One of the client's requirements when implementing an AIOps solution is that the entire Netcool on-premises environment must be decommissioned after AIOps is made available.

In other words, at the end of the project, all functionality offered by Netcool on-premises must be migrated to the cluster that supports AIOps. All servers allocated to Netcool on-premises must be decommissioned.

We have observed that it is possible to migrate Netcool Impact and ObjectServer to the cluster. We have already started this process.

The biggest question, the reason for this inquiry, is how to migrate the probes (basically JDBC, MTTrapd, and Message Bus) to the OpenShift cluster?

Will it be necessary to create a POD for each probe?

Or, must the probes remain on-premises?

We thank you in advance for all your support.

Hi Antonio,

There are some probes which have been made available for deployment in OpenShift; these can be found in our documentation here. For other probes, there is not an OpenShift deployment which is supported by the product team.

However, whilst there isn't official support for it, it is possible to build your own image from a probe and run this in OpenShift. There is some information in the following GitHub repository from someone who has done this in the past: https://github.com/styblope/probe-container .

Note that this all relates to probes (i.e. for collecting event/alert data), not gateways (i.e. for synchronising data outbound from, or between Object Servers). Gateways are not supported in AIOps; instead there are connectors as part of the product which serve some of their use cases. Depending on your use case for the JDBC gateway, this could be covered by the DB2 Connector or Impact integration.