A Deeper Look At Sybase: History of ASE covered some of the history of Sybase Adaptive Server Enterprise (ASE). Today’s post is the next part in theA Deeper Look At Sybase, and is about where ASE is today.
A Deeper Look At Sybase: History of ASESybase established itself as Wall St’s financial database of choice in the 1990’s. Since ASE has consistently proven its reliability, robustness and performance over the years, a large part of Wall St systems still run on Sybase ASE today.
Indeed, Sybase is still going strong in the world of finance. This was illustrated by Sybase showing record years of revenue and profit in 2008 and 2009, despite the credit crisis. One reason was that, once the crisis had broken out, the financial sector invested heavily in additional systems for risk management (the lack, or shortage, of which, as we all know now, was an important factor in how the credit crisis came about). Such systems require exactly the type of infrastructural data management software provided by Sybase, and consequently, Sybase’s business did well.
Another illustration of how deeply entrenched Sybase ASE is in the financial market segment, is that various 3rd-party vendors of financial software packages still offer their applications on ASE only.
For clarity, ASE’s use is not restricted to the financial segment, but is used in anything from astronomical satellite’s observation archives to logistics to medical systems. Just to give a non-financial example: the nationwide railway ticketing systems in China as well as in India both run on Sybase ASE. Just think, for a second, about the number of people taking the train every day in those countries, and what that means for the system’s availability and its ability to deliver predictable performance. Would you take the job if you were asked to design such a system tomorrow? ASE has proven to be capable of meeting such challenging requirements.
I’ve mentioned reliability as one of ASE’s strengths. Also worth mentioning is the virtual absence of security breaches in Sybase ASE over the years. While Oracle and Microsoft SQL Server have had significant publicity around security holes, this has hardly happened for Sybase ASE.
It has been argued that, as a smaller but specialised database vendor, Sybase systems might not be as attractive a target for hackers as for example, Oracle-based systems. However, that argument makes no sense: when you recall that many Sybase customers are financial institutions like banks, that would only make Sybase ASE an even more attractive target to break into. So the conclusion must be that, somehow, Sybase ASE has been designed and engineered such that there is little opportunity for security holes. Your data is safe in ASE!
Currently, the latest and greatest ASE release is version 15.x; the previous major release is version 12.5.x. You may wonder what happened to versions 13 and 14? Well, some superstitious folks felt that ’13’ should be avoided for its association with bad luck (a stupid reasoning if you ask me, since I was born on Friday the 13th myself — I’ll take 13 as my lucky number any day!).
As for ’14’, the story is that this was skipped because that number seems to have a somewhat ominous connotation in Chinese, much like ’13’ in the Western world. Given the strength of Sybase in the Chinese market, you can imagine why ’14’ was avoided. Apparently nobody found anything wrong with ’15’, so that became the next ASE version number.
(Okay, here’s another bit of history when it comes to version numbers: in the mid-1990’s, Sybase SQL Server jumped from version 4.9 to version 10. According to anecdote, some large customer had to decide between Oracle and Sybase. At that point Oracle was at version 6, whereas Sybase was at version 4.9 and heading towards version 5. The customer reportedly said “well, if Oracle is already at ‘6’ and Sybase isn’t even at ‘5’ yet, then Oracle must be more advanced. I’ll buy Oracle”. Not to be the victim of such well-informed decision making again, the next Sybase SQL Server version was then moved to ’10’).
Speaking of version numbering: what you may notice today is that unrelated Sybase products tend to have similar major version numbers. Sybase Replication Server, Sybase IQ and PowerDesigner all went from version 12.x to 15.x as well, following the ASE version. There is no deeper significance to this sharing of a major version number, other than to signify a common notion of ‘latest version’.
Due to its architecture, Sybase ASE happens to run especially efficiently on Linux. There is no better illustration I can give than a partnership between Sybase and IBM (this is still active, to the best of my knowledge) where IBM prefers to sell its P-series Linux servers with Sybase ASE rather than with IBM’s own DB2 or Informix databases. The reason is simple — ASE squeezes more performance out of the same hardware. IBM’s software division must not have been happy about this, but given that the IBM hardware folks got their way (they sell the P-series boxes), I guess that must mean the revenue numbers were convincing.
Sybase ASE runs on the main flavours of Unix, Linux, and Windows. That’s not different from most databases (although Sybase is still supporting HP-UX/Itanium, unlike certain other vendor’s plans). Indeed, today’s enterprise databases are broadly similar in terms of the general functionality they provide – after all, SAP Business Suite runs on various different database brands.
Nevertheless, one of the things that clearly sets ASE apart is its ease of administration: an ASE-based system requires significantly less DBA involvement than a comparable system running on Oracle or DB2. Many large Sybase customers also run Oracle in their enterprise, and we’ve consistently heard such customers report that less DBA effort is required to keep a Sybase database running compared with an Oracle database. Estimates vary, but based on what I’ve seen and heard over the years, I’d put that DBA efficiency difference at a factor 3 at least (to ASE’s advantage).
DBA efficiency may sound boring, but in reality such staffing costs are a sizeable chunk of an IT system’s operational cost.
I recall a customer success story at sybase.com about a large Wall St customer who were operating some 30.000 ASE servers and 1.000 Replication Servers worldwide, with less then 100 DBAs. Such numbers are impressive by any standard. (this was prior to 2005, this document appears to have been archived from sybase.com).
ASE comes in a number of flavours. First, there is ASE-SMP vs. ASE Cluster Edition. ASE-SMP (or ‘ASE classic’ as I like to call it) is the regular form of the ASE database server that most customers have, where ASE runs on a single host. In contrast, ASE Cluster Edition, which was first released a few years ago, is a shared-disk cluster architecture designed to run on a clustered server: an ASE server runs on each cluster node, but together they act as if there is only one ASE server: for an application, it doesn’t matter whether it runs on ASE Cluster Edition or ASE-SMP, since they behave identical functionally. What ASE Cluster Edition adds is better high-availability and workload management capabilities such as load balancing.
ASE Cluster Edition is similar in concept to Oracle RAC, but endlessly easier to configure and administrate.
On a different dimension, ASE-SMP is available as Enterprise Edition (for the largest deployments) and a Small Business Edition (can run on no more than 8 CPUs). Since Sybase ASE is a proprietary, commercial software product, these editions require you to buy a license from Sybase.
However, there are also two free versions: the Developer Edition (download here; allows only a limited number of users; not allowed for production use; most optional features included) and the Express Edition (download here; free for production on Linux; limited to 1 CPU and 5GB of data; no optional features included). For ASE Cluster Edition, there’s no free version but only what amounts to an Enterprise Edition (though it’s named, you guessed it, ‘Cluster Edition’).
Anyone for Sushi?
The technical quality of ASE has always been, and still is, recognised throughout the IT industry.
For a number of years, Sybase was perhaps less successful in effective marketing of its products outside the traditional Sybase market segments — Sybase’s corporate DNA has always been more about engineering. A former Sybase marketing manager once admitted in an interview that Sybase had suffered from the ‘cold, dead fish syndrome’(*) in earlier years. Fortunately, things have improved much since. With SAP having embraced Sybase and its products, I have good hopes that Sybase ASE may conquer a bigger piece of the database market.
(*) Never heard of the ‘cold, dead fish syndrome’? It describes a tendency of using too-technological, non-appealing names for products, something that strongly technology-driven companies are sometimes guilty of. The syndrome derives its name from sushi: “If Company X were selling sushi, they would market it as ‘cold, dead fish’. ”
Coming up next: A Deeper Look At Sybase: Sybase IQ