Database Reference
In-Depth Information
1.5 an apple a Day
Prepare
and
prevent
are two very good verbs in the infrastructure vocabulary. Prepare for
the possibility of data issues with tested backups. Prevent issues with system monitoring.
As an example, one early afternoon the day before a holiday break a long-time cli-
ent reached out after he clicked a wrong button during routine maintenance and then
answered a question improperly after that click. The result was his Essbase databases
were recreated and his data was dropped. We were able to work around this moment
of panic because the system had a tested backup process. Some people have not been
as fortunate as you may read on the technical forums where they are asking what
they can do to restore their data when they have no backup; at that time it's too late
to make things better. The moral of both of these stories is you better have a good
backup if you do not want to ruin your holiday, vacation, weekend, or continued paid
employment.
table 1.11 shows a minimum set of recommended maintenance activities for your
Essbase environment.
We monitor systems in order to reduce system down time and, for those of us in any
type of customer support role, to preserve our sanity. table 1.12 below outlines the most
important items I recommend that you focus on when monitoring your environment.
Table 1.11
System Maintenance
Maintenance Activity
Notes
Essbase data backup
• Cold backups of Essbase must be performed with the Essbase services
fully shut down.
• A warm backup operation can be made with level-0 export scripts.
Application server backup
• The oracle hyperion applications store key information in the system
registry, temporary directories, user profile directories, and the
application file system.
• Ensure the full backup captures the system registry, temporary
directories, and user profile directories, and the application file
system.
• note: In order to ensure the system backups have integrity, all services
should be shut down while the system is being backed up.
• The Framework Services repository Directory must be backed up at
the same time as the BIPLuS relational database.
relational database backup
• Setup procedures to have the transaction logs cleared after they have
been backed up.
• If the BIPLuS database is restored, it must be restored in conjunction
with the Framework Services repository. Failure to keep the BIPLuS
database and Framework Services repository synchronized will cause
issues with Financial reports and Financial report database
connections.
Archive logs
• maintenance archives of log files and remove old archives. Log files
older than one month are usually not valuable.
Defragment
• over the period of several days or weeks cubes that have changing
data can become fragmented. Fragmentation can increase the size of
database files on disk and also reduce performance. Enact a scheduled
process to load and calculated data or restructure the cube.
