ASF 013: James Rowland-Jones interview

ASF 013: James Rowland-Jones interview


James Rowland-Jones is a Principal Program Manager for Microsoft. He is passionate about delivering highly scalable solutions that are creative, simple and elegant in their design. James specializes in big data warehouse solutions such as Microsoft Analytics Platform System (APS), Azure SQL Data Warehouse, Hadoop ecosystems.

James is a keen advocate for the SQL Server community; both internationally and in the UK. He has previously served on the Board of Directors for PASS and helped organise SQLBits. James was awarded Microsoft’s MVP accreditation from 2008 – 2015 for his services to the community.

This talk has taken place during SQL Bits conference in London, on 23rd February 2018 (Friday).
Interviewers: Kamil Nowinski & Damian Widera (remotely).

Do you want to know why James has moved to Seattle and started working for Microsoft?
How he started his career in IT and what is important in your work? Which big conference is his baby and how much his community involvement has affected his professional life?
You will find out all the answers in the following talk.


* Bear in mind that term “compute optimized” for Tier 2 is no longer in use.

Kamil Nowinski: Our today’s guest is a Principal Program Manager of Microsoft since December 2015, former Director at Large in PASS for almost four years, former Organizing Committee Member for seven years and currently part of data platform teams supporting Massively Parallel Processing workloads in hybrid architectures. When you are working with Azure SQL Data Warehouse, you’ve heard about him very lightly. So our today’s guest is…

JRJ: James Rowland Jones.

KN: James, where do you live currently?

JRJ: I currently live in Sammamish which is just outside of Redmond and just past Lake Sammamish in Washington state.

KN: It is very close to Seattle, right? To Redmond basically?

JRJ: Yes, it is very close to Redmond.

KN: However, you used to live in the UK. Why do you decide to move to the USA?

JRJ: I’ve been working in consulting for many years – works in a partner eco-system for a long time. I realized that my impact was ultimately one customer at a time and I could do my best for that one customer. But then the opportunity came up for me to be able to transfer over and actually do whatever I could to help everybody all at once, so I can release one feature and help everybody. The opportunity to be a part of the team and making the product and be a part of the journey that we’re going on with Azure SQL Data Warehouse was too good to pass up. My family and I had some very long conversations. We decided that we want to do it. And that was it. We moved over just a year ago. Here at SQLBits this is my first time back. It is great to be here and see the SQLBits carrying on. I’m going from Strength to Strength and I can’t begin to tell you how proud I am of this event in terms of what it means to the global Community of data professionals. Even at Microsoft I can see an echo. Everybody cares very much about the Community and the success that we were able to achieve with SQLBits. It’s great to see it continuing and long may it do so.

KN: You have been with SQLBits from the beginning?

JRJ: Pretty much. I first came over to SQLBits as an attendee. I attended the first one. And at the second SQLBits I have spoken for the first time I have ever spoken about anything. I did a talk on I/O. And then the third one I started to get involved and at the fourth one I started to be actively involved in the Committee of SQLBits. And then pretty much what happened to the point where I left and joined Microsoft, I was working very closely with Simon, Darren, Aaron and Chris – all of the organizers over the years to make these events happen. This is one of the things I probably would say I am most proud of in all of my professional carrier. It has obviously in many ways nothing to do with my professional carrier because it is all volunteer. Like yourself, we would kind make it happen. It is a great thing. Something that was truly special.

KN: It is good to see how your child is growing.

JRJ: Absolutely.

Damian Widera: James, what are you doing for a living?

JRJ: I’m a Principal Program Manager at Microsoft now and I work on Azure SQL Data Warehouse. Up until very recently, I was actually working directly on product focus. And the current preview of the compute optimized* of the performance tier was the area that I was leading. And I was delighted to be able to get that into a preview. Just recently, like a week or so ago, I have slightly different role within the team. Now I lead a team of PMs and we look at the end to end ecosystem around Data Warehouse in analytics in general. So thinking about what SQLDW role is an analytics platform but also thinking about what are they associated, related services, how do they integrate together, how do we make sure…

KN: So you have a little bit wider view currently…

JRJ: Yes, much broader perspective on how we actually make the magic happen at the Azure.

KN: All the integrations.

JRJ: Yes, some of it is a big integration, some of it is making sure that we have clear patterns and guidance – stuff like that.

KN: I’ve been on your session where you have been showing the new generation of Data Warehouse in Azure. Could you tell us a bit more about it? What new is in the new generation of Azure Data Warehouse?

