Last night, I hopped a train to NYC and went to the NY Entrepreneurs Business Network Networking Event on behalf of my startup. There were good and bad things about this particular meetup. More after the jump…
[Read the rest of this entry...]
As many of you know, I’m not a fan of technical certifications — it seems like they reward people who can regurgitate technical trivia rather than the ones who can actually do the work (but may not be able to recall the canonical port number for LDAP traffic, as a quick example from both a past interview and a past exam). Having said that, last fall Oracle offered to select partners the opportunity to beta test some new certification exams. I signed up for about five and have gotten results back for most of them. So far, I’m three out of four (want to know which ones? Look here).
Here’s the thing: I signed up for two exams where I knew absolutely nothing about the product outside of what I have read on marketing slicks (Exadata and GoldenGate). I have since had the opportunity to do a GoldenGate deployment; still starting at zero with Exadata. I almost passed the GoldenGate one and did pass the Exadata one. So what does that say about the certification exam that I am now certified on a product I’ve never used?
– Update —
And now I’m four for five (although database security is a bit of a gimme for me…)
– Update 2 –
I reached out to Oracle about my Exadata exam result, and this is their response:
The problem arises from the cost and limited availability of Exadata Database Machines. Although most implementation exams require hands-on experience, we could not expect partners to purchase a Database Machine to prepare for the exam, and we don’t even have very many of them for courses and practice in house. This situation was amplified by the fact that many of the key features of Exadata just work – they are not configurable.
Because of this, the decision was made to make the “Implementation” exam for Exadata a detailed technical exam. Someone who understands how the Exadata technology works will be in a good situation to properly use the machine, and this corresponds with the course and test material.
Over the weekend, LinkedIn migrated their data center from Chicago to LA. To do this, LinkedIn offlined their service. In short, this was the wrong way to do it.
LinkedIn initially scheduled the outage to be five hours but it turned out to be much more (the official LinkedIn twitter account asks people to hold on just a bit more at the six hour mark and the official blog post about the migration has a comment from someone claiming the site is still down at the 13 hour mark. In today’s world of highly available computing, there is no reason why this had to happen.
From what I can tell, LinkedIn did a complete exchange of data centers; they apparently shut down the world to test the synchronization between the two sites. This just has me scratching my head. LinkedIn is a rather large enterprise, and I’m sure they have a few bucks they could set aside for a redundant testing environment which would eliminate the need for a complete outage.
Off the top of my head, there are four others ways I would have solved this problem. Check them out after the jump.
Today, I flew up to Boston and spoke at the Gilbane Boston conference. It was a long day, full of way more planes and taxis than I would like in eight hours and I learned several things today; some good, some not so good, but lessons none the less.
I’ve never been to Gilbane before. It’s my understanding that they are one of the larger conferences dedicated to content management. Perhaps; I have no frame of reference on the subject. However, as I was talking with a number of attendees, they universally talked about how fewer people were there this year and the ones that were tended to be from the Boston area.
From my perspective, the talk at which I presented was very sparsely attended. As we got started, the ratio of presenters/moderator to audience was 1:1 (this changed about halfway during the session when another person came into the room). On the one hand, that’s actually a nice challenge and an opportunity to learn what does or doesn’t work. One of the things I’ve learned from playing music is that it’s much easier to play for 4,000 people than 4 — if 2 people leave in the first case who cares; if 2 people leave in the latter case, that’s half your audience. So you have a very good chance to make connections with your audience and get immediate feedback on what does or does not work.
Content management is an important area of technology; it’s just not my area of technology. I’m a big data guy; this conference was all about enterprise systems of managing content from a high level, end user focused point of view. As opposed to optimal ways to store/index/search/process the content.
I spoke on federated queries. The way I have implemented federated query systems and the way the content people were talking about implementing them were not the same. It’s a good thing that I spent most of my time talking about the political issues surrounding a federated query system rather than specific technical implementation.
The people I met at the conference — and the two gentleman who were nice enough to share the stage with me — were quite interesting and knowledgeable. If for no other reason, the trip was worthwhile for that alone.
The 9am flight to Boston had 7 people on it. Seven. After the attendants gave us the nod, we drew straws and six of us moved to first class. The 5pm trip back was a bit better, in that there were about 20 people on the plane. Perhaps Delta might want to cut back on flights and, you know, lower prices a bit rather than burn all the fuel for seemingly no good reason?
