Installing engine tier on multiple machines, where to start?

laxman.ds · Post by **laxman.ds** » Thu Oct 15, 2015 12:27 pm

Dear Experts,

We have adopted below installation topology to maximize the utilization of products individually. As we are going with engine tier on multiple machines, once we install repository and services tiers, from where we can start installing engine tier? How it can make difference? And when we connect from DS-QS designer from client machine which engine machine project we have to select along with Services tier. (is it Project on QualityStage installed Engine machine or DataStage installed engine machine)? While designing the job, all stages for datastage and qualitystage will be available or is there any constraint? Kindly respond from your experience and expertise.

Machine 1; [Repository]
Machine 2: [Services Tier] + [Engine Tier] QualityStage
Machine 3: [Engine Tier] DataStage
Machine 4: [Engine Tier] Information Analyzer + Governance Catalog

Thanks in advance for your time n response.

priyadarshikunal · Post by **priyadarshikunal** » Sun Oct 18, 2015 12:37 am

I am not sure if I understood what is being asked, please correct me if the answer seems off topic.

You can have multiple engine tiers connected to same service tier. So when you login you will have to use service tier IP and port and then in project dropdown you can find the projects as "hostname:port\projectname" or "hostname\projectname" where the hostname will be of the engine tier where the project resides.

Information analyzer uses datastage engine under the covers so you will have to make sure that datastage engine is visible to it.

Quality stage is integrated with datastage, the only separation should be, rules defined that 1 server will have only quality stage jobs and other will be pure datastage.

ray.wurlod · Post by **ray.wurlod** » Sun Oct 18, 2015 3:59 pm

One you've installed Machine1 and Machine2 (in that order) you can install Machine3 and Machine4 in either order or, indeed, simultaneously.

When you connect clients this is initially to the services tier. From there you are presented with a list of engine-based projects, qualified with machine names, from which you can choose. There is no chance of error.

Machine2 (QualityStage only) will have a substantial subset of DataStage stages, but not all of them.

laxman.ds · Post by **laxman.ds** » Mon Oct 19, 2015 11:38 am

Thanks Ray, for your reply.

We couldn't see any difference for the stages availability between the projects which are on different engine machines (QualityStage / DataStage). If you could experienced any, can you please mention in specific, that would help us in planning the different projects. Also tested running the job on QualityStage installed machine by shutting down DataStage installed machine. I would see benefit of utilizing the second engine machine by spreading the nodes in Config file across the engine machine so that more power (diskspace & processes distributed across the engine machines). Is it that way?

ray.wurlod · Post by **ray.wurlod** » Mon Oct 19, 2015 4:11 pm

In a QualityStage-only licence (if I recall correctly) you probably do not have the real time stage types enabled, nor the restructure stage types.

laxman.ds · Post by **laxman.ds** » Tue Oct 20, 2015 1:54 pm

Thank you Ray, will look at that stages availability by connecting the project on QualityStage installed Machine.
As we already planned to use CDD for real time data integration, so we can't go with project on QualityStage engine for real time stages, we have to limit real time integration on DataStage installed engine machine, while we can use the project on QualityStage installed machine for daily batch processing jobs. I hope this is the way we can utilize both the engine machines. Right?