In our second conversation of our “I Am A Mainframer” interview series, Jeffrey Frey talks with Emily K. Hugenbruch, OpenStack Cloud Enablement Engineer, z/VM and Software Engineer at IBM about the OpenStack Newton release and her overall career experience as a woman working in the mainframe space.
You can listen the full recording and read the transcript of the interview below.
If you’re a mainframe enthusiast or interested in the space, we invite you to check out our new community forum.
Create a profile and post a selfie with your mainframe system, and you will receive an exclusive “I am a Mainframer” t-shirt.
Jeffrey: Jeffrey Frey, IBM fellow and retired CTO of IBM’s Mainframe platform, and very much a mainframe enthusiast. Today I am very pleased to host the “I Am A Mainframer” interview series, in a conversation with Emily Kate Hugenbruch. Emily is an IBM software engineer working on OpenStack and cloud enablement for IBM’s virtualization platforms, so we’re looking forward to having a great conversation.
The conversation is sponsored by the Open Mainframe Project. As a Linux Foundation Project, the Open Mainframe Project is intended to help create a mainframe focus, open-source technical community and to serve as a focal point for the development and use for enterprise Linux and a Mainframe computing environment.
The goal of the project is to excite the Linux community around the use of the mainframe and to foster collaboration across mainframe community and develop and exploit shared Linux tool sets, resources, and services within the mainframe environment. In addition, the project seeks to involve the participation of academic institutions to help assist in creating educational programs aimed at developing the mainframe Linux engineers and developers of tomorrow.
So Emily, welcome, it’s a real pleasure to talk with you today.
Emily: Oh thanks Jeff! Thanks for having me on.
IBM AND THE TEAM
Jeffrey: You bet. So listen, to get started maybe you can tell us a little about your role at IBM and your team, and a little bit about what you do.
Emily: Sure. So I work at IBM here in Endicott, New York, which is the birthplace of IBM, and I work on z/VM, and that’s the virtualization platform. It’s our oldest and our flagship virtualization platform For Z Systems, and within that, I am a team lead for the, what we call the cloud enablement portion of z/VM, so I have a team partially here in the US, partially in China that work on the OpenStack drivers, we work on them taking them up into what we call the cloud manager appliance, and we work on supporting that, so all of the documentation, all of the customer issues that come with it, and then we do a lot of the presentations about it and going out and talking about it.
So part of my job is the sort of boring project manager-y, team leader-y stuff kinda, hosting meetings, keeping track of where we are on milestones. And then kind of the more fun part of my job is working with the OpenStack community, and making sure that we’re contributing properly to the OpenStack community. That my team has the right focus on things, that we’re working on things in our drivers that we need to be working on and that we’re giving back to the community, because of course with an open source community, you don’t just want to be putting a bunch of code out there and asking them to take it.
So working on that part of it, and of course the part I really love is when I get to work with customers who are looking at trying to enable their z/VM system for cloud and how can they do that. I really love working with customers on that part of it.
Jeffrey: Very cool. I don’t know that you and I had ever met face-to-face but, I know a lot of the folks on your team and it’s been very important to get z/VM involved in participating in this open environment and to enable it so that it can plug into the various management tools like OpenStack so I’m thrilled to be talking with you today.
How long have you been with IBM and when did you first get involved with the mainframe?
Emily: This is my eleventh year as a full time IBMer and before that I had two internship projects while I was still in college, and they were actually both with the z/VM team, so I’ve kind of been a mainframer from the start. Even during my second internship I was working on some of the system management APIs and so I’ve been working on various cloud enablement systems management stuff for almost my whole career.
Jeffrey: Very good, very good.
OPENSTACK NEWTON RELEASE AND Z/VM
Jeffrey: So let’s talk a little bit more about some of the enablement work you’re doing especially around OpenStack and some of the industry standard stuff.
