[Veritas-bu] Status of 24's and 66's

--------------0B0A1059D4503BBDC40FDB38
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hello All:

We are running into problems with our 3.4 and 3.2 NT clients. It
seems when the bpstart_notify takes more than 6-7 minutes to
complete, the backup fails with 24's and 66's (depending on which
retry of the backup). These clients have "large" oracle instances
running on them and the bpstart has one line just shutting down
the instance (we are not using the NB Oracle Agents). I have
tried to up the timeout values but this has not worked. We
presently backup up Oracle for Unix on at least 100 clients (big
and small) and many smaller Oracle instances on NT clients and
all run fine, it's just these larger db's on NT. We have
determined that it is not the db that is causing the problem;
people have suggested that the Oracle agent would fix, but we
have experimented with placing a "sleep" of 10 minutes in the
bpstart (no shutdown of the db) and we get the same results
(errors). I have called Veritas and they cannot duplicate the
error via making their bpstart sleep for 10 minutes. We first
noticed this on 3.2 hoping that 3.4 would correct but it has not.
We have also changed the ports on the switch since many had
suggested that these are network related errors but that has also
not worked. The only work around that we have at this point is to
have the NT admin schedule the shutdown via the AT scheduler 15
minutes before the backup is supposed to take off and schedule it
to bring it up at a certain time; this works fine but I would
prefer to use the bpstart (safer).

Has anybody seen anything like this? Would appreciate any help.

Thanks.

Charlie Somogyi
BMS

--------------0B0A1059D4503BBDC40FDB38
Content-Type: text/html; charset=us-ascii
Content-Transfer-Encoding: 7bit

<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
Hello All:
<p>We are running into problems with our 3.4 and 3.2 NT clients. It seems
when the bpstart_notify takes more than 6-7 minutes to complete, the backup
fails with 24's and 66's (depending on which retry of the backup). These
clients have "large" oracle instances running on them and the bpstart has
one line just shutting down the instance (we are not using the NB Oracle
Agents). I have tried to up the timeout values but this has not worked.
We presently backup up Oracle for Unix on at least 100 clients (big and
small) and many smaller Oracle instances on NT clients and all run fine,
it's just these larger db's on NT. We have determined that it is not the
db that is causing the problem; people have suggested that the Oracle agent
would fix, but we have experimented with placing a "sleep" of 10 minutes
in the bpstart (no shutdown of the db) and we get the same results (errors).
I have called Veritas and they <b>cannot</b> duplicate the error via making
their bpstart sleep for 10 minutes. We first noticed this on 3.2 hoping
that 3.4 would correct but it has not. We have also changed the ports on
the switch since many had suggested that these are network related errors
but that has also not worked. The only work around that we have at this
point is to have the NT admin schedule the shutdown via the AT scheduler
15 minutes before the backup is supposed to take off and schedule it to
bring it up at a certain time; this works fine but I would prefer to use
the bpstart (safer).
<p>Has anybody seen anything like this? Would appreciate any help.
<p>Thanks.
<p>Charlie Somogyi
<br>BMS</html>

--------------0B0A1059D4503BBDC40FDB38--