I write this blog in response to this blog and associated comments left by the community members mostly SAP Mentors.
A consulting manager's view according to SAP Mentor John Appleby:
Unfortunately the CVs that cross my desk are more often than not people have the useless TZHANA or TZH100 certifications. The certification in HANA devalues consultants.
TZHANA or TZH100 certifications, useless and devalues consultants? Too strong statement in my opinion. It is not true statement - see below for details - however; and even if it is true, I - who spent ~US$6,000 on training/certification in 6 months - don't want to hear it from an SAP Mentor in a public forum.
and I do not value certification. I value great people with great skills and customer focus. Some of them are certified. There is no correlation between great people and certification and there is a correlation between under qualified people and certification.
Since some of those great people are certified, I can't explain why the consulting manager pre-judges the certificate as useless. Does anyone else think it is fair - specifically to people like me who spent ~US$6,000 in the last 6 months + loss of income + time/efforts investment?
SAP's training programs and certification examinations - if attended sincerely and appropriately - would be very good resources to gain knowledge. If one believes in "knowledge is power", then I would highly recommend SAP's training programs. If one's sole motive is just to get SAP job after getting a certificate, then I wouldn't recommend it.
I'm going to tell 3 stories to illustrate how SAP's training program helped me acquire knowledge and perform my job better.
Before I begin telling my stories, here is a brief background:
I'm an independent consultant working as Basis Administrator. In the last 6 months, I've attended 3 training programs and 2 certificate exams:
- OS/DB Migration Consultant Training
- OS/DB Migration Consultant Certificate
- TZHANA Training
- TZ300 Training
- TZHANA Associate - Certificate
Story 1: Reduction of EXP/IMP time from 48+ hours to less than 12 hours
In March 2011, I was working on BW unicode conversion project. The production system's size was 1.1TB. The conversion took 48+ hours.
In Dec 2011, I attended OS/DB migration training program; I received the certificate in Jan '12.
Since November 2011, I've been working on virtualizing the same BW system. It has grown larger since March '11 and was 1.3TB when we migrated that DB from a server to virtualized zone in April '12. We also upgraded the DB to Oracle 11g and took advantage of "Compress for OLTP" feature(available since 10g).
We asked for 48 hours outage in April '12. Two days before the scheduled migration, some emergency came up and the customer suggested the migration be postponed as the business users needed the system on Sunday 8am instead of 8am Monday. At the same time, the customer also explained the need that the migration be completed sooner due to go-live of new reports in two weeks. So they asked us to explore the possibility of completing the migration before 8am Sunday. We accepted the challenge and completed the DB migration in 12 hours.
How did I reduce the runtime from 48+ hours to 12 hours?
From training program in just 3 days (+ time spent in preparing for the certification exam), I learned:
- the availability and the usage of Time Analyzer tool
- How to read the report generated by the Time Analyzer tool
- Advantages of table splitting
- How to implement table splitting based on Time Analyzer's output
- Running Export/Import in parallel
- The number of available options to run export/import in parallel
- How to choose a right parallel option for a customer
- How to use some undocumented parameters to SAPINST
- and more....
I certainly am not making more money due to the training program and/or the certificate I earned. However the customer thinks highly of me due to:
- Considerable reduction in the migration time and
- My willingness to take initiative and learn newer ways of performing tasks
They may not pay me more;however they may keep me longer. Regardless I learned something new through training program and the customer provided me an opportunity to test what I learned. As a result, not only I've theoretical knowledge but practical experience as well. Knowledge is Power, isn't it?
I really wished I had attended that training program sooner.
Story 2: Designing Partitions in SAP-HANA
I've decided, as a result of my success with OS/DB migration training program/certificate, not to waste anytime in signing up for SAP-HANA training programs. Believe me, had I thought attending a training program was a waste of time/efforts based on OS/DB migration training, I would not have signed up for TZHANA and TZ300 training programs. First I signed up for TZHANA;enjoyed that training and am glad I attended it.
One of the slides from that training program is shown below:
Based on the slide shown above, I knew we shouldn't create more partitions than the number of blades. And this blog - Great blog. John Appleby took time to explain what he learned during data load - discussed creating 26(+1) partitions in one-node server. As a result of what I learned in HANA class, I was able to ask this question:
On partitioning: I assume - based on 512GB size - you used one-blade SAP-HANA. From performance standpoint, is there any advantage(other than 2b records limit) of partitioning tables on one-blade SAP-HANA appliance? Wouldn't this be counter-productive in some cases as CPUs now need to spend time splitting the job/scheduling/monitoring etc?
And John responded:
Yes a single 512gb 40-core system is what this is on. If you have >2bn rows you have no choice but to partition. Does this reduce query performance? It's a great question that I can't answer yet but in theory you could be correct. What I can say is that queries do use 100% of all 40 CPUs when running against 27 partitions. Next steps are to play with the data model a bit and then see what happens when we vary the number of partitons.
Would it perhaps perform better with just 2 or 4 partitions rather than monthly? That's for another blog!
Once again, the knowledge I gained was powerful and was able to ask thoughtful questions.
Story 3: High Availability
SAPPHIRENOW '12 demonstrated High Availability. IMO, what SAP demonstrated was not HA - they didn't demonstrate what will happen when all nodes fail. I wouldn't have realized it had I not attended TZ300. The training programs certainly helped me understand what is hype and what is real. My thoughts on HA are here.
PS: Since I'm traveling next 10 days with mostly no internet connection, I wouldn't be able to respond to any of your comments promptly.