Skip to main content

Denise Kalm on Coaching and Consulting Mainframers

Reg Harbeck talks with Denise Kalm, coach and consultant, about how younger mainframers can make an impact on the platform and logical next steps for experienced mainframers. Listen to the interview via the orange play button or read the transcript below.

Reg Harbeck: Hi, this is Reg Harbeck and today I'm talking with Denise Kalm. Denise is one of those extraordinary people who has worked on the mainframe and on a whole bunch of other things at the same time that are non-mainframe but are connected to the mainframe. She is an author of a number of books about careers; she is a coach and she also has been on the CMG executive among many other things. Well Denise, why don't I get you to introduce yourself to us and tell us about your background with the mainframe, with IT, with coaching and anything else you would like to tell us about?

Denise Kalm: Well thank you Reg. The first thing I would say is this: IT was my second career. I actually started out in biochemical genetics a number of years before that became a hot field and with no jobs around, I went into IT almost by accident. I think a lot of people my age have made that move because there were jobs. They looked well paying and I thought ‘I'll go work there for a while until I figure out what I want to do.’ Well 35 years later, I haven't figured that out yet I guess. So I started on mainframes for the phone company and migrated briefly to Tandems for awhile, back to mainframes, toyed with UNIX performance for a bit and then went to work for vendors. So I've had a lot of variety in my career and I think that is what keeps it fresh is that doing different things, doing different roles, playing with different platforms, none of which actually are as good as the mainframe in my opinion. Sometimes exploring what you think you would like to do when you retire. In one case, in my case, that was going back to school in my 50s to get my coaching certificate, so I did coaching on the side for quite a number of years until I formed my own business which is comprised of doing consulting primarily on the mainframe, coaching and doing some outplacement coaching for one of the outplacement firms.

Harbeck: Now that's really relevant to today's mainframe workforce because our mainframe workforce is shaped almost kind of like a dumbbell. You know on the one end you got the big weight and on the other end, the big weight and the big empty bar in between because on the one hand we have this massive number of people who are at or past retirement age and trying to decide what they want to do next in their lives. On the other hand, you've got this influx of new people who haven't really spent the time learning and mentoring up enough to sort of you know have the mainframe handed over to them entirely with not a whole of lot of us inbetweeners that are sort of midway through our careers on the mainframe. So you must have some interesting coaching advice both for people who are at the end of their typical mainframe career and trying to decide whether to move onto something else or redouble with mainframe expertise and people at the beginning who are trying to decide how to go forward best with their careers. Maybe you could offer some thoughts in both of those areas.

Kalm: Well, let's start with the new mainframers because that's where we really have a strong need. I'm not the only person at my age thinking about retirement, thinking about next steps so there is going to be a lot of mainframe jobs out there. Because so many people think it is sexier to work on servers, I think you're going to find that there is less competition and presumably better pay. Here is the challenge: Where are you going to get the training? When I started out, first of all it was a lot less complicated. I started with CICS 1.4. I don't remember what MVS release it was but it was early on and we had 3 byte addresses, you know virtual address space was very, very limited so you know the evolution has done what the evolution of every product on computers has done, made it more complex. So the relatively simple job I had of learning enough to become useful has become much more complicated. I would advise people are thinking about a mainframe career to get into it as soon as possible because those expert mainframers, once they retire, that institutional knowledge will be gone. Now maybe we can bring them back as consultants and that leads to the next step but if you want to work side by side by an expert and get a much quicker path to the 20 years or more of experience that they have, you want to get into those jobs while they're still working there. The best way to do this is to befriend them, be willing to work hard and look things up a lot of the time first.

Harbeck: Oh, that is so important.

Kalm: Make the effort. Don't assume the person is like another Google search.

Harbeck: Right.

Kalm: That you can just go up to and ask. It's not very flattering. Make notes. Don't ask the same question two or three times. You know develop a learning plan with them. They may give you some really good guidance on best ways to start, how to approach it and don't try to bite off the whole thing. In one year, you will not become the mainframe expert that you're learning from. It isn't possible. There is just too much. Pick an area that you think you can start to master. Say you start in operations, a lot of people have. I did a few months as a temporary person in operations and that gave me a really good grounding because that's how the real world works. That's production and it gives you an understanding of what the people in operations are up against particularly because now with the complexity most of them don't know what those applications represent. They are trying to manage a conglomerate of obscure names and systems on multiple platforms and they don't know what the business priorities are. Once you have done a little operations, you move into an area that sounds appealing like database management. Schedulers actually, workload automation is a great place to start because it is a nice blend of operations and more back office work where you can learn quite a bit about how things work but, whatever you choose, try to define it a little more narrowly so that you can build your competency and create value early. You will feel better about it and you are not trying to eat the entire 747. You are just trying to take a bite at a time.

Harbeck: Hmm. Makes sense. So now having taken a look at some really good advice for how a new mainframer is starting out, let me add my own thought to that just because I'm reminded of IBM-MAIN-that famous mainframe emailing list: “IBM-MAIN”—Google it up—and how one of the biggest mistakes that newbies on IBM-MAIN make is asking questions which have already been answered if they were just to search through the archives or questions to which the answer is obvious if they were just to look up the answer themselves before asking a “dumb” question, but on the other hand of course then a lot of the established mainframers are people who tend to be you know very focused on work and don't much like being interrupted by casual and insufficiently considered questions but these are going to be our future experts. These are going to be the mainframe established wisdom people in a lot of cases who move into a consulting expert role. What thoughts do you have for people who are moving toward the end of their traditional mainframe career and considering whether to move beyond the mainframe or redouble their mainframe involvement?