JRJ: This compute optimized* performance tier which we have is like the next generation of SQLDW. It introduces new hardware from Azure. It is new NVMe’s SSD capability into the compute node itself. And with that, is means that we can add to our story around separation of computer storage this intelligent cache locally in the computer nodes which eliminates beneath remain I/O when querying large volumes of data.

Second thing – to make use of that cache is that we externalize all of the columns to segments and we hold them independent of the actual SQL server data files. So you can imagine how impossible this would be in an on-premise world. We’re taking that data and we’re holding it in the remote storage. And then – as customers query their column-store data, we move that column-store segment up into the intelligent cache and we manage it there so that queries – especially weak queries – run exponentially fast.

KN: And the cache is pretty amazing – 400 GB.

JRJ: The nodes themselves are quite large physically. Every node today is about 400 GB more of physical memory and then you basically come down to the cache itself and the NVMe SSD disk space cache has about 1.5 TB of cache in that node and that is actually what delivers the hottest data in the warehouse which you can imagine in Data Warehouse obviously like tens of the tail of the table, keeps that nice and tight to the CPUs use so that we can really deliver the best in class performance. It’s very exciting to see that.

KN: It’s 100 better improve performance, right?

JRJ: Yeah. My favourite individual customer experience that had one of our specific customers were able to get some of that queries to run kind of 100 times faster but also they were running with half the compute costs. They literally were running half the compute power but got this massive performance benefit on some of their queries. Which is really great to see. That is a great value for customers.

KN: Yes, it is. You mentioned also the data movement service.

JRJ: One of the things that should be in the system is that in order to get the resolve to query you have to be able to transparently move data across the nodes in order to make it works properly. So what we do to achieve that is we use the technologies like the data movement service. But what we’re doing right now is actually that we took some of the core data movements type – one in particular called the shuffle move – and we made it an in‑process database movement and because of that we’ve also seen phenomenal performance improvements because the data itself just moves that much faster around the system.

KN: Those things are all amazing and probably the customers will be asking when it will be generally available. [this talk has taken place on February 2018].

JRJ: We’re currently targeting for Spring of this year [2018] to bring it to the general availability. Where we are at right now customers can and are using the preview in about six regions of Azure and we’re aggressively onboarding new regions as we go. Azure SQL Data Warehouse currently is available in 33 regions in the world. I think it is over double that any other Cloud Data Warehouse available in the market today. It’s a great story for us and it’s a very important product.

KN: And you are as an employee and very important person from that perspective – from Data Warehouse in Azure – you are working with customers very closely. And from that perspective, how do you think how fast customers are adopting those new technologies in their environment?

JRJ: These are some very interesting patterns, actually. What we have seen in a lot of the time was that customers moving into Data Warehousing in the Cloud really wanted to take advantage of what the Cloud can offer them rather than just pick it up and just flip it into the Cloud. They wanted to see what they can really do to modernize the platform, they’re separating out their data, they’re trying to leverage also like a data lake driven approach in many cases. Sometimes they don’t even realize they’re doing it, the way as they transfer their data into BLOB Storage and then move the data through into the DW. So we’ve seen customers of all shapes and sizes across all the major verticals which is great to see. It’s not just kind of landed in one vertical. That’s always very important for us. And then the other thing that we’ve seen is that you get these new styles of workloads where customers are integrating basically like if you can think about transactional feed with your Cloud data sources – like dynamic CRM or Salesforce – whatever – integrating that into that flow and then also maybe even some sensor data IoT style. All those different types and flavours of data, all coming in and then using the DW to serve that out organizationally. It’s interesting because the more typical workloads and then you get the people to see the potential doing something different and then even going on and building out some new capabilities. It’s been great to see customers integrating ML [Machine Learning] modules over their data in the DW doing things like that to actually take advantage of all the connected services available in Azure.

KN: If they move, they already see new possibilities.

JRJ: Absolutely. That’s a very real thing for people.

KN: Also those new features like Graphical Execution Plan in the newest SSMS. I think it’s a very cool direction. But how much can it help right now?

JRJ: The important thing we found when people were using Explain plans in the old way is that it took quite a lot of explaining to explain how to use the Explain plan – that’s quite a tongue twister – but it meant ultimately for us that we were making our customers lives more difficult. We wanted to have something that we give them like the conceptual view and visibility of the plan in a way that people knew and understood and were happy with. It’s great to see the overwhelmingly positive feedback we’ve had when I’ve shown the use-case both in my Pre-Conf with the latest on Wednesday and also when we show it in the session today. That’s been fantastic.

KN: But the plan is not producing by the engine itself.

JRJ: Yes, it’s producing by the MPP engine.

KN: Ah, is it?

