Search results for query: *

alternatively, if space is an issue and your partitions aren't too big individually, you add the new surrogate key column to the table, create a new empty table with the same columns and then do a partition exchange to the new table. You could then use a combination of cursors, varrays/bulk...

expdp and impdp also have parallel options so you can spread the export over several files.

There are a variety of tools. in our organization favourites are PL/SQL Developer and SQL Navigator. Both have a relatively simple to use GUI, but require a license fee. SQL Developer is also available free from Oracle, but its rather flaky to say the least.

Hi, Didn't realise you were using windows. I'm not totally sure how you set the oracle_sid, whether set oracle_sid=usclus4 does it or not. However, that error appears to be down to dbconsole not being set up. You can use the following command: emca -config dbcontrol db -repos create You'll...

Sorry forgot to mention, set ORACLE_SID to the new database and then try the emctl command.

on the server where you have created the new database try this command: emctl status dbconsole

Do you have Grid Control or dbcontrol set up? If so can you review database activity for that time?

It does sound like a possible datatype issue as Dagon suggested. Just out of interest, are you able to add optimizer hints to the code/queries to get the expected behaviour?

The rman command 'show all' will show you the default parameters. I think the default for disk is one: e.g. RMAN>show all; .. CONFIGURE DEVICE TYPE DISK PARALLELISM 1; I believe there is an equivalent parameter for sbt_tape if you are using a tape device. Otherwise you need to check your rman...

If this is an NFS mount, there are a few things you can do: 1) set event 10298 in the database (from i think this requires bouncing the database ) 2) get the sysadmin to adjust the rsize and wsize parameters for the NFS mount.

If you are using Enterprise edition, why not investigate using Oracle Streams to replicate the data?

If your source and destination O/S are the same, standby would probably be best. At the moment I am looking at migrating a 1tb database from linux to AIX with minimal downtime, and because the endian formats are different, I am using Datapump to create the destination database and Oracle Streams...

have you tried running it like this $ORACLE_HOME/bin/sqlplus?

as well as exp/imp, there's also Datapump and the transportable tablespaces option - well worth looking into, but you will need to check the supported/unsupported types...

Never used Toad, but sql*loader has a command line interface: sqlldr help=y hope this helps you get started.

If i remember rightly vmstat shows you the memory in 4K pages so some small arithmetic is required :)

My experience with our application here is that once I have removed a shared memory segment or message queue, I have to restart a component in the application (effectively reloading shared memory). You should be able to relate shared memory back to your app components by looking at the start...

You will need to add these to your package control file. Review the files under /etc/cmcluster/<package name>. Once changed these files will need to be distributed to the other nodes in the cluster.

man ipcrm

I am finding HP's tech support the most troublesome at the moment. I will always log a webcase with them now and only tick Reply electronically. The other annoying thing is that HP's overseas support centres never seem to have the 'appropriate knowledge' and have to refer it somewhere else....