Vertica projections — an overview
Partially at my suggestion, Vertica has blogged a three–part series explaining the “projections” that are central to a Vertica database. This is important, because in Vertica projections play the roles that in many analytic DBMS might be filled by base tables, indexes, AND materialized views. Highlights include:
- A Vertica projection can contain:
- All the columns in a table.
- Some of the columns in a table.
- A prejoin among tables.
- Vertica projections are updated and maintained just as base tables are. (I.e., there’s no kind of batch lag.)
- You can import the same logical schema you use elsewhere. Vertica puts no constraints on your logical schema. Note: Vertica has been claiming good support for all logical schemas since Vertica 4.0 came out in early 2010.
- Vertica (the product) will automatically generate a physical schema for you — i.e. a set of projections — that Vertica (the company) thinks will do a great job for you. Note: That also dates back to Vertica 4.0.
- Vertica claims that queries are very fast even when you haven’t created projections explicitly for them. Note: While the extent to which this is true may be a matter of dispute, competitors clearly overreach when they make assertions like “every major Vertica query needs a projection prebuilt for it.”
- On the other hand, it is advisable to build projections (automatically or manually) that optimize performance of certain parts of your query load.
The blog posts contain a lot more than that, of course, both rah-rah and technical detail, including reminders of other Vertica advantages (compression, no logging, etc.). If you’re interested in analytic DBMS, they’re worth a look.
Derived data, progressive enhancement, and schema evolution
The emphasis I’m putting on derived data is leading to a variety of questions, especially about how to tease apart several related concepts:
- Derived data.
- Many-step processes to produce derived data.
- Schema evolution.
- Temporary data constructs.
So let’s dive in. Read more
| Categories: Data models and architecture, Data warehousing, Derived data, MarkLogic, Text | Leave a Comment |
Data management at Zynga and LinkedIn
Mike Driscoll and his Metamarkets colleagues organized a bit of a bash Thursday night. Among the many folks I chatted with were Ken Rudin of Zynga, Sam Shah of LinkedIn, and D. J. Patil, late of LinkedIn. I now know more about analytic data management at Zynga and LinkedIn, plus some bonus stuff on LinkedIn’s People You May Know application. 🙂
It’s blindingly obvious that Zynga is one of Vertica’s petabyte-scale customers, given that Zynga sends 5 TB/day of data into Vertica, and keeps that data for about a year. (Zynga may retain even more data going forward; in particular, Zynga regrets ever having thrown out the first month of data for any game it’s tried to launch.) This is game actions, for the most part, rather than log files; true logs generally go into Splunk.
I don’t know whether the missing data is completely thrown away, or just stashed on inaccessible tapes somewhere.
I found two aspects of the Zynga story particularly interesting. First, those 5 TB/day are going straight into Vertica (from, I presume, memcached/Membase/Couchbase), as Zynga decided that sending the data to some kind of log first was more trouble than it’s worth. Second, there’s Zynga’s approach to analytic database design. Highlights of that include: Read more
| Categories: Aster Data, Couchbase, Data models and architecture, Games and virtual worlds, Greenplum, Hadoop, Petabyte-scale data management, Specific users, Vertica Systems, Zynga | 27 Comments |
Virtual data marts in Sybase IQ
I made a few remarks about Sybase IQ 15.3 when it became generally available in July. Now that I’ve had a current briefing, I’ll make a few more.
The key enhancement in Sybase IQ 15.3 is distributed query — what others might call parallel query — aka PlexQ. A Sybase IQ query can now be distributed among many nodes, all talking to the same SAN (Storage-Area Network). Any Sybase IQ node can take the responsibility of being the “leader” for that particular query.
In itself, this isn’t that impressive; all the same things could have been said about pre-Exadata Oracle.* But PlexQ goes somewhat further than just removing a bottleneck from Sybase IQ. Notably, Sybase has rolled out a virtual data mart capability. Highlights of the Sybase IQ virtual data mart story include: Read more
| Categories: Columnar database management, Data warehousing, Oracle, Parallelization, Sybase, Theory and architecture, Workload management | 1 Comment |
Renaming CEP … or not
One of the less popular category names I deal with is “Complex Event Processing (CEP)”. The word “complex” looks weird, and many are unsure about the “event processing” part as well. CEP does have one virtue as a name, however — it’s concise.
The other main alternative is to base the name on “stream processing” instead.* The CEP-or-whatever industry is split between these choices, with StreamBase currently favoring “CEP” (despite its company name), IBM emphatically favoring “stream”, and Sybase seemingly trying to have things both ways.
*And then, of course, there is “event stream processing”, regarding which please see below.
Hadoop evolution
I wanted to learn more about Hadoop and its futures, so I talked Friday with Arun Murthy of Hortonworks.* Most of what we talked about was:
- NameNode evolution, and the related issue of file-count limitations.
- JobTracker evolution.
Arun previously addressed these issues and more in a June slide deck.
Read more
| Categories: Hadoop, MapReduce, Parallelization, Workload management, Yahoo | 7 Comments |
HP/Autonomy sound bites
HP has announced that:
- HP is buying Autonomy.
