So on Monday, Hana celebrates first birthday ! I will be in Palo Alto to be part of the celebration. It is amazing that one year has passed since the GA announcement came for HANA. Looking back over the last year, SAP can definitely be proud. They put their heart and soul into it – from Hasso, Jim, BIll and Vishal to all the way down the labs and sales organization hierarchies.
I have had many discussions with Hasso and Vishal on Hana in the past couple of years. Their passion is infectious – and they have an excellent vision on where Hana will take SAP to. Vishal has a rockstar team – and I know most of them, and that talent runs deep many layers down the hierarchy. Engineering for Hana is definitely in good hands – and as time progresses, it will only get better.
Hype or not – last year showed the power of good PR and marketing. Jonathan Becher deserves major kudos for this. Same goes for the global communications people under Hubertus Kuelps and Mike Prosceno. Without their extreme effort, Hana probably would not have had the coverage it got amongst analysts, bloggers etc. The power of social media was fully utilized for HANA – probably more than any other product before it. It is now in a state where if someone has a question about HANA on twitter, some one from the ecosystem can clarify most of the times without needing an SAP person to jump in.
On the sales front – they got the best GM for the job in Steve Lucas. He is a solid leader and an excellent communicator, and is supported by an amazing team. They are going all out to sell Hana to customers. And from what I have seen – they have all the support from Enslin, Poonen and other big executives.
SAP also got around to enabling HANA developers on the cloud. That is a great step in the right direction. Another thing that excites me a lot is the set of Hana start up companies that SAP is supporting. Again – a great step.
The one area I am not equally happy about is SAP education for HANA. I saw somewhere that 2000 people are certified. I seriously doubt even if half those people can hold their own in a HANA project. Education needs to keep pace with product, not lag it . Parts of SAP like the CSA team have done a fair bit to help partners keep pace, but that is not sufficient. If education does not contribute in a significant way – several of those projects that start now will end in disasters next year, and none of us want that. I was extremely happy to talk to Marcus Schwarz few days ago. He is the SVP leading SAP education. He understands this better than me, and he is doing all he can to turn this around.
Last I heard , they have about 350 Hana customers and about 140+ live installations. That is not bad for a product that is only a year old by any stretch. The question is how much of acceleration will Hana see in 2012 and beyond? Will it be what moves the needle for SAP in a big way in terms of revenue and profit? I am cautiously optimistic.
Hana works now as one of the possible databases for BW. There are apparently something like 17000 BW installations around the world. If half of these customers will switch their databases to hana – that itself will make the investment SAP made in HANA worthwhile. But that is not easy to pull off for many reasons
1. Other DB vendors will do all they can to prevent erosion of their instal base. And they are all bigger companies than SAP with more money to throw at communications and product development. Interestingly, they are all good partners of SAP too. So it will be a fun dance to watch from the cheap stands.
2. Not all BW installations have the prospect for a big sale due to size of DB. So SAP will need short sales cycles to get them to act. This is not easy in several cases since SAP has not actively sold BW in the past. I haven’t seen a lot of SAP sales people articulate value of BW on HANA in a convincing manner.
3. Kind of related to above, and it is about SI partners. Despite the top leaders trying their best – partners have not really had a significant role in HANA yet. Unlike SAP themselves, SIs have articulated BW before and this is their game. By not removing the hurdles for SIs, SAP will slow down the uptake of HANA as a BW database. This is changing, but not at a pace I would like to see.
4. Although I have no real idea why – it was clear this past week on twitter that several people, including some SAP Mentors and big customers of SAP, still think HANA replaces BW. Steve Lucas posted an excellent blog, and I wrote parts of it. https://www.experiencesaphana.com/community/blogs/blog/2012/06/13/does-sap-hana-replace-bw-hint-no but that did not seem to help since questions keep coming on twitter. I also did a podcast with my DSL buddies on this topic – http://dslayer.net/dslayer/view/130-big-data-hana-vs-iq and guess that did not help either . I guess this will not get solved easily when SAP influencers themselves are not clear on the details.
5. Some of the HA questions did get answered after Hasso’s keynote showed a live demo. But there are plenty of people who need more convincing on data center sturdiness of Hana. SAP needs to step up its efforts big time on this front – and fast.
6. SAP also needs to articulate the IQ vs HANA some more. Generic statements doesn’t help. Listen to the podcast link above and listen to Clint and the DSL guys.
