What is the Accuracy of CONSILIO IT on MRP on HANA Performance?
Executive Summary
- This is the Consilio S/4HANA, which is part of our research study.
- We evaluate the accuracy of this case study.
Introduction
CONSILIO IT is an SAP consulting partner, and consulting partners usually provide highly inaccurate information about SAP. CONSILIO IT makes many peculiar claims in this article, including the idea that MRP processing time is a major problem at many companies and is resolved with HANA. CONSILIO performed a test on MRP with HANA to prove this. CONSILIO also claims that HANA is vastly superior for MRP simulation and that they have performed a comprehensive TCO and ROI calculation on HANA. You will see how accurate CONSILIO IT statements from this article.
Our References for This Article
If you want to see our references for this article and other related Brightwork articles, see this link.
Notice of Lack of Financial Bias: We have no financial ties to SAP or any other entity mentioned in this article.
HANA on MRP
Here is how the article begins…
“In Material Requirements Planning (MRP), time-optimized and transparent processes are a critical success factor, but every minute, especially in economic turbulent times. Innovative MRP solutions based on the real-time platform SAP HANA convince with an unparalleled performance gain.”
This description starts off using deception. MRP processing is not an issue at companies unless they have misconfigured their ERP system. Also, there is no evidence that current times are more turbulent than previous economic times.
Finally, the real-time nature of HANA is not proven. HANA performs worse than Oracle 12c in analytics and has no track record of delivering well in processing or transaction processing, which is why SAP publishes no benchmark for transactions, while it did in the past. This point is covered in the following quotation from the book SAP Nation 2.0.
“It has not helped matters that SAP has been opaque about HANA benchmarks. For two decades, its SD benchmark, which measures SAP customer order lines processed in its Sales and Distribution SD module, has been the gold standard for measuring new hardware and software infrastructure. It has not released those metrics using a HANA database. One of the (unsatisfactory) excuses offered is that the expensive hardware needed to support such a test in a lab is better shipped to paying customers.”
“In the meantime, database players like Oracle and Microsoft are offering their own in-memory switches while supporting the systems management and IT talent infrastructure most CIOs have built over the last few decades. So in a couple of years SAP may have to reverse source about database portability. SAP has already shared dates for continued support of the Oracle Database 12c in-memory switch and its Engineering Systems products like Exadata and the SPARC SuperCluster.”
HANA is an analytics database that has yet to master (updated information coming to us in September 2018) dual model database processing with anywhere near the promised performance.
Let’s move to the next quote, which addresses the performance improvement.
“27 minutes instead of 22 hours for material requirements planning – this is the performance improvement achieved by an MRP solution with SAP HANA technology compared to a classical MRP. These impressive figures have recently provided a test scenario for the specialist solution provider CONSILIO IT-Solutions, a company founded by former SAP and KPMG managers, who is constantly testing and developing new developments for its customers in the logistics, production, organization and information technology industries in innovative applications for its customers.”
This performance improvement is not matched by any observations that Brightwork Research & Analysis has collected (which include both client data points and benchmarks).
Quite the opposite.
Our observations point to MRP performing quite a bit worse than MRP on even older versions of Oracle or SQL Server. This result was either found through rigging the test or by merely making up this number. Remember, SAP has never produced a benchmark indicating anything like this. We cover this in the following article What is the Actual HANA Performance?
Was This Article a Paid Placement?
CONSILIO IT wants to sell HANA and S/4HANA, and they are not a reliable guide for benchmarks. Is this publication validating any of these statements, or do they come straight from CONSILIO IT? Is this entire article looks like a paid placement from CONSILIO IT?
“The latest new development from SAP, the HANA database system, is proving to be a powerful efficiency lever thanks to in-memory computing.”
All computing is “in memory,” which we covered in the article How to Understand Why In-Memory Computing is a Myth. So this lack of distinction makes no sense. HANA proposed that it put the entire database into memory, a technique considered inefficient, but then later, we found that HANA does not work this way. Its memory optimizes. HANA uses a great deal of memory, as we cover in the article How HANA Takes 30 to 40 Times the Memory of Other Databases, and that is the distinction. But benchmarking shows that HANA has difficulty addressing the memory in the hardware footprint.
“The technology offers increased speed in compressed storage, networking and real-time analysis: the availability of materials and their stock ranges can be displayed in real-time.”
If HANA cannot outperform other similar databases with far smaller hardware footprints, why would any of this be true?
Material and stock ranges are already displayed in real time in every system we have reviewed. It is unclear what CONSILIO IT even means here. It may only be repeating what SAP marketing.