Kalm: Well I think it depends, Reg, on what you call moving beyond the mainframe. In some cases, that may naturally be retirement. One thing you discover when you get into your 60s is that your memory doesn't work as well. It's a simple fact. It's not Alzheimer's. When you hit 90, you are not going to remember very much and it is a big challenge so it becomes more difficult sometimes to do the job. So some people will really retire. Now that might still involve some consulting, teaching at some place like Phoenix University, you know there's a lot of different ways you can go, but there are people who will actually just plain leave. Some people as the mainframe gap gets bigger will see that the best thing they can do is take their retirement buyout package and then come back as a consultant. What that is going to take though is having your name out there. If all you've ever done is work for one or two companies, you've never published anything, you've never written on anything, you are going to be going up against consultants who have that experience as Reg and I know. Having a name in the field is going to be the difference between success and failure in consulting, but if you go back and teach classes and do various things, mentor for people if you are in a town where they are teaching mainframe skills at the university, anything you do that gets your name more widely out there, it can add the LinkedIn and start a posting things you've learned and you know. That will make a big difference to what options are open to you. Now an older mainframer may be able to get another job relatively easily because companies in desperation are starting to look at the fact that there are older mainframers are quite talented, quite valuable and maybe they've slowed down a little bit but it is better than getting a raw newbie that knows nothing. They are less biased against it. The baby boomers had a great work ethic and that's something that companies are increasingly desirous of having so that's a really important point there.

Harbeck: Well these are good thoughts. Now before we sort of wind up our conversation, I mean I've talked to a lot of people on these conversations who've got experience with SHARE and you have of course as well that SHARE is such an important user group but of course you also have a very special experience with CMG. We don't get enough opportunities to talk about the really important role, the performance monitoring measurement management has across IT that really started on the mainframe and has its in many ways highest manifestation on the mainframe. Would you like to talk a little bit about your experience with CMG?

Kalm: Yeah, I got involved fairly early on because my second major job was doing Tandem performance. I went from being a programmer, CICS assembler programmer to going to do Tandem performance, which is a story for another day. I started hearing about this organization Computer Measurement Group. When I could, I went. Obviously as we all know, you can't always get to a place every year but I tried to get more involved and we had a regional group called Northern Cal CMG that I got to attend and got more involved. It was great training, very tightly focused on the areas I was interested in, performance and capacity planning. As my career moved on, I got more interested in being involved and once I became a vendor, I started writing papers every year and found out how much fun it was to present at a conference. I then moved into being a director for one term and then came back. In the last four years, I've been vice president and that gives you the opportunity to guide the future of the organization, try to understand where the future should be and make that happen. So it is very exciting to be involved. I think this is probably my last year. I'm going to let another generation take over because I think we need some new fresh ideas but I probably will still try to speak every year at CMG because I really enjoy staying connected with those people. For your listeners, look on the CMG.org website. Anybody can subscribe to Measure IT, a newsletter that has really great content. If you are interested in performance and capacity planning areas, it is something to look for.

Harbeck: And maybe just as one more thought before we get to our winding up thoughts on this talk, if you could just maybe give some insights into why does measurement still matter in IT generally and in the mainframe specifically?

Kalm: Well, that's an interesting question. There was an assumption particularly when we moved into the Windows servers that if you needed more capacity, just buy more but buy more what? Too often, I saw back in the UNIX days people buying a bigger box but what they really just needed was more memory. If you don't understand what the problem is, you can't buy the fix. There was a period where IBM had powered down CPUs but they had lots of them but after people installed them, their batch jobs started running very badly because on the whole, batch runs on a single processor so if it is slower than the old one, having 10 of them doesn't help you. It is sort of like a woman, if you have nine women that are pregnant, you can't get a baby in a month. If you have one batch job running on 10 processors, it is still going to only run on one. If it is slower, that's going to hold you up. Without measurement, you don't know what the problem really is and sometimes the fix is a lot cheaper than just going out, throwing out hardware and buying new. In addition, you have to know what to measure, what metrics are important and what do they mean to predict and prevent problems. By collecting measurements and having a discipline, you actually ensure that you can prevent most problems that are related to performance and capacity, keep people up and running which is increasingly a deal breaker for companies. If you are slow, if you're down, they will find someone else to deal with. There is always another option so I think it is absolutely critical. IBM and other vendors have touted for years, well we can do this automatically for you but I don't believe that that has really proven out. I still see plenty of people fully employed doing this work because in fact everything comes down to “it depends”—and it depends, the fuzzy logic of it depends relies on a person to interpret, not a computer.

Harbeck: It comes down to taking a professional approach in order to get the really best results then.

Kalm: Yes, absolutely.

Harbeck: Okay, well we've gone a bit long and this has been absolutely worth it but maybe if you have any closing thoughts that you wanted to share with everybody.

Kalm: I think the one thing I would say is that if you’ve started going into autopilot with your career, put more effort into it. It will really make a difference to ensure that you're continually challenged, stimulated, using your best stuff and doing something you love. If you can't say your job’s like that, then start thinking about what you should do to change it because we should all be loving our jobs a lot more than most people are. You can take charge of that.

Harbeck: Excellent. Well thank you very much Denise. This has been a really neat excellent call and really informative. I appreciate you taking the time.

Kalm: Thanks so much, Reg.