AWR Warehouse and Data Guard – Conclusions

Unfortunately AWR Warehouse and Data Guard do not work together seamlessly, however by understanding some of the implementation details, and following some simple steps you can help to mitigate most problems that may be encountered.

It is not possible to add an instance to AWRW when it is in standby mode, worse still if you do configure the primary and then switch over to an instance not configured in AWRW, the extract jobs will continue to dump files to local storage, but those files will never be transferred or loaded.

For this reason, I would suggest if setting up a pair of instances in data guard configuration from scratch to consider adding both to AWRW as part of initial configuration process. If this is an existing system where you cannot switchover to add the standby to AWRW, you should consider adding it the next time you do switchover.

Whenever you add a standby instance to AWRW make sure to backup the extract metadata table beforehand, and restore it before performing the first extract (remember to fix the sequence on this table too), to make sure you do not extract the same snapshots from multiple instances.

While it is possible to update the mapping table to force AWRW to upload from both instances to the same DBID, I think it is not worth the risk you run of breaking things by performing such changes. The other downside is that removal of either instance from AWRW triggers the removal of AWR history from both instances from the repository.

The downside of having the different instances of the same database mapped to different DBIDs is extra complexity in queries you run against the AWRW for analysis, especially if you switchover for significant periods of time. In addition, AWR reports or ASH analytics will not be work on consolidation of both instances, you will have to perform those tasks against each instance separately.

In addition as long as you are on DB plugins 13.2.2 or later (and if you are not you should think about upgrading), you may also consider updating the properties file such that data is extracted to a location accessible from both instances. That way files extracted from one instance will still get uploaded from the other after switchover. If you do not, you may need to perform some manual steps to ensure snapshots are not missed during switchover.

One thought on “AWR Warehouse and Data Guard – Conclusions

  1. Pingback: AWR Warehouse and Data Guard (Part One) | hkpatora

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s