I know that z/VM, first of all for people who may be unaware, z/VM is an extremely robust virtualization platform for the mainframe. Some even credit IBM for inventing virtualization and it’s used by some of the largest customers in the world to virtualize their mainframes. Now with Linux running on that platform, there’s all kinds of opportunities that have been opened up to use the mainframe in an open environment for people who are seeking a more standardized, open kind of world, both in terms of the applications they run on and how to manage the system.
How is the OpenStack work going? I know that when I was with IBM we had started to enable the platform for OpenStack based management, how is that going?
Emily: It’s going really well, actually on Friday we celebrated our Newton release, so OpenStack Newton is the latest OpenStack release and we released our cloud manager appliance that matches OpenStack Newton, so that part of it is going really well, and as you said, this has been a bit of a journey with OpenStack and z/VM and it’s a fun thing to be a part of because there are a lot of things in the OpenStack community that really come from having that kind of distributed world background, as opposed to the mainframe background and some of those very basic differences are things like on z/VM, we’re familiar with running hundreds or thousands of guests on one system and in the distributed world, people have done things like that, but it’s not how you run a production. Whereas we’re really used to running in production that way, and so there are a lot of different design assumptions that you make when you’re designing something that has to manage, maybe at top ten guests on the system, as opposed to managing maybe hundreds of guests on the system.
So that’s where some of our challenges have been, and how to take things that maybe OpenStack on the distributed platform didn’t have to be as efficient about, but when you get onto the mainframe platform you have to be a lot more careful about efficiencies.
Jeffrey: Right, right.
I know that one of the things that we wrestled with was, you know, when you drop in with the history that people have with mainframe systems and the dependence that people have on these platforms running, as you said, thousands of guests in production today, one of the things that when you introduce enablement for a new management approach for a platform like that, it has to be integrated with the rest of the environment. It has to be integrated with what clients already have in terms of their tools, in terms of their process, their automation etc.
I imagine that balancing the new stuff with OpenStack while also preserving for compatibility and investment protection is another one of the challenges you face.
Emily: Yeah, that’s definitely true. So we’ve had a number of things that we’ve added as sort of special features in our cloud manager appliance that are not part of the OpenStack community. An example would be the discovery function.
Emily: So we have the idea that you could have an existing guest and then we will import some information about that guest and let you do some management of that guest through OpenStack without having to recreate it.
Other examples are, one thing we have for our Newton release that just came out on Friday, is that we offer CMS configuration wizards. So if you’re familiar with CMS at all, that’s an operating system that runs in guest on the Mainframe, it’s kind of our basic operating system that comes with z/VM, and so a lot of customers are very comfortable with using CMS wizards, like IP wizard for example, so we wanted to give them a CMS wizard that would help them configure their cloud manager appliance, rather than having them dig through Linux configuration files or something like that.
Jeffrey: Yeah, yeah.
Well this is really exciting because the IBM strategy around Linux on the mainframe has traditionally focused on making sure Linux and the capabilities of Linux, and the administrative control and the way Linux surfaces, expressed itself on the mainframe is standard Linux, Linux is Linux is Linux, and that’s been a major objective of Linux on the mainframe platform. Historically the other side of that and what people face is the management of these systems, and now especially with cloud and all of the advances in automation and using the virtualization of a platform as kind of the operational control of the platform.
It’s very important not just to get the Linux stuff standardized, but also the management of this platform more standardized or at least able to plug into a wider set of tooling choices that the customers are faced with today.
Emily: Yeah, that’s definitely true and when you get into especially the cloud space, there have been questions about, do we want to do something that’s more Z specific because we have a lot of really great features on Z that don’t necessarily translate into concepts on other platforms, but it’s really important when doing a cloud solution and you’re looking at enabling your end user developers, to collinate the I, or log in through a GUI to interact with our servers, then its very important for them to have a way to do that, that looks exactly the same as it does on other platforms.