JRJ: Yes. So literally what you are seeing when you run the Graphical Explain Plan is exactly the approach that had you run the query that is what SQLDW would have used as a plan to actually execute your query. So, it’s a real-real.

KN: Has it been finished or would it be extended?

JRJ: I think we’re at the point when we’re happy with what we’ve got, we want to see what customers want and what feedback we get now. We’re more in the actively listening mode again for that particular feature.

DW: What do you think – what will be the future of the cloud?

JRJ: I think that the future of on-premises is more questionable than the future of the Cloud The future of the Cloud has a pretty clear message. I mean the vibe here in SQLBit since I’ve been here, obviously doing this for many years – the question is not like “Should I move to the Cloud?” or “When can I move to the Cloud?” but the question is: “How? How do I get there? What do I need to do?” All of those sorts of questions. I don’t see it as a question about the future of the Cloud. I think it’s more about how customers are moving their workloads to the Cloud, how quickly they can do that and what’s driving that. And then for the workloads – it’s stay on-premise and there are clearly legitimate reasons for that. It’s about who can support you best there and what customers solutions do we provide in that space. And that’s what it’s exciting about working in Microsoft – we support customers where they’re at with their requirements. And I feel that we have some very competitive offerings whether you’re on-premise or in the Cloud.

DW: Yes, that’s right. And the question could be in the future of course why am I still not in the Cloud? I should be there because that’s the future.

How do you start working with data?

JRJ: I actually have a law degree. I’m not a computer scientist in any stretch. I happened to been sat at home very unemployed one day. I was watching a TV show in the UK called Working Lunch. It was a lunchtime business show. And it was in about 1998. There was a chap on TV. They were interviewing him from LOGICA – now the CGI is the current version of this company – they were talking about an international phenomenon known as Y2K and they were talking about the shortage of technical resources to solve the problem of Y2K compliance and coding. CGI’s response for that at that time was a 4-week training course as an apprenticeship. And if you turned up at the university and pass the interview, they would put you on this 4-week training course and turn you onto the Y2K programmer. That’s how I got into it. I actually turned up. I could honestly tell you that I failed the math test but I managed to Ace the interview and they let me in. And the rest is history. Five weeks later I was having my first interview and then I spent my first two years working on Ford of Europe’s mainframes making Y2K changes to their vehicle accounting system.

KN: Wow! That’s amazing history.

DW: And now you’re at Microsoft so what is the most important part of your work?

JRJ: I think wherever you go the most important part of your work is the people who you are working with and actually making sure that everyone is aligned to the goals that you have and they are actually working together to achieve those goals. When you have that alignment and that energy and everybody is pulling together that’s a phenomenal feeling. It’s great to see things land – you win together, you lose together. That’s a very important lesson not just professionally but also in live.

KN: Working as a team.

Ok, can you tell us how are you preparing for a speech exactly?

JRJ: It’s really thinking about the audience. What do you think people need to know? What outcomes do you want them to have from having attended your session? One of the things about coming to this is like there is the wealth of choice so appreciating the people who choose to attend your session is very, very important. And once you have that, the next piece is really about making sure that you give them something that is somehow actionable. Sometimes knowledge for knowledge’s sake has very limited value in my opinion. Whereas if I give you something with the purpose that you can go and do or you can go and try, or you can explore and you take it forward, that tends to have much more meaningful impact overall. But just me telling you about my product, you’re just saying that it was really interesting and we all go away that doesn’t necessarily change anything for anybody. That’s why I try to think about showing things like we showed three things today in the session which haven’t been really shown publicly before but these are the three things that people can go and do. And they can go and do them immediately and they can improve their solutions when they go back to work. One of the most important things when SQLBits were founded was the experience which the speakers had: Was is the real-world experience? Did they were able to communicate the value and actually help people who were attending with their everyday problems?

KN: Show the practice from the real field.

JRJ: So I try and do stuff that’s more true to that spirit.

KN: How much important is passion for you?

JRJ: I think personally it’s very important. You have to care about what you’re doing. There’s lots of people that are very passionate in this space and it kind of shines through. When you’re speaking about something you have to really care about it and you would have to able to communicate the value of it and I think it helps make your sessions come to live.

KN: This podcast is not only about the technical stuff but also about the personality. How about your work-life balance?

JRJ: As an interesting one – I’m sure if you asked Mrs. Rowland Jones, she would give you a very different answer. I work pretty hard I have to say in that regard. There’s lots of things that I care about: family, children – not really very small children but their still my children, nonetheless. Obviously, I love and care about them very much. I would say things that I can do with them are obviously very important to me. I love also my sports. For my sins I did go and see Arsenal play last night and that was fairly miserable experience but I couldn’t come all the way back to London and not see them play even if they were quite horrible yesterday. But you can’t have everything.

