the first and likely not the second.
> > How nuanced are your statements in your third or fourth language?
>
> It's a pitty you can't read Dutch (S.'s first language). It's even MORE rude
> than his English.
>
> Shakespeare
funny. I was going to make a comment about him knowing English well
enough to realize he's being an ass.
Ben
Actually, I'll be surprised if he doesn't blast Mr Morgan with an
insult for implying he doesn't realize he's being rude.
joel garry
>
> >now that we've lost the catalog, spfile, controlfiles, and everything
> >else. Is there a way to get the controlfile and spfile out of that
> >backup?
>
> If you didn't have controlfile autobackup on, which you didn't
> specify, shortly : NO.
I have a recently installed hp-ux 10.2.0.3 with autobackup on. It's
just idling along while I do app development on another machine. This
thread made me take a closer look at the autobackup. I check the
alert log, I see it is writing controlfile backups to specific files
in directories with each days date on them, makes sense since I'm
taking a full backup each day (I haven't set up a repository yet). It
seems to be keeping a weeks worth of these directories, but I only
just noticed it only is keeping the latest controlfile backup. Seems
not paranoid enough to me. I have CONFIGURE RETENTION POLICY TO
REDUNDANCY 2;
The script as generated by the customization wizard is this:
$rman_script="backup device type disk tag '%TAG' database;
backup device type disk tag '%TAG' archivelog all not backed up;
allocate channel for maintenance type disk;
delete noprompt obsolete device type disk;
release channel;
";
Am I wrong to expect the delete obsolete to keep two of the
controlfile/spfile backups? Or is it just also in the regular backup
files?
Looking at the rman log, I see what is happening: It is backing up
the controlfile/spfile once for the archivelog backup, then once again
for the regular backup, then deleting all the others. So it is
keeping two, from the same time - the log appears to have them before
and after the backup, but the unix time is the same (I assume because
rman must have closed the files at nearly the same time).
This seems strange, how should I be doing this? Shouldn't I want to
have some spread of controlfile backup times, to avoid just
propagating a messed up one when restoring?
jg
--
@home.com is bogus.
“I like candy in general, I'm pretty favorably disposed toward
candy.” - Warren Buffet
Shakespeare
"Ben" <bena...@yahoo.com> schreef in bericht
news:d4d30824-457c-475c...@m44g2000hsc.googlegroups.com...
On May 20, 7:54 am, Ben <benal...@yahoo.com> wrote:
> On May 20, 5:56 am, "Shakespeare" <what...@xs4all.nl> wrote:
>
> > "DA Morgan" <damor...@psoug.org> schreef in
> > berichtnews:12112354...@bubbleator.drizzle.com...
>
> > > Many of us disagree. I suspect S. speaks a minimum of four
> > > languages of which English is not the first and likely not the second.
> > > How nuanced are your statements in your third or fourth language?
>
> > It's a pitty you can't read Dutch (S.'s first language). It's even MORE
> > rude
> > than his English.
>
> > Shakespeare
>
> funny. I was going to make a comment about him knowing English well
> enough to realize he's being an ass.
- Actually, I'll be surprised if he doesn't blast Mr Morgan with an
- insult for implying he doesn't realize he's being rude.
No, S. knows well who to blast and who not...
Just give yourself an 'Indian like' nickname and see what happens.....
Shakespeare
sybr...@hccnet.nl
Joel,
I can only investigate this for 9i. I don't think 10g has changed in
this regard, so that should be OK. I ran into something strange today
which I need to investigate anyway: my autobackups also seem to
disappear, though my retention is 9 days.
One question though:
Do you happen to use backup optimization.
I noticed using backup optimization in 9i you can only backup the
archivelogs once, whatever you tell it to do. It begins to brag you
already have a backup of the archivelogs.
Sounds equally scary.
The behavior with autobackup controlfile may be the same issue.
The reasoning might be you have records of multiple backups in the
controlfile anyway, so you should be OK.
Hth
joel garry
Correction, the two backups bracket the archivelog backup, since there
is no activity on the db for now, that just takes seconds.
>
> >This seems strange, how should I be doing this? Shouldn't I want to
> >have some spread of controlfile backup times, to avoid just
> >propagating a messed up one when restoring?
>
> >jg
>
> Joel,
>
> I can only investigate this for 9i. I don't think 10g has changed in
> this regard, so that should be OK. I ran into something strange today
> which I need to investigate anyway: my autobackups also seem to
> disappear, though my retention is 9 days.
That's why I said something, I figured either I missed something or
there is a consequence one doesn't hear often about and many people
might have misapprehensions, or issues waiting to blow, even. I
haven't settled on window or redundancy yet.
Must be Frank is right, you do always just want the latest one, the
doc I quoted to him may be misleading if there is only effectively one
controlfile to choose.
>
> One question though:
> Do you happen to use backup optimization.
It's the default (off).
> I noticed using backup optimization in 9i you can only backup the
> archivelogs once, whatever you tell it to do. It begins to brag you
> already have a backup of the archivelogs.
> Sounds equally scary.
> The behavior with autobackup controlfile may be the same issue.
> The reasoning might be you have records of multiple backups in the
> controlfile anyway, so you should be OK.
>
Strange reasoning to me, given that only the controlfile has the
records.
jg
--
@home.com is bogus.