Jeffrey: Yup, yup, yup, that is a very interesting challenge and one that actually celebrates the
value proposition in the mainframe is to be able to deal with administrative control and management control of a system, but also leverage all of the inherent benefits of the mainframe, it’s scale, it’s quality of service, especially it’s security and availability attributes. So this is a really cool thing that you guys are doing.
Jeffrey: Emily what do you think the biggest challenges are for the mainframe going forward. We’ve talked a lot about, you know, what we’re doing here to enable the platform. What do you see as the challenges in front of you?
Emily: I think one of our biggest challenges is really talking about what cloud means, and especially now that we’re getting into container technologies, how a lot of times those container technologies and how they interact with projects like OpenStack are still not very well defined yet. So when you add the in the extra layer of Z systems then that’s an extra challenge of how should these things interact? What do we need to do to make sure that our customers are happy with them, that they like how these things fit together and that it’s easy to put together.
There are a lot of things that you can do if you play around and really hack it but when we’re talking about the mainframe, our customers want something that’s very reliable, very easy, straightforward to put together and very well thought out and backed up by IBM so yeah, there’s a tremendous amount of work to be done with defining that interaction.
Jeffrey: Yeah, the other thing is, you know, more and more every day I see that things are just moving so fast and you have an entire world community, open community developing new tools, new approaches, new frameworks, that get adopted extremely quickly.
I’d imagine that you and your team have to be very agile in responding to the changes that are occurring and the new approaches that are being used to manage a system. It’s not like it used to be.
Emily: Yeah that’s definitely true, if I looked over even just the last ten or eleven years of my career and the different types of system management projects that we’ve seen then getting into cloud, it’s just really exploding when Gawker, and Kubernetes and Mesos, and probably lots and lots of other projects that we haven’t even heard about yet but are going to become very popular.
Our team does a lot of monitoring different listservs, getting involved in different communities. I think one of the great things about IBM is that for nearly any software project out there, there’s probably some team within IBM who’s working on it and so working with the research teams, or teams within other parts of IBM, keeping those lines of communication open is really important to us so that we know about these new things as they come up and we can start having that conversation about how they fit on the mainframe.
Jeffrey:Yeah, no, that’s very cool.
The thing that struck me about one of the challenges that the mainframe faces with respect to cloud, is that there are, I used to say that there are if you asked ten different people what they mean by cloud you’d get at least nine different answers and one of the things that I think could be most beneficial to our clients with respect to mainframe, it’s virtualization and its use in a cloud environment is to try and have the discussion with our clients that cloud isn’t a place, it’s a control mechanism, it’s a model for service delivery.
So one of the things that I think the main frame is very well positioned for, especially in our existing large customer base is to kind of cloudify the enterprise. You don’t necessarily have to move stuff out of the data center to some other service provider in order to get the benefits of the automation and the service delivery model of cloud. You’ve got this mainframe that’s ready to go, it’s ready to plug into these new management approaches so that you can incrementally transform your data center and your IT environment into a cloud service provider.
Emily:Yeah, I mean that’s definitely something that we see going on and the definition of what is cloud and why do I need cloud, sometimes you get system administrators that are not, or assistant programmers that are not really willing to give up kind of control of those things to their developers and other system programmers who maybe are completely okay with that but they don’t want to change the way that they’re building guests, so we’ve put things in like a cloning capability for OpenStack that lets you make an exact copy of another guest.
On the OpenStack side of things, and the developers point of view, it looks like a regular deploy, but in terms of what’s happening underneath it’s just a copy of another guest which then an allow your other maybe existing management tools to set things up like the IP address, and the host name of the guest and things like that and so coming up with tools like that that can hybridize this approach between the new world and the old world is very important to us.
Jeffrey: Yeah, you know what, I could talk with you about this all day if we had the time. The challenges part of his discussion is very interesting to me because, I mean you mentioned some of the perspectives and the attitudes that existing IT staff have related to the mainframe.