KN: True.

DW: I have a question about MVP program. What do you think about it right now? How important MVPs are for Microsoft?

JRJ: I don’t know which forms people listen to but when I was an MVP, one of the things I valued the most as an MVP was the NDA alias that MVPs have access to. Likewise, I look on that alias as an employer of Microsoft and the feedback I get from that I take very seriously. I think MVPs have a very unique insight, a depth-level of your product and that it is in itself valuable. Clearly, they don’t speak for every person because they are specialists and probably in many cases the top 0.1% of professionals in the world and so their perception is normality of what’s normal everyday problem versus to the standard of which their code is like exponentially higher. You have to obviously take that into consideration as you balanced at the feedback. The feedback that we get especially from the SQL MVPs generally is very valid, very well respected but like I said – I’m giving you the broader context. You have to put it into the context that this is an absolute expert in the field of what they’re doing. Or it’s a very influential community person maybe speaking for a large group of people. You need to think of it in terms of those perspectives.

DW: A little bit different question – what hints would you give to young people who want to start working on the IT market?

JRJ: That’s a good question. I would say – getting involved. For example, I know where my career was at when I first got involved in the Community and I know where my career is now.

KN: That’s why we’re asking that question to very experienced people.

JRJ: And I owe an enormous amount to Simon Sabin who first introduced me to the Community and into being more of a volunteer and actually giving up time. I can honestly say I would never be able to pin point the specific ROY in doing a specific thing because it’s largely filled with intangibles. I guess at some point I just gave up carrying about that and I cared more about what I was able to achieve and what we were able to contribute. At the same time, I found that my passions took me in the direction that has obviously benefited me both personally and professionally. Without the Community investments I don’t think I would be where I am today and have the role that I have today. But in terms of just starting out, I think getting involved, going to use group meetings, going to SQL Saturdays is an integral part. But I think you should actually try stuff, being involved, be passionate about things and actually try to use technologies. I think that one of the things that a data professional is more of a boarding horizon and looking into new technologies and how they complement their already deep technical knowledge is definitely a thing.

KN: It’s very valuable spot because I spotted that some young people has got that approach: „Oh, you know, it’s SQL Saturday but I’m not going because it’s my weekend.”

JRJ: When I used to run my company called the Big Bang Data company which I run before I started my data profession and I used to take my son and he used to be on the stand with me on SQL Saturdays and he could deliver the pitch about SQL Data Warehouse as good as any Microsoft sales person that you would have ever meet. It was a lot of fun. I was very proud of him and seeing him getting involved and engaged with the Community and getting involved in a family business. I’m not sure that IT or life of data professionals is what he ultimately wants to do but even so – I was very proud of him for being involved and actually getting immersed and experiencing it for himself.

KN: Thank you very much. At the end of our conversation tell us where we can find you.

JRJ: So the obviously ways to find me is Twitter. You can get me also on LinkedIn. These are the two main social media platforms. But also I try to contribute to both the documentation for Azure SQL Data Warehouse as well. So when you are actually commenting on our docs and giving us feedback on things, you are connecting directly with the PMs and a writer as well. I think that’s a very important point and it’s not one to be forgotten when we think about how we engage with one another.

KN: Ok, thank you again for this conversation. I really appreciate it.

JRJ: Thank you. Take care.


Useful links:

James’s twitter: @jrowlandjones
James’s profile: LinkedIn
James on Channel9
Conferences: SQLBits

Previous Last week reading (2018-08-19)
Next Last week reading (2018-08-26)

About author

Kamil Nowinski
Kamil Nowinski 200 posts

Blogger, speaker. Data Platform MVP, MCSE. Senior Data Engineer & data geek. Member of Data Community Poland, co-organizer of SQLDay, Happy husband & father.

View all posts by this author →

You might also like

Podcast 0 Comments

ASF 016: Chris Webb interview

Introduction Chris Webb is independent consultant specialising in Analysis Services, MDX, Power Pivot, DAX, Power Query and Power BI in general. He has been blogging about Microsoft BI for over

Podcast 0 Comments

ASF 029: Tomaz Kastrun interview

Introduction Tomaž Kastrun is BI & DEV developer, data analyst & data science. With more than 15 years of experiences in the field of databases, business warehouses and development, with

Podcast 0 Comments

ASF 002: Greg Low interview

Introduction Dr Greg Low is well known in the SQL Server community, an internationally recognized consultant, trainer and author of a number of Microsoft white papers, blog posts. Greg holds a


No Comments Yet!

You can be first to comment this post!

Leave a Reply