I have a two node HACMP/ES4.4.1 cluster configured with one resource
group on the A box and none on the B box. The cluster is configured as
fallover without fallback cascading. The systems are two 6H1's running
AIX5.1ML04 and connected two two 7133-020 fully populated SSA
enclosures. We have 8 volume groups configured in the cluster and all
eight migrate successfully during our monthly fallover test. VG's 7
and 8 are currently empty and were created for future growth. I
recently added an LV and FS /prodsite6 to the pievg6 and thought it
was done successfully. However when the fallover is conducted
/prodsite6 does not come over and an lsfs does not show it as a
recognized file system on the B system. Any ideas?

Thanks in advance for the help
Mike

Mike wrote:
> I have a two node HACMP/ES4.4.1 cluster configured with one resource
> group on the A box and none on the B box. The cluster is configured
as
> fallover without fallback cascading. The systems are two 6H1's
running
> AIX5.1ML04 and connected two two 7133-020 fully populated SSA
> enclosures. We have 8 volume groups configured in the cluster and all
> eight migrate successfully during our monthly fallover test. VG's 7
> and 8 are currently empty and were created for future growth. I
> recently added an LV and FS /prodsite6 to the pievg6 and thought it
> was done successfully. However when the fallover is conducted
> /prodsite6 does not come over and an lsfs does not show it as a
> recognized file system on the B system. Any ideas?


Did you reimport the VG on the failover machine?
Look at "importvg -L"


--
Dana French

On 29 Oct 2004 dfrench@mtxia.com wrote:

>
>
>
> Mike wrote:
> > I have a two node HACMP/ES4.4.1 cluster configured with one resource
> > group on the A box and none on the B box. The cluster is configured
> as
> > fallover without fallback cascading. The systems are two 6H1's
> running
> > AIX5.1ML04 and connected two two 7133-020 fully populated SSA
> > enclosures. We have 8 volume groups configured in the cluster and all
> > eight migrate successfully during our monthly fallover test. VG's 7
> > and 8 are currently empty and were created for future growth. I
> > recently added an LV and FS /prodsite6 to the pievg6 and thought it
> > was done successfully. However when the fallover is conducted
> > /prodsite6 does not come over and an lsfs does not show it as a
> > recognized file system on the B system. Any ideas?
>
>
> Did you reimport the VG on the failover machine?
> Look at "importvg -L"
>

If you are not comfortable with 'importvg -L' (there are some caveats),
you should just export and re-import the pievg6 VG on node B. 'importvg
-L' will not work in all situations.

Brandon

As I suspected a reimport is my best bet, Now I have to get the user group
to give me a little downtime to do this. Thank You both for taking the time
to answer.

Mike

"Brandon" <bb@zbeba.com> wrote in message
news:Pine.LNX.4.44.0411030310130.25880-100000@boomer.boles.local...
> On 29 Oct 2004 dfrench@mtxia.com wrote:
>
> >
> >
> >
> > Mike wrote:
> > > I have a two node HACMP/ES4.4.1 cluster configured with one resource
> > > group on the A box and none on the B box. The cluster is configured
> > as
> > > fallover without fallback cascading. The systems are two 6H1's
> > running
> > > AIX5.1ML04 and connected two two 7133-020 fully populated SSA
> > > enclosures. We have 8 volume groups configured in the cluster and all
> > > eight migrate successfully during our monthly fallover test. VG's 7
> > > and 8 are currently empty and were created for future growth. I
> > > recently added an LV and FS /prodsite6 to the pievg6 and thought it
> > > was done successfully. However when the fallover is conducted
> > > /prodsite6 does not come over and an lsfs does not show it as a
> > > recognized file system on the B system. Any ideas?
> >
> >
> > Did you reimport the VG on the failover machine?
> > Look at "importvg -L"
> >
>
> If you are not comfortable with 'importvg -L' (there are some caveats),
> you should just export and re-import the pievg6 VG on node B. 'importvg
> -L' will not work in all situations.
>
> Brandon
>

Followup - the reimport worked like a charm. After reading up on the -L flag
I decided to do a straight import with the cluster down and everything
worked fine. In the future I will run the export/import cycle a couple of
times to ensure success. Now if I can only get the user group to get enought
down time to test the changes..........

"yarboon" <mhsrdh@yahoo.com> wrote in message
news:haqjd.18836$As5.3212@tornado.tampabay.rr.com. ..
> As I suspected a reimport is my best bet, Now I have to get the user group
> to give me a little downtime to do this. Thank You both for taking the
time
> to answer.
>
> Mike
>
> "Brandon" <bb@zbeba.com> wrote in message
> news:Pine.LNX.4.44.0411030310130.25880-100000@boomer.boles.local...
> > On 29 Oct 2004 dfrench@mtxia.com wrote:
> >
> > >
> > >
> > >
> > > Mike wrote:
> > > > I have a two node HACMP/ES4.4.1 cluster configured with one resource
> > > > group on the A box and none on the B box. The cluster is configured
> > > as
> > > > fallover without fallback cascading. The systems are two 6H1's
> > > running
> > > > AIX5.1ML04 and connected two two 7133-020 fully populated SSA
> > > > enclosures. We have 8 volume groups configured in the cluster and
all
> > > > eight migrate successfully during our monthly fallover test. VG's 7
> > > > and 8 are currently empty and were created for future growth. I
> > > > recently added an LV and FS /prodsite6 to the pievg6 and thought it
> > > > was done successfully. However when the fallover is conducted
> > > > /prodsite6 does not come over and an lsfs does not show it as a
> > > > recognized file system on the B system. Any ideas?
> > >
> > >
> > > Did you reimport the VG on the failover machine?
> > > Look at "importvg -L"
> > >
> >
> > If you are not comfortable with 'importvg -L' (there are some caveats),
> > you should just export and re-import the pievg6 VG on node B. 'importvg
> > -L' will not work in all situations.
> >
> > Brandon
> >
>
>