You know part of the reason that the mainframe has a reputation for being secure and available is not just because you people at IBM engineer it that way, it’s also because it’s managed that way. There’s discipline and rigor and process, how it’s employed in customer shops in terms of change control and making sure the system runs effectively. And so part of the fear I think some of these people have is that loss of control where they’re afraid that if they don’t tightly control things and everything has to go through the traditional processes, that the system will be degraded in some way.
So I think the challenge for you and others, is to make sure that as the system plugs into these new management approaches and gives more control and freedom to users of the system, that those IT guys can be assured that we’re not breaking any rules, and we’re not introducing any risks or putting the system and the services that are so critical to those customers at risk by doing it.
Emily: Yeah and so one thing that we often like to talk about is the idea of having cloud security separate from your z/VM security so that you have the idea of separate users in your cloud that are not maybe the same as z/VM user ID’s. You could have that linkage but you can also keep them separate and so having things like having our cloud management appliance work with RackAFx for example, so you have those audit records and things like that are definitely very important. To some extent I think that some of the projects like OpenStack and others, are still catching up a little bit to the maturity that the mainframe has around things like security and audit controls.
ADVICE TO OTHER WOMEN WORKING IN TECH
Jeffrey: Alright, let me hit on a little different subject.
As a woman in the industry, what advice would you have for others who are interested in working in this space, IT in general, but even specific to mainframe. You must have perspective on that.
Emily: Definitely I worked more with mainframe at first and now I’m moving a little bit more into the open source world and definitely in terms of the male to female ratio, the mainframe world is a lot more even and to some extent. I think this is because some of the women who started maybe during the eighties and nineties and things like that are now technical leaders in the mainframe world. There have been a lot of really great women who have been mentors and role models and things like that in the mainframe world.
Then in the OpenStack world, I’m involved with groups like the Women of OpenStack and the OpenStack Mentoring Project and we’re struggling to get double digit attendance for women in some of the OpenStack conferences.
So definitely if you’re a woman, finding the support groups for other women in the industry is really important, but I will say that the mainframe is a really great place to be from that perspective, there are just a lot of inspiring women here.
Jeffrey: Good, good.
THE FUTURE OF THE OPEN MAINFRAME PROJECT
Jeffrey: So one last question for you, as an open mainframer project member, what do you hope to see within the Open Project in the community in the future. What objectives would you like the group to achieve in the near future?
Emily: We’re just starting this, as you know the Cloud Stack Consortium as we’re calling it in the Open Mainframe Project, so I’m hoping that this will help answer some of these questions that still remain about exactly what cloud means on the mainframe act, what these technologies, the container technologies and things like that.
I always hope from the perspective of myself and my team working on the OpenStack drivers that we get participation from the other companies. SUSE has been great with that but we’d love to see more partner companies really jump in with both feet to the Open Source projects and get involved with spreading the good news about mainframes.
Jeffrey: Yeah, no that’s great because you know, you mentioned SUSE and others to be able to really demonstrate an effective integration with some of these enterprise management tools around cloud and virtualization and demonstrate the power in using a set of open approaches like OpenStack on the mainframe would be extremely powerful and I think maybe the project could assist in incubating some of that, but also act as a way to spread the good word on successes that we have in that area.
Jeffrey: So hey listen, Emily this has been terrific. Thank you so much for having the discussion and you’re in Endicott that’s not actually far from me, maybe I’ll get down to see you and some of the old members of the team. I’ve got friends in Endicott and so maybe in the near future, I’ll drop by to see you.
Emily: Well we’d love to see you and we have lots of speedies here so.
Jeffrey : Alright good. So that does it for this session, thank you very much.
Emily: Thank you Jeff.
If you’re a mainframe enthusiast or interested in the space, we invite you to check out our new community forum.
Create a profile and post a selfie with your mainframe system, and you will receive an exclusive “I Am A Mainframer” t-shirt.