Product Information
Connect Multiple SAC Environment To One SAP Cloud Connector
Introduction:
I recently came across a requirement to test and create comprehensive list of functions SAC (SAP Analytics Cloud ) supports, as a planning environment. As an organization we have a large footprint in terms of planning in different line of businesses. Not to mention we have multiple non SAP planning tool used for years and the biggest challenge is to have a consolidation environment. Every planning tool/ environment we use have pro’s and con’s one its own. So the goal is to find out whether SAC can support large volume and complex planning scenarios. This may let us find a sweet spot of planning features and performance that we are looking for.
Though, this is not the topic I am going to discussion today, may be another time. Today I want to share what I found and tested even before I started testing SAC planning component.
This blog post is to share my experience and steps involved in switching SAC environment from Neo to CF (AWS data center).
Why we switched from Neo to CF (AWS)?
One of the most hyped feature of SAC planning is, its predictive/ training (Classification/ time series/ regression) function and not to mention your existing planning licenses should let you take advantage of this feature. Predictive comes out of the box in SAC, no additional cost (apart form licenses). Our business process can take advantage of Forecasting done via SAC predictive and later adjust numbers as per requirement. Business user or Business Analyst would really appreciate a AI based training option at the start of their planning and forecasting cycle. I assumed that I can start testing these features in our existing SAC tenant (Neo – Host on SAP Data Center).
Surprise! Surprise! “predictive” feature is not available in our SAC environment. Even though we have planning licenses and other planing functions are available. We checked with check with our SE and found that some features (including predictive) is not available in Neo clients. Beside there is no plans to include predictive in applications hosted on Neo. We were instructed to switch to CF (AWS) without any additional cost. Move all our objects and connections to new tenant.I think describing migration process and our experience would required a separate blog post.
So here we are, planning to switch from Neo to CF.
Prerequisite:
Assumption is your existing SAC environment is already setup to pull on premise data via Cloud Connector > Cloud Agent ( for legacy protocol). You are planning to connect a new SAC environment, reusing the existing Connector and Agent setup. This is setup environment (both Sac’s) for transition period.
1.Reassign System Ownership:
Setup “System Owner” access, once you receive confirmation from SAP SE. This is an optional step if “System Owner” is not granted to a user planing to perform initial steps after system takeover. You can even request SAP to assign “System Owner” before handover. Though swapping ownership can be done at later-stage and I prefer this as it is quick.
Assuming someone from your organization or someone from SAP may be designated as “System Owner”. You can request System Owner to add you (system admin) as BI user.
At this point system will prompt you to assign new role to previous system owner. You can assign any role (as per requirement).. System will not let you delete user without assign a role first. In case you want to delete previous user.
2.Cloud Agent and Connector entry:
At this stage you need three information –
- Region Host : Where your SAC tenant is hosted (data center)
- Sub-account information
- Email ID linked to S-User ID
Region host can be found from the SAC URL itself.
Trivia Info –
SAC URL usually consist of customer info (name) + data center code + “hcs.cloud.sap”/ “sapanalytics.cloud”
example- <Customer Name>.us10.hcs.cloud.sap
Here US10 is the Region Host.
How can I confirm if I am on Neo (SAP Data center) or CF (non SAP data center) tenant?
Neo tenants usual have 1 digit and CF data centers have 2 digits. Hence in above example SAC tenant is on non-SAP data center in US.
Login to cloud connector URL. https://<Cloud Connector Host>:8443/
At this point you should see your existing sub-account name on top of the screen.
Fill in details found in step 1. Password will be your S-ID password.
At this point list will show two Subaccounts (new and old one). Please note that Status will “missing resources”. We will update resource in subsequent steps.
Steps:
Check the status of the newly added subaccount. It should show Connected now.
3. Add datasource in SAC
Host will be virtual host name in step 2; Port 8080; User Name and Password from tomcat-users.xml (location: \Apache Software Foundation\Tomcat 8.0\conf) file. Set during initial setup (already set & available).
Conclusion:
SAP provided a easy integration steps to configure and connect multiple SAC tenant to your existing Cloud Agent (no need to setup separate Agent).
Some features (like predictive) are only available on applications hosted non SAP data centers (Cloud Foundry – CF). This trend of releasing new cutting edge feature to environment hosted on non SAP data center only, makes me feel like SAP is trying to get away from SAP data center. Rely more and more on other cloud service providers. I guess this is to going to be SAP’s cloud strategy going fwd and we will see this trend with other applications. Not to mention new collaboration with cloud service providers (MS Azure and GCP) speaks for itself.
Disclaimer:
The observations that I am sharing in my blog posts and the information I am presenting are collected in a careful manner by myself with the help of my fellow colleagues. However, they are not “official” guidelines released by SAP. Also, the information can be outdated as the platform is developed further. So I personally cannot guarantee for the correctness of everything I am writing about in these articles and I do not speak for SAP when I am giving you recommendations.
Thanks for sharing this. We recently setup a separate host for object migration from old tenant.
Is there a limitation on how many tenant can connect to single cloud connector/ agent?
Not that I am aware of. I think the limiting factor will be resources on the host (hosting connector and agent). If require you can separate agent and connector in separate hosts to better manage resources.
Thanks!
Multiple tenant is common from life cycle point of view. Though SAP supports import/ export and content network sharing have its own limitations - quality check is one of it.
Hopefully we will see something on LCM side in 2021.
Thanks for sharing your valuable insights.