Still, I don’t regret going. The more speaking I do, the better at it I get.
As this is being published, I am probably on stage at Gilbane Boston, participating in the Applications Of Search In Sharepoint panel discussion. Which is interesting, as I am not a SharePoint person. I do know data, though, so I’m talking using federated queries to search both structured and unstructured data.
In a departure from my more typical technical deep dive, this chat is more about the politics of implementing a federated query system. I’ve deployed federated systems for a few federal clients and the politics of each process were far more challenging than the technical issues. Tackling semantic search is a breeze compared to trying to get a forty year holdover from the LBJ days to loosen his grip on his precious, precious data.
But I digress.
The deck is available on Slideshare. I believe Gilbane will be publishing a video of the panel; I will update this post when that occurs. Thanks again to the Gilbane Group for giving us the opportunity to speak!
Daniel Abadi of Yale (and, for me, of Hadoop World fame) has a great article about a Teradata white paper focusing on Oracle’s Exadata:
I recently came across a whitepaper from Teradata, written by a senior consultant for Teradata, Richard Burns. This is a very well written piece, and has one of the best overviews of Exadata I’ve seen. I did not notice any obvious inaccuracies in the description of Exadata itself, and even the anti-Exadata arguments (presented after the overview), though at times biased and misleading, do not have many clear factual errors. Hence, it is quite a professionally done whitepaper, even though it is devoted to attacking a competitor. Reading it will probably make you smarter.
Both the original whitepaper and Daniel’s commentary are worth your time to read.
Mint.com has an article about ways to getting out of the canned customer service script.
One shortcut to a supervisor is to ask a question at the beginning of the call, says C. Scyphers, a former customer service representative who currently works for Daemon Consulting: “Ask if you’re allowed to record the call.” You don’t necessarily need permission to tape, since some states require only one side to consent to record. “Most first-tier support person are going to hear that question, quickly decide the caller is going to be too much hassle and then transfer to a more senior person just to get them out of their hair.”
Way back when, I got my start as a technical support rep. The above thing only happened to me once or twice, and it was always an immediate hand off to the tier two support team. It usually works for me when I try it.
This time, it’s about two occasions in which I hired the person who waited on my table for the companies where I was working at the time. Both people went into various forms of customer service (one in tech support, the other in the more traditional customer support); in both cases, it worked out very well. Good customer service is good customer service, regardless of the industry, and I have found it usually translates well.
“Good customer service is good customer service,” Scyphers says, reflecting on his recent restaurant-inspired hires. “[The waiters] were very charming, displayed excellent attention to detail and were well-informed about the world in general.”
I will be one of the speakers for Gilbane Boston this year. My particular contribution will be “Pitfalls & Profits In Federated Search” with a special focus on Sharepoint. More details after the jump.
[Read the rest of this entry...]
Now that both Hadoop World and Open World are done, let’s compare and contrast:
|
|
|
|
| Cost To Go | $300 | $1800 and that’s early bird |
| Location | NYC, NY | San Francisco, CA |
| Number Of Attendees | 900 | 41,000 |
| % of women | 5%. At most | 30-40% |
| Laptop Use During Talks |
33% | 10% |
| Typical Dress | Jeans & Company / Logo Polo Shirt |
Business Casual/Suits & No Ties |
| General content | Either very technical or sales/marketing | What’s this technical thing of which you speak? |
| Vendors In Expo Hall / Number Of Halls | 12-14 / 1 | ~600 / 5 |
| Number of concurrent sessions | 4 | 26 |
| Sessions Filmed & Available Online | All | Keynotes Only |
In the spirit of the NoSQL/OpenWorld comparison of last year.
Today was Hadoop World 2010 in NYC. Put together by the good folk at Cloudera, this conference was a long day focusing on big data, analytics and the Hadoop ecosystem.
All in all, this was a good conference. Hadoop and I were decent friends before today, with me being pretty good on how it could be used, where to use it and (generally) how to implement it. I’m now on a much stronger footing on all items, not to mention learning about some of the new tools. Well, new to me, at least.
Recap after the jump.
[Read the rest of this entry...]
That’s the title of a great article in TechCrunch about Oracle, HP & Andreessen (yes, him). The main thrust of the article is Oracle does better when it has a serious rival on to which it can focus it’s competitive attentions.
Only — pity for reporters– Ellison seems to have tired of the spotlight. Few magazine covers, no billion-dollar hostile takeovers, no wacky press conferences, no weird Shakepearian dramas with former proteges. Just a well-performing stock and growing business. Yawn.
