Red Hat OpenShift OpenShift Commons Gathering 2021 Kubecon EU, 4 May 2021

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: Banking App Modernization via the IBM Cloud Pak for Automation on OpenShift - BOK Financial

Description

OpenShift Commons Gathering at Kubecon/EU 2021
May 4th, 2021

Title:
Banking App Modernization via the IBM Cloud Pak for Automation on OpenShift Container Platform

Speakers:
Lance Preston (Bank of Oklahoma)
Faizal Khader (Red Hat)

Abstract:
BOK is an acronym for "Bank of Oklahoma". The company's banking subsidiary BOKF, NA operates in 8 states under the brands Bank of Oklahoma, Bank of Texas, Bank of Albuquerque, and BOK Financial (in Arizona, Arkansas, Colorado, Kansas, and Missouri). Bank of Oklahoma made a decision to modernize their IBM FileNet Content Management Suite in early 2020 by migrating to a container-based infrastructure. With the overall goal to improve development, delivery, and reliability for developers & operations on-premise & in the cloud for the bank. With this new environment based on Red Hat OpenShift 4x is able to scale its content management solution, reduce application delivery times while creating a scalable, vendor agnostic foundation for future cloud adoption.

https://commons.openshift.org/gatherings/OpenShift_Commons_Gathering_at_Kubecon_EU.html

A

Hello, all I'm faisal kadur, I'm a specialist essay in the sales team that manages this account. I specialize in openshift and related red hat technologies, I'm kubernetes and openshift certified.

A

I closely work with lance preston, the infrastructure architect from bokf in setting up the different openshift clusters until production environment on these clusters, ibm cloud pack for automation with its filename product is hosted.

A

I closely worked with him, helping help, helping him navigate through all the challenges with the infrastructure and provided the necessary enablement to help him complete this effort.

A

Working together, we managed to tackle all the roadblocks and have led to this great success story for the bank and red hat I'll hand it over to lance to share his experience.

B

uh Hello, my name is lance preston. I am a an infrastructure architect with the bok financial corporation, which is headquartered in tulsa oklahoma.

B

I'm here today to talk about banking, app modernization via the ibm cloud pack for automation on the openshift container platform.

B

So we started our app modernization journey, basically in the fall of 2019, uh when we began planning to replace our legacy filenet platform- um and I I believe I'm sharing that information on screen now.

B

um So our uh I'll talk about what our legacy platform looked like here in just a minute, but uh basically at that time, what we needed was a solution that was going to be supported for the foreseeable future as well as being more easily scalable, and that could support the needs of our business, which included adding some additional products within that content.

B

Management platform uh provided by ibm um so in the fall of 2019 is when we really began talking seriously with uh ibm and red hat about building out a new environment uh for our content management platform, and it was at that time that they suggested that, um instead of using the you know, just upgrading the components in our legacy platform that we look at going with containers and kubernetes and specifically the openshift platform.

B

So I'll talk a little bit about what our our legacy environment looked like. So we basically had uh the the ibm filenet p8 platform, which was uh version 5.2.

B

We were running the ibm content, navigator uh kind of separately as a front end to file net for our users to access the the documents and the content within filenet. This was all running within the websphere application, server network deployment, edition uh version, uh 7-0 and 8.5.

B

We had a combination of both and these were connecting to an oracle database that was clustered within power. ha All of these uh were running on our uh our ibm power. Eight hardware running the aix-72 operating system, so I will show you what that looked like from a visual perspective here, I'm not sure if I can get that any larger there we go so um basically, uh so all of these are aix l cars in our ibm power platform right.

B

Our oracle database that the uh these lpars, which are the content platform engine, uh which is where the the majority of the content is served up right uh and it is stored on our uh network attached storage platform uh on site. This is all on premise by the way um and then the content, navigator applications uh which were load balanced and would front end the content uh platform engine applications.

B

We also have kind of a legacy piece: that's called the application engine which is used by a couple of other lines of business. So all of these are right. So the the ae servers, the application engines were running websphere network deployment, edition version 7.0, uh the cpe servers were websphere nd85 and the content navigator was also websphere nb85, but these these are all separate, aix ltars in our on-prem power, 8 environment, so growing this and expanding this right is a little bit of a chore.

B

We can obviously add additional resources to each of these servers as needed, but if we need to stand up another server right to scale out the environment, um that's obviously a lot of work. We have to build another aix lpar. We have to install websphere. We have to join that into the existing cluster and and get the applications deployed to that.

