RE: AIX and Informix
Posted in 2006
Topics: High Availability & Replication, Storage & Space Management, Migration, Import/Export & Data Conversion, Platform-Specific Issues, Clustering, Grid & MACH11
This prompts me to ask whether you are using raw volumes or some sort of
cooked space.
DC
> -----Original Message-----
> From: informix-list-bounces@iiug.org [mailto:informix-list-
> bounces@iiug.org] On Behalf Of Link, David A. (Omaha)
> Sent: Friday, October 20, 2006 10:11 AM
> To: tomcaml@yahoo.com; informix-list@iiug.org
> Subject: RE: AIX and Informix
>
> If memory serves me correctly IBM does write some of the LV data in
the
> first 512 bytes of the volume. Now that data is accessed very rawly
and
> is written when you issue some system commands which manages the LVs.
I
> am not sure which commands.
>
> I think when you fail/fall over using HACMP it will rebuild the data
in
> the first 512 bytes of the LV. This ends up overwriting your informix
> data if you don't offset it.
>
> I have always used an offset of the LV in AIX.
>
> Here is a small document which kind of covers this.
>
> http://www.iiug.org/forums/ids/index.cgi/noframes/read/4433
>
> You may find more info out on the web.
>
> Hope this points you in the correct direction.
>
>
> David
>
> -----Original Message-----
> From: informix-list-bounces@iiug.org
> [mailto:informix-list-bounces@iiug.org] On Behalf Of tomcaml@yahoo.com
> Sent: Friday, October 20, 2006 10:47 AM
> To: informix-list@iiug.org
> Subject: Re: AIX and Informix
>
> Doug
>
> yes - Fishy like fresh from the sea
>
>
> Tom
>
>
>
> Doug Conrey wrote:
> > This sounds fishy to me. We have plenty of dbspaces with an offset
of
> 0
> > (on 5.2 and 5.3), and have never had any problems with
> > adding/dropping/using them.
> >
> > DC
> >
> > > -----Original Message-----
> > > From: informix-list-bounces@iiug.org [mailto:informix-list-
> > > bounces@iiug.org] On Behalf Of tomcaml@yahoo.com
> > > Sent: Friday, October 20, 2006 9:22 AM
> > > To: informix-list@iiug.org
> > > Subject: Re: AIX and Informix
> > >
> > > Thanks for the comments.
> > > We are running into several things.
> > >
> > > One is that we set the offset of the dbspace to 0. We are being
told
> > it
> > > needs to have an offset of "something" - that AIX wants this for
the
> > > volume group management. Our current environ is HP which we have
> > always
> > > used an offset of 0 - it has never been an issue. We have our
> > physical,
> > > log and root spaces mirrored and there looks to be some kind of
> > > corruption, according to support. Kinda wondering if the HACMP
> > > clustering might be involved also but unsure.
> > > What we are seeing is this: I dropped a large database and when I
> went
> > > to check out the spaces using onstat -d pertaining to the space
the
> DB
> > > used to be in, the % used was a negative number. So the total size
> > free
> > > is greater than the total size allocated.
> > >
> > > Tom
> > >
> > >
> > > Obnoxio The Clown wrote:
> > > > tomcaml@yahoo.com said:
> > > > > Hello
> > > > >
> > > > > We are running into some major walls trying to configure and
use
> > AIX
> > > > > 5.3 and Informix v10 with HACMP clustering and Informix HDR.
AIX
> > looks
> > > > > to be very behind the times, as does the clustering piece.
This
> is
> > for
> > > > > an intended migration of the main "cash cow" application. Does
> > anyone
> > > > > run their "important stuff" on AIX ? I have been researching
and
> > most
> > > > > of what I see is people using AIX from years back or they only
> use
> > AIX
> > > > > for smaller apps that they don't have great concerns being
down.
> > > > > Comments, Suggestions, Rants or Raves?
> > > >
> > > > Personally, I think AIX tin is very good. I can't comment on
> HACMP,
> > > other
> > > > than that I've never met anyone who used ANY clustering software
> > (HACMP,
> > > > Sun, Veritas, Compaq) who didn't prefer HDR.
> > > >
> > > > --
> > > > Bye now,
> > > > Obnoxio
> > > >
> > > > "... no bill is required as no value was provided."
> > > > -- Christine Normile
> > > >
> > > > --
> > > > This message has been scanned for viruses and
> > > > dangerous content by OpenProtect(http://www.openprotect.com),
and
> is
> > > > believed to be clean.
> > >
> > > _______________________________________________
> > > Informix-list mailing list
> > > Informix-list@iiug.org
> > > http://www.iiug.org/mailman/listinfo/informix-list
>
> _______________________________________________
> Informix-list mailing list
> Informix-list@iiug.org
> http://www.iiug.org/mailman/listinfo/informix-list
> _______________________________________________
> Informix-list mailing list
> Informix-list@iiug.org
> http://www.iiug.org/mailman/listinfo/informix-list
We use raw. HACMP folks say use offset 512 - AIX folks say no offset is
needed.
We are going to try some with 512 and see what blows up - or not
Thanks again
Tom
Doug Conrey wrote:
> This prompts me to ask whether you are using raw volumes or some sort of
> cooked space.
