High Availability
Micro Focus CMS supports application clustering. UCMDB High Availability (HA) solution is desirable when your CMS users are expecting zero downtime and CMS data is being used for critical business applications.
Facts on HA - Application Clustering Model
- This model requires external access to the UCMDB servers via the load balancer, where two or more UCMDB servers are configured.
- All of the UCMDB servers are active at any given time and can handle both read and write requests.
- Requests are distributed to the UCMDB servers in the cluster by the load balancer.
- All read requests are shared evenly among all of the UCMDB servers (Readers), only one UCMDB server (Writer) is also responsible for write requests at one time.
- Any write requests received by a Reader are passed to the Writer.
- Any of the UCMDB servers can take over the Writer role in the case that the Writer becomes unavailable.
Load balancer Requirements
- Cookie insertion (in all responses to clients)
- Health monitors
- Redirect rewrites (for SSL)
HA Logs
- cmdb.ha.log – This log displays informational messages in short format. In case of exception only the message appears.
- cmdb.ha.detailed.log – This log also has exception stack trace, and provides more details in TRACE mode. For example, each HA request has a request ID, and you can follow the request’s trace between the servers.