By end of this year, SAP will announce ECC running on HANA in some form. On the sales and communication side – SAP will hopefully manage to fine tune its act with BW so that they can go all out for ECC. In my view, ECC on HANA is not a big deal really for customers. Not everything in ECC will benefit from having HANA as the DB. Some long running batch jobs etc might improve – but I cannot imagine the majority of ECC functionality benefiting from HANA. However, for HANA to gain credibility as a pure database compared to Oracle, DB2, MS etc – it needs to work on ECC. And for that reason – I think it is necessary for SAP to get HANA on ECC right.
Finally the quest for the killer apps continue. BPC on HANA so far is the best application I have seen, and that will evolve more with time. But given the developer and start up enablement – I hope this is just a temporary issue.
So – Happy First Birthday SAP HANA ! I am looking forward to celebrating with everyone at Palo Alto and Monday, and many happy returns of the day.
10 thoughts on “Happy First Birthday SAP HANA !”
What about the mobility on HANA platform ? . I assume mobility is another major aspect that could come to limelight when ECC is migrated to HANA. This will open up new doors for various Operational aspects currently running on ECC.
very good blog and excellent post. I would add a couple of cents of thought into this. Which might not be the ones that SAP would favor most but at least reflects what we are seeing.
HANA is definitely a brutal accelerator to SAP, which has a lot of value. The question always remains if the current pricing is justified by the value given. I have seen more than just one customer saying this is great but it is way to expensive for what it is. This leads to another point SAP will have to treat sooner or later. Right now all the pricing pressure is down on their ecosystem which in turn is needed to make HANA a success, there will for sure be changes but they must be significant for the ecosystem to push.
BPC will for sure be a great asset on HANA but for those who have already BW on HANA this could mean a significant license upgrade or new HANA Landscape just for BPC…
Another piece is that the SAP Messaging must get clearer and more detailed. You mentioned Datacenter readiness. There are still things which today are not 100% solved and responses like “talk to your hw partner” are not always the best ones to give.
The race has just started and it is a lot of fun to be part of it, now we come to the second round and we need to make sure that all are rowing at the same rhythm to keep up the speed and success ! Happy 1st Birthday
Very nice blog.
thx for responding.
i don’t think we need to worry about decompression of OLTP tables as HANA will take care of the OLAP piece, no need to write back to ECC, but compressing the line item tables before throwing the data over the fence seems like an important failover/competitive point for the OLTP/OLAP DB2/HANA integration.
are you aware of any implementations addressing this particular scenario?
I have not tried it. But this would be limited only to table level replication Which does not conserve semantics of data. All other forms of etl will need to uncompress first I would think.
I would like to comment on Thiru’s comment in SIT-NYC(I speculate your question stems from that session. My apologies if I’m wrong. I wanted to make this point in my blog; however I didn’t want to digress).
Oracle’s 10g – 2 years back – released a new feature called COMPRESS_FOR_OLTP. By using that compression, customers could easily reduce the BW DB size from 18TB to less than 10TB – Thiru mentioned 18TB became 12TB using DB2 compression. Remember if DB has not been reorganized for a long time, then the savings in oracle could be more than 50%. And the performance improves in spite of little more CPU requirements to compress/decompress in Oracle environment.
In TechEd ’11, Tony(SAP Mentor from Australia), mentioned his ECC Oracle DB size went from 9TB to 4TB (one could expect more compression in BW DB); and the day after compression, he stated, help desk started receiving calls asking if Basis team deleted any data over the weekend because the performance was much better:)).
I don’t believe DB2 has an edge over Oracle on compression. However I heard from another customer last Friday, DB2 requires less DBA resources to manage.
can you share something about DB2’s row compression? seems like an ideal fit for columnar compression done in HANA, especially for those large ECC tables.
I am not a big expert on DB2, and will ask my colleagues to see if they can add some color.
But here is my understanding – DB2 does significant compression of large tables. But we need to be careful on what is compressed. It takes more CPU to compress and uncompress – so if you are constantly changing data and then reading back, such tables might not be good candidates for compression. Hana compression will be great for several ECC tables – since several columns have a small list of possible values, and several columns are left blank in each row. However – for data already compressed by DB2, I doubt if HANA can compress it a lot further.
Very good summary Vijay. To your point on ECC on HANA, I partly agree with your comment. In my opinion, the real impact is not in the as-is scenarios but what can be done differently. For e.g, if I am a utility customer and want to analyze my unbilled revenue, I need to generate pseudo bills first and that complex logic is built in ECC routines. Having a side car HANA DB doesn’t help me today because it is not efficient for me to rewrite that business logic in the seperate HANA DB. Cases like this is where ECC on HANA will trigger excitement. It is the new doors that it will open which makes it interesting, thoughts?