- HP is pulling back from WebOS.
- HP may spin off its PC business altogether.
On a high level, this means:
- HP is doubling down on enterprise IT.
- HP is taking a more software-centric approach to the enterprise IT business.
- HP is backing away from the consumer electronics business.
- HP in particular is backing away from the generic desktop/laptop PC business, which may with only moderate exaggeration be regarded as:
- The intersection of the enterprise IT and consumer electronics businesses.
- The least attractive sector of each.
My coverage of Autonomy isn’t exactly current, but I don’t know of anything that contradicts long-time competitor* Dave Kellogg’s skeptical view of Autonomy. Autonomy is a collection of businesses involved in the management, search, and retrieval of poly-structured data, in some cases with strong market share, but even so not necessarily with the strongest of reputations for technology or technology momentum. Autonomy started from a text search engine and a Bayesian search algorithm on top of that, which did a decent job for many customers. But if there’s been much in the way of impressive enhancement over the past 8-10 years, I’ve missed the news.
*Dave, of course, was CEO of MarkLogic.
Questions obviously arise about how the Autonomy acquisition relates to other HP businesses. My early thoughts include: Read more
| Categories: HP and Neoview, Market share and customer counts, Structured documents, Text, Vertica Systems | 10 Comments |
Couchbase technical update
My Couchbase business update with Bob Wiederhold was very interesting, but it didn’t answer much about the actual Couchbase product. For that, I talked with Dustin Sallings. We jumped around a lot, and some important parts of the Couchbase product haven’t had their designs locked down yet anyway. But here’s at least a partial explanation of what’s up.
memcached is a way to cache data in RAM across a cluster of servers and have it all look logically like a single memory pool, extremely popular among large internet companies. The Membase product — which is what Couchbase has been selling this year — adds persistence to memcached, an obvious improvement on requiring application developers to write both to memcached and to non-transparently-sharded MySQL. The main technical points in adding persistence seem to have been:
- A persistent backing store (duh), namely SQLite.
- A change to the hashing algorithm, to avoid losing data when the cluster configuration is changed.
Couchbase is essentially Membase improved by integrating CouchDB into it, with the main changes being:
- Changing the backing store to CouchDB (duh). This will be in the first Couchbase release.
- Adding cross data center replication on CouchDB’s consistency model. This will not, I believe, be in the first Couchbase release.
- Offering CouchDB’s programming and query interfaces as an option. So far as I can tell, this will be implemented straightforwardly in the first Couchbase release, with elegance planned for later down the road.
Let’s drill down a bit into Membase/Couchbase clustering and consistency. Read more
| Categories: Cache, Clustering, Couchbase, memcached, Memory-centric data management, MySQL, Parallelization, Solid-state memory | 8 Comments |
Couchbase business update
I decided I needed some Couchbase drilldown, on business and technology alike, so I had solid chats with both CEO Bob Wiederhold and Chief Architect Dustin Sallings. Pretty much everything I wrote at the time Membase and CouchOne merged to form Couchbase (the company) still holds up. But I have more detail now. 😉
Context for any comments on customer traction includes:
- Membase went into limited production release in October, and full release in January. Similar things are true of CouchDB.
- Hence, most sales of Couchbase’s products have been made over the past 6 months.
- Couchbase (the merged product) is at this point only in a pre-production developer’s release.
- Couchbase has both a direct sales force and a classic open-source “funnel”-based online selling model. Naturally, Couchbase’s understanding of what its customers are doing is more solid with respect to the direct sales base.
- Most of Couchbase’s revenue to date seems to have come from a limited number of big-ticket “lighthouse” accounts (as opposed to, say, the larger number of smaller deals that come in through the online funnel).
That said,
- Most Membase purchases are for new applications, as opposed to memcached migrations. However, customers are the kinds of companies that probably also are using memcached elsewhere.
- Most other Membase purchases are replacements for the Membase/MySQL combination. Bob says those are easy sales with short sales cycles.
- Pure memcached support is a small but non-zero business for Couchbase, and a fine source of upsell opportunities.
- In the pipeline but not so much yet in the customer base are SaaS vendors and the like who use and may want to replace traditional DBMS such as Oracle. Other than among those, Couchbase doesn’t compete much yet with Oracle et al.
- Pure CouchDB isn’t all that much of a business, at least relative to community size, as CouchDB is a single-server product commonly used by people who are content not to pay for support.
Membase sales are concentrated in five kinds of internet-centric companies, which in declining order are: Read more
The Ted Codd guarantee
I write a lot about whether or not to use relational DBMS. For example:
- In May I surveyed relational vs. non-relational pros and cons at some length.
- Last November I mused about when it might be OK to do without joins.
- The question is implicit in a variety of posts about, say, document-oriented or object-oriented DBMS.
Before going further in that vein, I’d like to do a quick review of what E. F. “Ted” Codd was getting at with the relational model in the first place. Read more
| Categories: Data models and architecture, IBM and DB2, MOLAP, NoSQL | 3 Comments |