Thank you for attending our webcast "Modernize and Automate your IBM OMSV10 on RedHat Openshift". We received more questions than we could answer, so we are posting the answers in this forum. If you have additional questions, please post them below and I will get them answered.
1. Can you please indicate the cost implications for a customer on premise to move to RHOCP? are there components that need licensing. Also, does this apply to a customer, who has non-IBM stack for app server or database.
- ANSWER - The Certified containers have a new part number
2. Can the images run on Windows also or is it Ubuntu-based?
- ANSWER - The certified containers are supposed to be deployed on Kubernetes based platforms. For development purpose these containers can be deployed on an environment where docker or podman is available.
3. Can OM app container be built with JBoss instead of websphere
- ANSWER - The certified om-app container is WAS Liberty based. The scripts provided are also to create custom images based on WAS Liberty only.
4. Can you give more insight about service layer usage
- ANSWER - Please refer to the Helm charts readme.
5. How long is typical build runtime compared to previous ant build?
- ANSWER -Not sure if we can directly compare ear build to an image build.As with the ear build you just get an application ear, but with the image build you get the app ear, Liberty & OS all bundled in a single image that is readiy runnable in a docker or quickly deployed on OpenShift. However, the amount of time taken for applying the customizations is same as traditional installation and even that can be expedited with CI/CD model.The build will be running in om-base container.
The advantage is in the deployment time compared to traditional deployment as the files/OS/Liberty images are all minimal needed for deployment and execution.
6. Will we be able to get a copy of today's slides?
- ANSWER - Yes, the webinar recording is available here and sent to participants
7. Is it possible to include Automated unit testing and test coverage into the build and deployment process?
- ANSWER - The product has provided the hooks to create custom build scripts and execute them on the om-base container. You can add your additional scripts to build the test suites to those hooks.
The helm charts provided are only for the certified oms containers
8. For autoscaling do client needs to pay if traffic increses and new pods are created?
- ANSWER - Autoscaling is a feature already available in the OpenShift platform and most of the Kubernetes platforms. There is no additional cost fo using autoscaling, however if there is additional infrastructure usage due to more number of pods/containers, there might be extra cost. This depends on the cloud/infrastructure provider where OpenShift/Kubernetest cluster is installed and the contract with them.
9. Suppose I have a file bases Integration Server and in that case if my pod goes down then what happens to those files?
- ANSWER - For file based integrations, persistence volumes has to be used to place the input files so that if the containers goes down also, they are preserved. File based integrations are generally not supported in a K8s container model to maintain higher level of security
10. Is the same concepts for auto scaling used for IBM SaaS OM customers?
- ANSWER - IBM SaaS autoscaling is based on Customer specific environment/needs
11. How can one take care of OS upgrade? Sterling OMS upgrades?
- ANSWER - It depends on where the OpenShift cluster is installed. For public cloud providers, the upgrades are handled by the cloud provider itself. For on-prem installations, OpenShift provides processes to upgrade the Platform/Underlying OS.
The upgraded OpenShift version should be supported by the OMS certified containers' System requirements.
In either case, Sterling Upgrade has to be handled by the application team. Product will provide new versions of certified images which can be downloaded from entitled registry/website and also apply the upgrade steps to update the database.
12. Need more insights on how traffic is handled by router and service
- ANSWER - Please refer to OpenShift Routes/Services documentation
13. From which FP onwards does OMS 10 support containerization?
- ANSWER - From OMSv10.0 onwards containerization is supported.
Certified containers are separate part numbers to be bought separately and will not be released via Fix Packs. Customers who have the appropriate license keys can download the images directly from IBM Entited Registry. THe product KC covers those instructions
14. Sterling is heavily dependent on CDT and that is common for all the containers. How are interdependent configurations handled during deployment? Eg: I want to deploy some changes for webstore but with CDT that would impact the agents as well.
- ANSWER - CDT should be done as part of the deployment to each environment. Om-base container can be used to execute CDT export and import scripts.
<We may not be able to separate cache changes from not propogating to agents uness the specific JVM has disabled cache for that table in question -- @Murthy to confim>
Helm commands/OpenShift UI can be used for stopping and starting current running containers in one shot.
15. Any idea of extending this to 9.5v?
- ANSWER - OMSv10 is minimum version required for using containerization and reference helm charts
16. Is the CPU usage matrix used for agent/integration server autoscaling or there are other matrix like quedepth.
- ANSWER -There is no autoscaling based on QueueDepth.
17. Do we have support for Oracle and Weblogic MQ?
- ANSWER - Oracle and DB2 are the supported databases. Any supported JMS provider can also be used. Please see System Requirements document.
18. Is there a support for Oracle based DBContainers ?
- ANSWER - Oracle based db containers can be used for development
19. Will there be a DTK for Openshift?
- ANSWER - There is no DTK specific for OpenShift. The certified containers can be deployed on local machines with docker or podman. Please see the customizing the OpenShift container section in KC
20. Are the autoscaling/self-healing capabilities features of Kubernetes, or RHOCP?
- ANSWER - Yes, this is a feature of Kubernetes/RHOCP
21. What considerations/risks we have while we do an upgrade from an existing Sterling Version to V10 using containers?
- ANSWER - Consideration on File Based Integrations using PV. There is no physical access to the Kubernetes Worker nodes.
Refer KC for OMS traditional to Container Upgrade.
22. What are the reasons for recommendation "DB Server as Standalone" on production?
- ANSWER - DB2 servers require vertical scalability. Certfied DB2 containers with production ready HA/DR capacbilities will be made available in upcoming releases & OMS product will consider support of it in future roadmap
23. Are there any strong recommendations on using container based mq, db2 versus explicitly installed mq, db2
- ANSWER - Certified MQ containers are available for IBM Cloud Pak for Integration. Please refer to DB2 & MQ documentation to read about the containerized versions
24. Is there any alert mechanisms in case of memory leaks or container restarts for the application team to debug and perform RCA.
- ANSWER - The alert mechanism is based on the Kubernetes Platform Chosen. Monitoring console in Openshift has mechanism to configure for alerts. In addition the events in the pods should also given an indication of whats happening to the container.
25. On what basis pods can scale up and scale down automatically?
- ANSWER - Based on CPU and Memory Usage (beta)
26. Can we implement this in on prem application deployment?
- ANSWER - OpenShift can be installed on on-prem provided the infrastructure/system requirements are met at on-prem. You can get cloud experience in on-prem.
27. Is this session recorded?
- ANSWER - Yes, the recording of the webinar is available here.
------------------------------
Tanvi Kakodkar
Supply Chain Community Manager
IBM Sterling
------------------------------
#OrderManagementandFulfillment#SupplyChain