>
> DC
>
> > -----Original Message-----
> > From: informix-list-bounces@iiug.org [mailto:informix-list-
> > bounces@iiug.org] On Behalf Of Link, David A. (Omaha)
> > Sent: Friday, October 20, 2006 10:11 AM
> > To: tomcaml@yahoo.com; informix-list@iiug.org
> > Subject: RE: AIX and Informix
> >
> > If memory serves me correctly IBM does write some of the LV data in
> the
> > first 512 bytes of the volume. Now that data is accessed very rawly
> and
> > is written when you issue some system commands which manages the LVs.
> I
> > am not sure which commands.
> >
> > I think when you fail/fall over using HACMP it will rebuild the data
> in
> > the first 512 bytes of the LV. This ends up overwriting your informix
> > data if you don't offset it.
> >
> > I have always used an offset of the LV in AIX.
> >
> > Here is a small document which kind of covers this.
> >
> > http://www.iiug.org/forums/ids/index.cgi/noframes/read/4433
> >
> > You may find more info out on the web.
> >
> > Hope this points you in the correct direction.
> >
> >
> > David
> >
> > -----Original Message-----
> > From: informix-list-bounces@iiug.org
> > [mailto:informix-list-bounces@iiug.org] On Behalf Of tomcaml@yahoo.com
> > Sent: Friday, October 20, 2006 10:47 AM
> > To: informix-list@iiug.org
> > Subject: Re: AIX and Informix
> >
> > Doug
> >
> > yes - Fishy like fresh from the sea
> >
> >
> > Tom
> >
> >
> >
> > Doug Conrey wrote:
> > > This sounds fishy to me. We have plenty of dbspaces with an offset
> of
> > 0
> > > (on 5.2 and 5.3), and have never had any problems with
> > > adding/dropping/using them.
> > >
> > > DC
> > >
> > > > -----Original Message-----
> > > > From: informix-list-bounces@iiug.org [mailto:informix-list-
> > > > bounces@iiug.org] On Behalf Of tomcaml@yahoo.com
> > > > Sent: Friday, October 20, 2006 9:22 AM
> > > > To: informix-list@iiug.org
> > > > Subject: Re: AIX and Informix
> > > >
> > > > Thanks for the comments.
> > > > We are running into several things.
> > > >
> > > > One is that we set the offset of the dbspace to 0. We are being
> told
> > > it
> > > > needs to have an offset of "something" - that AIX wants this for
> the
> > > > volume group management. Our current environ is HP which we have
> > > always
> > > > used an offset of 0 - it has never been an issue. We have our
> > > physical,
> > > > log and root spaces mirrored and there looks to be some kind of
> > > > corruption, according to support. Kinda wondering if the HACMP
> > > > clustering might be involved also but unsure.
> > > > What we are seeing is this: I dropped a large database and when I
> > went
> > > > to check out the spaces using onstat -d pertaining to the space
> the
> > DB
> > > > used to be in, the % used was a negative number. So the total size
> > > free
> > > > is greater than the total size allocated.
> > > >
> > > > Tom
> > > >
> > > >
> > > > Obnoxio The Clown wrote:
> > > > > tomcaml@yahoo.com said:
> > > > > > Hello
> > > > > >
> > > > > > We are running into some major walls trying to configure and
> use
> > > AIX
> > > > > > 5.3 and Informix v10 with HACMP clustering and Informix HDR.
> AIX
> > > looks
> > > > > > to be very behind the times, as does the clustering piece.
> This
> > is
> > > for
> > > > > > an intended migration of the main "cash cow" application. Does
> > > anyone
> > > > > > run their "important stuff" on AIX ? I have been researching
> and
> > > most
> > > > > > of what I see is people using AIX from years back or they only
> > use
> > > AIX
> > > > > > for smaller apps that they don't have great concerns being
> down.
> > > > > > Comments, Suggestions, Rants or Raves?
> > > > >
> > > > > Personally, I think AIX tin is very good. I can't comment on
> > HACMP,
> > > > other
> > > > > than that I've never met anyone who used ANY clustering software
> > > (HACMP,
> > > > > Sun, Veritas, Compaq) who didn't prefer HDR.
> > > > >
> > > > > --
> > > > > Bye now,
> > > > > Obnoxio
> > > > >
> > > > > "... no bill is required as no value was provided."
> > > > > -- Christine Normile
> > > > >
> > > > > --
> > > > > This message has been scanned for viruses and
> > > > > dangerous content by OpenProtect(http://www.openprotect.com),
> and
> > is
> > > > > believed to be clean.
> > > >
> > > > _______________________________________________
> > > > Informix-list mailing list
> > > > Informix-list@iiug.org
> > > > http://www.iiug.org/mailman/listinfo/informix-list
> >
> > _______________________________________________
> > Informix-list mailing list
> > Informix-list@iiug.org
> > http://www.iiug.org/mailman/listinfo/informix-list
> > _______________________________________________
> > Informix-list mailing list
> > Informix-list@iiug.org
> > http://www.iiug.org/mailman/listinfo/informix-list
tomcaml@yahoo.com wrote:
> We use raw. HACMP folks say use offset 512 - AIX folks say no offset is
> needed.
>
> We are going to try some with 512 and see what blows up - or not
>
> Thanks again
>
>
> Tom
Just to clear up a possibly misunderstanding. '512' is bytes. When you give an offset using onspaces it's on kilobytes.
Further the offset must always be a multiplum of the page size, which is 4 Kb on AIX, so use '-o 4' with onspaces.
'onstat -d' will then show offset = 1 (page).
Related threads
- IDS 10 table-level restore
- Informix Development Webinar December 11, 2007
- ontape -p/r with changed ROOTPATH
- Migrate from HP PA-RISC to HP ITANIUM by ontape