Moving transactional data - upgrades only

Transactional audit log and flow history data has been moved from Neo4j to Elasticsearch to improve the maintenance of and access to this data.

Note: This section only applies for upgrades from versions of Spectrum Technology Platform earlier than 2019.1. New installations and upgrades from the 2019.1 version are not affected in this release.
If you are performing an upgrade from a version of Spectrum Technology Platform earlier than 2019.1, the upgrade process will export all transactional data into two ZIP files labeled by the prior Spectrum version number (such as 18.2):
  • Archive_Upgrade_SpectrumVersion_history.zip
  • Archive_Upgrade_SpectrumVersion_audit.zip
The sizes of these ZIP files depends on how much data you have.
Important: To avoid building large export files, we suggest that you purge or archive the data before you begin the upgrade process.
Standard installation (server and database on the same machine)
When the Spectrum Technology Platform server and the database are installed together on the same machine, the upgrade process creates the export ZIP files in SpectrumDirectory/server/import directory. Then, the first time that the Spectrum Technology Platform server starts after the upgrade, it automatically imports data from the files located in this directory. Standard installations require no manual steps to accomplish this task.
Separate installation (server and database on different machines)
When the Spectrum Technology Platform server and the database are installed separately on different machines, you must manually move the exported files from the machine that hosts the database to the machine that runs the Spectrum Technology Platform server.
Before you start the Spectrum Technology Platform server for the first time after the upgrade, move the exported files from the following location on the machine that hosts the database:
SpectrumDirectory/repository/import
To the following location on the machine that runs the server:
SpectrumDirectory/server/import
Spectrum Technology Platform then imports the files automatically the first time that you start the server after the upgrade.

Depending on how much data you are moving, you can expect the Spectrum Technology Platform server to start much slower than usual while the transactional data is imported back into the system and into the Elasticsearch repository. After the Spectrum Technology Platform server successfully imports all the data, it automatically deletes the ZIP files from its host machine.