Compression is irrelevant as storage is quite inexpensive. Do we cover this in the article What Was John Appleby’s Accuracy in Moving BW to HANA?
The comment on networking is a non-sequitur.
With or without HANA, “real-time” analytics are already possible on all data platforms if the data is prepared. It depends upon what data is a question — but again, HANA does not show performance advantages over competitive databases.
“This demonstrates a lot of potential, rightly demonstrated in the right applications, such as the practice test mentioned at the outset. HANA-based MRP solutions can dramatically accelerate complex material supply scheduling and capacity planning.”
We will discuss our observations which contradict this further in the article.
The Supposed Practice Test with Realistic Data Volume
“Background of the practice test: Many companies suffer from a too slow material requirements planning run because the technical possibilities of classical databases do not give a faster data processing.”
No, that is incorrect. The only reason for the slow MRP observed is when the system is not configured correctly.
Second, as HANA is an analytics database, it should not be expected to improve MRP performance. Real life observations indicate that none of the statements SAP has made about performing OLTP and OLAP out of the same database are true.
“Frequently, material requirements planning is subject to constraints, especially when new concepts, such as restriction-based and optimizing methods, are to compensate for the deficits of the MRP. The data volume increases and the run times of the programs are even longer.”
This paragraph is virtually incomprehensible. So we have to guess what CONSILIO IT actually means.
New methods for supply planning have existed for some time. But they were developed not because of performance problems with MRP but because the methods were considered improvements on MRP regarding the planning output. These methods are covered in the book Supply Planning with MRP, DRP, and APS Systems. But how does running MRP faster address those deficits? For instance, unlike inventory optimization and multi echelon planning, MRP does not know the stocking levels of locations surrounding a particular planning location. Running MRP faster does not change that limitation. Running MRP faster also does not allow MRP to constrain resources.
CONSILIO IT seems deeply confused about the issue of MRP processing speed versus what MRP does versus other supply and production planning methods.
And Now on to Big Data?
“It was clear to us from the outset that the Big Data technology HANA offers completely different possibilities for our customers, not only in terms of optimizing the supply chain”, says David Reibnegger, project manager at CONSILIO.”
This is a massive switch in gears on the part of CONSILIO IT.
- First, Big Data is not related to anything written above.
- Secondly, HANA has nothing to do with Big Data; as described in the following article, HANA Big Data Equals Big Failure.
In terms of optimizing the supply chain, there are several methods available in supply and production planning that are “optimal.” One is inventory optimization, and the second is cost optimization. Why use the term optimization as a general term (i.e., meaning to improve) when there are two methods available that deal with real optimization?
CONSILIO IT’s Fake MRP Test
“An ECC 6.0 system with HANA DB, the so-called MRP Live, and a realistic business scenario with three plants, were created for the practice test he carried out, creating more than 300,000 materials (including configurable materials) with up to 17 disposting points. The comparison data provided a classic ECC with a MaxDB and identical master data.
The larger the amount of data, the greater the effect
The planning process had to resolve 80,000 pre-planning requirements for finished products as well as 300 customer orders for configurable materials over a period of 12 months. The products were partly planned, partly consumption-controlled. “In order to present the effect of the HANA database as much as possible, we began with small amounts of data and successively increased the volume,” says Reibnegger, summarizing: “The result speaks for itself: The classic MRP based on MaxDB Parallelization needed 22 hours for the planning of the requirements and generated 10 million planned orders – with 52 million secondary requirements and 7 million purchase requisitions including stock transfer requirements.”
All of this seems quite unlikely. This has never happened in any of our observations, and it makes no sense from the perspective of HANA’s design and what it has proven to do well previously.
It also makes no sense in terms of what HANA has shown that it is capable of. CONSILIO IT goes on to make the following ridiculous claim.
“In addition, the HANA database and the newly developed HANA MRP were able to play out their strengths with increasing data volume as the data volume increases. The more data, the greater the time-optimizing effect of the HANA in-memory database. Businesses know more quickly when they need something.”
This quotation is headed right into crazy town. This is particularly deceptive because HANA’s memory management is significantly lower performance than the competing offerings. This means that CONSILIO IT knows this entire paragraph is false. Furthermore, no database processes with infinite improvement as more data are used. This is how a person who works in marketing thinks that databases work. If CONSILIO IT intends to lie, they could at least try to make the lies somewhat credible. This quotation fact checks itself.
In-Memory Computing for more Efficient Production Planning
“The enormous run-time gain opens up completely new possibilities for MRP Live on HANA-based supply chain planning. For example, planning is possible during the day so that companies can react quickly and flexibly to market changes.”
This is based upon the premise of faster MRP runtimes proposed to have been documented by CONSILIO IT. At Brightwork Research & Analysis, we question whether any of this happened.