Part of that is because Oracle either bought or trounced arch-rivals with its buy-not-build application strategy, and like most companies, Ellison becomes more interesting when he has a foil. It looks like that new foil is HP. Hallelujah, enterprise software is getting interesting again.
Because without that foil, we end up getting such brilliance from San Mateo as Beehive and the network computer.
Some of the more “established” players in the cloud marketspace take a potshot at Oracle’s Exalogic over on ReadWriteWeb today (I say “established,” because the market is far too young to have any truly established players as of yet). Essentially, their gripe is the “cloud in a box” marketing metaphor Oracle is using to push the hardware. Specifically, John Considine says
The Exalogic server is a contained set of resources that is purchased, operated, and maintained as part of the enterprise infrastructure. You can scale your applications up and down within this solution, but in the end, you are limited to the number of cores, amount or RAM, and size of the storage you purchased. While you can add more racks to the solution, you are stuck paying for the whole thing independent from what you really use – not exactly elastic or pay for only what you use.
To which I say, “Um, yes. And?” Oracle isn’t marketing Exalogic to startups who want to host their environment on the cloud; Oracle seems to be marketing Exalogic to large companies who want to provide their own internal cloud to host their own applications. Once you look at Exalogic from that perspective, then the criticism stops being very persuasive. Any cloud platform is going to be limited to the numbers of cores/RAM/storage available; within any cloud platform, you can scale up and down only within your available resources; if you need to add more capacity, you have to pay for the entire gamut of capacity you want to add, not what you actually use. This becomes a somewhat standard capacity provisioning problem at this point, one of which can be simplified by the unified forklift install of Exalogic.
Don’t get me wrong; I see lots of things wrong with Exalogic. Not the least of which is my view that Oracle is fighting off the coming disruption (of an inventor’s dilemma type) of a commodity everything cloud (hardware/software/infrastructure/VMs/etc where “good enough” gets by cheaper) by trying to sell the performance side of their engineering genius. But John’s comment isn’t one of them.
Oracle releases their recap of the last five years of Open World highlights. Strangely, they left Beehive off. Sorry guys, I was there; you were pushing it pretty hard. That it didn’t work out doesn’t excuse airbrushing away that little speck of unfortunately history…
Sorry for the delay in this post; I was a big jetlagged this morning from a late flight into IAD. Anyhoo…
[Read the rest of this entry...]
Day four
And now, we’re in the home stretch….
[Read the rest of this entry...]
Pam Barker quotes me on virtualization and DR:
“One of the more common ways large hosting environments provide virtualized DR solutions is to use para-virtualization in order for the physical machines to support multiple virtual machines,” explains C. Scyphers.[...] “However,” Scyphers adds, “if the hosting physical machine is insufficiently sized, a DR situation may cause the VMs – which are now acting as primary servers – to overload the resources of the underlying physical server.”
I spent some time on Monday and Tuesday wandering through the exhibition floors of Oracle’s OpenWorld. From that, I can safely make two conclusions:
1) Oracle has largely sucked the oxygen out for most companies trying to create database bolt-ons.
2) Oracle’s applications (eBusiness, Hyperion, JD Edwards, etc.) have significant issues.
Let’s start with the last one first.
If you were to go on the exhibition floor of Oracle’s Open World this year, you might think that everyone was giving away an iPad. That’s because lots of them are: specifically, 55 vendors are giving away at least one iPad and 10 more are giving away various other Apple products (Nanoes, iTouches). I saw another 7 vendors giving away eBooks (mostly Kindles, but there was one Nook in there — perhaps they have a relative that works for Barnes & Noble). For those a bit more mathematically inclined, that’s 19% of people giving away iPads, 23% giving away Apple products and 25% giving away electrical toys that can support reading eBooks. However, these stats do not take into account the several vendors (EMC, for example) who were giving away an iPad a day, if not an iPad an hour.
An iPad fits well into the bribery model of an exhibition floor; cheap enough to be economical to give away, spiffy enough to attract attention and expensive enough that people would be willing to sign up for spam in order to maybe win one.
As a side note, one company is giving away a Tesla Roadster (a $100k car, only about 1,000 have been made), but they are doing it in such a way as to practically ensure no one will ever win it. A different company, however, is giving away a Lexus RX-440h every day of the conference. I guess the recession is over, eh?
Book a meeting with me
