Recent Posts by Peter Jeffcock

Big Data SQL has been out for nearly two years and version 3.0 is a major update. In addition to increasing performance (next post) we've added support for clusters that aren't built on engineered systems. And alongside several other Oracle big data software products, Big Data SQL now also supports Hortonworks Data Platform. Before 3.0, the requirements were simple. You could use Big Data SQL to deliver "one fast, secure SQL query on all your data" as long as you were using...

I enjoy a trip to the hardware store as much as the next person. I like the feeling of achievement after I've built, painted or repaired something. And I know I've saved money when I compare the cost of the parts I bought, with the cost of paying a pro to do it all for me. Of course, I don't account for my own time or how much extra time it takes to complete the job (or even the risk that I'll have to pay a pro more to fix a mistake). But if you're working in IT, those costs...

We did a series of 10 predictions for 2016 around big data. The fourth one was around data virtualization, a key component of which is Oracle Big Data SQL which we’ve blogged about before. But the prediction was a little broader: "Look for a shifting focus from using any single technology, such as NoSQL, Hadoop, relational, spatial or graph…”. You know about NoSQL, relational and Hadoop. You’ve used a map, so can figure out spatial (though there’s much more to it than just a...

Oracle Big Data SQL has been out for a while and we announced Oracle Big Data SQL Cloud Service back in June. So what's different when it moves to the cloud? The pricing is different and it's running in our datacenter, not yours. But technically it offers the same capability. Which means these entries from earlier in the year still apply. So if you want to see how Big Data SQL Cloud Service handles metadata, has the best SQL on Hadoop, extends Oracle Database security to...

Guest Blog Author: Eric Sammer, CTO and Founder at Rocana I’m very excited to announce our partnership with Oracle. We’ve been spending months optimizing and certifying Rocana Ops for the Oracle Big Data Appliance (BDA) and will be releasing certified support for Oracle’s Big Data software. Ever since we worked with the Oracle Big Data team in the early days of the Big Data Appliance Engineered System while still at Cloudera, we’ve had a vision of the power of a pre-packaged...

We recently shipped a new big data product: Oracle Big Data Spatial and Graph. We’ve had spatial and graph analytics as an option for Oracle Database for over a decade. Now we’ve taken that expertise and used it to bring Spatial and Graph analytics to Hadoop and NoSQL. But first, what are spatial and graph analytics? I'll just give a quick summary here. Spatial analytics involves analysis that uses location. For example, Big Data Spatial and Graph can look at datasets that...

I posted here and here about showing how Oracle Big Data Appliance (BDA) is much cheaper than building your own DIY cluster. I know for some it's very counter-intuitive, but the numbers back this up. And it turns out it's not just me saying so. We first worked with ESG two or more years ago. They compared the BDA with a DIY cluster and we updated the paper last summer. This version shows the BDA 21% cheaper than a comparable cluster. (Note that we've since significantly...

Somebody looked at this earlier post and said "you're not comparing apples with apples". That's because the Big Data Appliance (BDA) is based upon a high specification 2 RU server, rather than the smaller and simpler 1 RU servers that I used in my comparison. Fair enough. Although the reason I used that config is because that's what almost everybody is thinking of when they say "I can build a Hadoop cluster cheaper than the BDA". And I can do that comparison quickly on a...

I've spent a lot of time on trade show stands. And I've lost count of the number of times that somebody has asked about the Oracle Big Data Appliance (BDA) and then said "I (or my organization) can build that cheaper myself". There are a couple of approaches that usually come up. Let's look at the first one. The most common approach is to find a cheap 1U, commodity server and build out a rack of those. I looked at the Dell site and you could go with either the R320 or the...