But even if we assume it is true, the related benefits of running MRP during the day don’t make much sense.
- Currently, almost all companies perform a nightly MRP run, called a net change run. Remember, the forecast is already in the system.
- As orders come in, they decrement the forecast. Therefore, when new sales orders come in, they don’t necessarily increase demand.
The illusion being presented by CONSILIO IT is that just because a system can be updated quickly, the supply network can be rapidly updated. But it can’t.
CONSILIO IT’s Confusion Between Planning and Execution
Real supply chains have real lead times and real constraints. The overall presentation by CONSILIO IT is an oversimplification of how supply chain planning reacts to changes. MRP is a planning method — emphasis on planning. Planning means an endeavor for the future, while execution requires instantaneous changes (on some occasions) but not planning.
Yet, the overall presentation by CONSILIO IT is an oversimplification of how supply chain planning reacts to changes. MRP is a planning method — emphasis on planning. Planning means an endeavor for the future, while execution requires instantaneous changes (on some occasions) but not planning.
And then, to top it off, in addition to misrepresenting/misunderstanding CONSILIO is basing its proposals based upon an inaccuracy regarding HANA’s improvement in MRP processing speed. CONSILIO IT is producing inaccurate information in an uncountable number of dimensions.
MRP on HANA for Better Simulation?
Here is what CONSILIO IT has to say about simulation.
“Simulation runs for the assessment of different scenarios can thus be generated and evaluated over the course of a day, unlike classical MRPs, which require several days. “With HANA, the MRP program can work substantially, unlike conventional systems, where such actions are typically shifted to night because of the amount of data,” explains Reibnegger. “With MRP Live you always have the latest planning results.”
ECC was never a proper simulation environment, and there is no evidence that S/4HANA will be better. It is sporadic to find ECC environments that perform simulation, and not enough S/4HANA environments are live to say much of anything about how companies use it.
Simulation encompasses much more than speed. MRP is better simulated in an external system that is designed for simulation. This is covered in the book How to Repair Your MRP System.
Some of the concepts are covered in the following article, An Analysis of the Commonly Listed Problems with MRP.
“A more efficient planning with the MRP Live is also possible for companies with highly complex products, a large number of orders or industries with short-term demand fluctuations. As a result, companies can react very quickly and flexibly to new customer requirements during production planning and control, or to exceptional situations in production.”
This seems to imply that products with extensive BOMs could be planned better. However, current ECC systems typically don’t have a problem exploding very complex BOMs. Companies with complex BOMs tend to have fewer overall SKUs, so it manages to balance out.
The later part of the quote goes back to the topic of reacting quickly. However, just because the computer can recompute rapidly does not mean the overall supply chain or the production floor can respond quickly.
CONSILIO IT Implies it Has Performed Full TCO and ROI Analyses of MRP on HANA
“Even if the MRP Live is not yet fully developed and certain functionalities such as the termination of planned orders or long-term planning are still in development, the costs for an MRP solution based on HANA will have been quickly amortized, especially because of the supply Chain ‘time’ is de facto just another word for ‘money’. And here is already true: Fast, faster, MRP Live.”
Seems like something written on the back of a cereal box.
CONSILIO IT seems to be underplaying the very high costs associated with HANA. And of course CONSILIO IT’s consulting costs.
- Does CONSILIO have any idea if this is true?
- Have they calculated a full TCO for ERP on HANA? A full ROI?
- How do they know what they are writing is true, or are they merely guessing. They should say.
The Real Story on MRP on HANA from a Research Entity Rather than Someone Trying to Sell HANA and S/4HANA Services
In the previous Brightwork article, How to Interpret the Performance Problems of MRP on HANA,
We covered the observations that have been coming in about how the many problems have surfaced when running MRP on HANA.
Conclusion
CONSILIO IT is making some outlandishly false claims about MRP on HANA and most likely lying about the performance improvement that it “tested.” They both provide incorrect information about HANA performance and then top it off by misunderstanding how supply and production planning work. The cherry on top is to propose further, falsely, that MRP run times are currently a problem at many companies that run ECC (something SAP also says but did not seem to say back when they were selling ECC for some reason).
This is all standard practice. SAP or the SAP consulting company produces exaggerated claims tantalizing to buyers, who then bring in the consulting company only to be disappointed.
Every piece of information provided here by CONSILIO is inconsistent with what we know about both supply chain planning and our research into HANA.
- This article receives a 1 out of 10 for accuracy. Nothing in it was true.
- It is obviously a paid placement, and most of the article is “written” by CONSILIO IT and then simply released through IT Daily.net.