Hi - newbie type question alert!  I've googled but no joy...  I have
inherited a netbackup solution and am trying to come to grips with
it... it is running netbackup datacentre 4.5 on a Sun V480 running
Solaris 8.  There are 4 media servers and policies exist for vault runs
for reach of them.
There are some vault failures that I am having trouble understanding -
when I checked the activity monitor this morning I noticed that a vault
run for the master server (policy named "vault-master" - a daily
automatic vault with a schedule window of 06:00-noon which runs the
command "/opt/openv/netbackup/bin/vltrun master-offsite") had failed
with an error status 275 - "a session is already running for this
vault".  Its start time was today 06:03.

However, when I check what else was running to see why it had failed
with such an error I see that the previous time that policy had run was
yesterday at the same time ...  it had completed successfully at 09:46
the same day.  In other words this policy was NOT running at the same
time as today's attempt...

However, further investigation showed that a DIFFERENT vault policy WAS
running at that time.  i.e. a vault policy for one of the other media
servers (policy named "vault-ms2") which runs the command
"/opt/openv/netbackup/bin/vltrun ms2-offsite".

master-offsite and ms2-offsite  are both definied within vault
management the former for all backups on all clients on media server
"master" for all policies and schedules, and the latter similarly
except for media server "ms2".  The same source volume group is used
for both...

Having checked all policies are set on the master server - there are no
policies defined on the actiual media-server itselkf (ie ms2).

So - the basic question is...

does "vault-master" fail because "vault-ms2" is running, even though
they were presumably intended to be different jobs handling different
backups from different media servers?

Secondary question(s) :  If the above is correct should the
configuration be (to make the initial intention work as originally
planned) that each media server should be controlling its own
backup policies and schedules etc.  (although there doesn't appear to
be an option to do this via tyhe gui which suggests this sin;t
feasible), or should it be (given ALL policies etc are defined on the
master server) that their should be one single vault policy that vaults
all four media-servers in one hit?


And if the above ISN'T correct (ie the failure 275 is not due to the
other vault job running) then has anybody any suggestions as to its
cause?

Thanks In Advance!

Ian

[ Post a follow-up to this message ]

From reading the vault manual while doing an install last week I noticed
that you cannot have 2 instances of vault running at the same time. You must
change your policy schedules to make this work. No overlaps or possible
overlaps.



"didds" <didds2@excite.com> wrote in message
news:1117189244.571611.302080@g14g2000cwa.googlegroups.com...
> Hi - newbie type question alert!  I've googled but no joy...  I have
> inherited a netbackup solution and am trying to come to grips with
> it... it is running netbackup datacentre 4.5 on a Sun V480 running
> Solaris 8.  There are 4 media servers and policies exist for vault runs
> for reach of them.
> There are some vault failures that I am having trouble understanding -
> when I checked the activity monitor this morning I noticed that a vault
> run for the master server (policy named "vault-master" - a daily
> automatic vault with a schedule window of 06:00-noon which runs the
> command "/opt/openv/netbackup/bin/vltrun master-offsite") had failed
> with an error status 275 - "a session is already running for this
> vault".  Its start time was today 06:03.
>
> However, when I check what else was running to see why it had failed
> with such an error I see that the previous time that policy had run was
> yesterday at the same time ...  it had completed successfully at 09:46
> the same day.  In other words this policy was NOT running at the same
> time as today's attempt...
>
> However, further investigation showed that a DIFFERENT vault policy WAS
> running at that time.  i.e. a vault policy for one of the other media
> servers (policy named "vault-ms2") which runs the command
> "/opt/openv/netbackup/bin/vltrun ms2-offsite".
>
> master-offsite and ms2-offsite  are both definied within vault
> management the former for all backups on all clients on media server
> "master" for all policies and schedules, and the latter similarly
> except for media server "ms2".  The same source volume group is used
> for both...
>
> Having checked all policies are set on the master server - there are no
> policies defined on the actiual media-server itselkf (ie ms2).
>
> So - the basic question is...
>
> does "vault-master" fail because "vault-ms2" is running, even though
> they were presumably intended to be different jobs handling different
> backups from different media servers?
>
> Secondary question(s) :  If the above is correct should the
> configuration be (to make the initial intention work as originally
> planned) that each media server should be controlling its own
> backup policies and schedules etc.  (although there doesn't appear to
> be an option to do this via tyhe gui which suggests this sin;t
> feasible), or should it be (given ALL policies etc are defined on the
> master server) that their should be one single vault policy that vaults
> all four media-servers in one hit?
>
>
> And if the above ISN'T correct (ie the failure 275 is not due to the
> other vault job running) then has anybody any suggestions as to its
> cause?
>
> Thanks In Advance!
>
> Ian
>

[ Post a follow-up to this message ]