Stopping Cassandra nodes for data center maintenance

When bringing a data center down for maintenance, applying fixes, or applying updates to Cassandra nodes, you must stop the Cassandra nodes in the data center, so that metadata is not replicated to the data center.

Procedure

To stop Cassandra nodes for data center maintenance, complete the following steps:

Go to the <InstallDIR>/MailboxUtilities/bin directory for each Cassandra node in the data center you are bringing down, and run the stopGMData script.
Important: While a data center is down for maintenance, all trading partner activity executed through a protocol adapter (FTP, SFTP) must continue to function properly. However, the following subsystems do not operate at functional capacity until the Cassandra instances are brought back online:
  • Event processing
  • Scheduled jobs that run periodically in the background
  • One-time jobs (some UI actions and CLI scripts)