B

So the solution that we came up with uh with ibm and red hat was to deploy uh what we consider our modern environment right, which is the uh basically the ibm cloud pack for automation right. So we are currently at uh release. 20.0.3 fix pack two for the the cloud pack for automation platform um that is running in and red hat, open shift, 4.6 environment um uh on our on-prem vmware, vsphere, 7 platform right. So these all of our red hat openshift nodes are just essentially uh red hat linux, core os vms running.

B

In our on-prem vmware platform, um we have the ibm cloud pack for automation, deployed to that environment. uh We are still running the uh the content platform engine, which is the ibm filenet p8 product, but it is the the later version right. It's version 5.5.

B

It also is still front ended by the ibm content navigator, which they refer to as the uh the band application within the cloud pack for automation, uh but these are now running websphere liberty within the containers within this application.

B

So here's kind of a visual of what that looks like well, and we all are so we all of these environments are still connecting to an oracle database on-prem that still resides in our aix platform in the the ibm power8 hardware. So this now looks like this essentially.

B

So we have our our on-prem vmware cluster, which is a you know. A couple of dozen servers, at least in our production environment and those vms, are our red hat core os nodes. We've got the three master nodes. We currently have four worker nodes that are all running uh red hat 4.6.16.

B

I believe the version we're currently at um these are all on. You know: on-prem vm data source in our vmware environment, on our enterprise, san. We have a management node that is used to deploy the applications right. That's where we run the the oc or the coupe control. You know command line interface for the cluster.

B

We do also use the the openshift gui for management of that cluster from time to time, but a lot of the the tasks that we do. We do command line through this management mode.

B

um This all is still front-ended through our on-premise load, balancer appliances um and then all of the cloud pack for automation, applications, those containers and pods right. They reside within this cluster right. They run on various nodes within this cluster and then the persistent storage for those applications, as well as our content.

B

Within the filenet platform is still on our nas platform right, our network attached storage um and then those environments still connect to our on-prem oracle database uh instances which are clustered together in a power ha environment uh in running on our aix-72 platform and again, this is all on premise, um so it took us quite a while to actually get to this point. To get this environment stood up.

B

uh There was a pretty steep learning curve from our perspective, to be able to get this environment in place, uh but we did receive a lot of help from both red hat and from ibm red hat, specifically on getting the openshift cluster stood up and working and any problems we had with that and then, of course, ibm with the cloud pack for automation packages that we got deployed.

B

We struggled with it a little bit, but we were able to overcome any any issues or difficulties that we had with assistance from from both of uh red hat and ibm and have a test. Dev cluster stood up as well as a production and a disaster recovery cluster stood up.

B

This image is depicting our production cluster, um but now that we have this in place and now that it's stood up right scaling out, these applications is much easier right. We can just scale up the deployment and immediately have additional instances of the application that are serving our customers as well as uh upgrading the cluster itself is relatively straightforward.

B

It's a very simple process and then upgrading the cloud pack for automation uh packages themselves that bundle uh it's a much easier process than the legacy platform right when we used to have to do an upgrade for filenet or for content navigator, but that was pretty much uh an entire weekend right.

B

Several employees uh working together for the majority of a weekend, uh which is when we could take the down time for the application to have it offline for hours or half a day at a time in order to do the upgrades and test it out and make sure everything was working before we went back into production. You know monday morning now that upgrade process right can occur basically within a couple of hours at most generally less than an hour, and it can be done with the applications online right.

B

We just deploy the new application, uh apply, a new uh configuration file right, a yaml file and the upgrade kind of happens by itself in the background, um with the the old version of the application still available until the new application is fully deployed and those containers come online and it's just like flipping a switch and then backing it back out is just as simple right so much easier in the new environment than the legacy platform.

B

I can't really speak to the performance of the new environment. Yet while we do have this production environment built out, we are still in the process of loading content and data to the filenet platform to serve to our customers. So they currently are still using the legacy platform.

B

But as we migrate that content to the new platform and make that available, more users will eventually cut over to this new platform, and we will have a much better idea at that point of you know, load and the performance that the new environment can provide we're optimistic. We don't have any reason to believe that the new environment won't perform as.

A

B

If not better than the the legacy platform, but we have uh yet to to have any concrete data on that. um That's basically what I wanted to present to you. So thank you for your time. uh I appreciate you paying attention. I hope you find this information useful have a great day. Thank.

A

You thank you lance for sharing this wonderful insight on this topic and thank you all for listening. Lance myself and my associates are available to take any questions in the.

B

B