In this episode of LogicTalks, Mark Banfield, Chief Revenue Officer at LogicMonitor, is joined by Ken Perry, Product Manager and Technology Monitoring Architect at Harvard University to discuss how LogicMonitor has shifted Harvard’s monitoring strategy. Ken explains how LogicMonitor has helped to break down monitoring silos between departments, solve unique challenges and use cases within such a large organization, and how LogicMonitor aided Harvard in it’s transition to remote learning during the COVID pandemic all while helping to build operational confidence throughout Harvard and it’s leadership.
Mark Banfield – Hey there, everyone, and welcome to LogicMonitor’s “LogicTalks,” a conversation with industry experts and some really innovative customers to dive into topics and technologies that make IT so crucial to the modern business world. Thanks for joining me on this episode. My name is Mark Banfied. I’m Chief Revenue Officer here at LogicMonitor. So today, we have a really exciting and innovative LogicMonitor customer from one of the most finest universities in the world, Harvard University. Ken Perry, who is Product Manager and Technology Monitoring Architect is with us today. Ken has been at Harvard for more than 12 years, holding positions ranging from Senior DevOps and SecOps Engineer to Microsoft Systems Engineer during his tenor. He’s a true expert when it comes to monitoring and observability systems, as well as large scale system design, deployment and administration. Welcome, Ken.
Ken Perry – Hey, Mark. Happy to be here.
Technologies and Complexities at Harvard
Mark Banfield – Yeah, we’re really happy to have you here. And I’m really excited to learn about some of the innovative ways that you have used LogicMonitor and continue to get value out of it. So let’s just kick off. So Harvard University is obviously one of the most famous universities in the world, and it’s a huge organization. Can you give us just a brief understanding of some of the complexities of the technologies that you support?
Ken Perry – Yeah, so some of the technologies and services we use include a merging of the oldest and the newest technology from Harvard. So we have CAT-3 at a lot of our Harvard houses, and then we’re also using the most sophisticated cloud-based infrastructure and auto-scaling in our DevOps teams and development cycles. One of the more interesting use cases we’re doing right now actually is single pane and glass monitoring for our faculty deans. Harvard professors live in each of these houses, along with the students, and they’re in charge of helping them succeed. And right now, in response to the COVID pandemic, what we’re working on is building the real time dashboards for each of those houses with Zoom services, all of their wifi access points and the performance and utilization in those houses, and then a real time monitoring for Canvas, they’re teaching software. And so we’re able to achieve that single pane of glass for these faculty deans, so that they can ensure that their students can learn online when they have come on campus.
Pain-points and Hurdles with Previous Monitoring Solutions
Mark Banfield – So one could imagine that monitoring such complexities throughout such a vast organization, really isn’t an easy task at all. Can you speak to some of the hurdles and pain points that you and your team had with the previous monitoring solutions that you were using?
Ken Perry – Yeah. So one of the biggest pain points that we had was it was hard to investigate and respond to outages. You know, oftentimes these service interruptions happen at 2:00 AM, and you’re trying to log into all these different tools and systems at the same time. And so we needed a unified single pane of glass, but we also needed a system that integrates with our other systems that we use on a daily basis. And so that’s one of the places where LogicMonitor really succeeded is that we can tie all of those technologies together. Some of the other other issues we encountered is that, you know, we have developers developing some really sophisticated software, and they’re spending a lot of their time actually building the monitoring and the monitoring solutions instead of actually developing, improving their app. That’s a drain on your resources. With LogicMonitor, you know, you have support for some of the newest things, some of the things we’re using right now. And so that’s an amazing thing where I can go in and meet with these developers and say, “Hey, it’s already monitored. LogicMonitor found it. When you spun it up in your AWS cloud, what thresholds do you want to be alerted on?” And so that was a big pain point that we’re absolutely super happy to have overcome.
LogicMonitor Use Cases across Harvard
Mark Banfield – Touching more on that, since being tasked with fixing monitoring at Harvard, how are different schools and different departments using logic models? So for example, the medical school, or for example, different user communities such as DevOps?
Ken Perry – Yeah, we have monitoring across all the schools directly or indirectly. What we’re doing with medical school right now is integrating monitoring into their backbone infrastructure and then also into their finance and alumni infrastructure. So some of the highly critical stuff, they’re very pleased that we can give them that single pane of glass. I can say, “Here’s your dashboard of your MySQL database, the VMware vCenter, the front-end website, the multi-factor authentication step through using our Caz Harvard key. And we can give them that single pane of glass in one window. And that really gives them the confidence to say, “You know, Hey, my system is healthy.” So the medical school’s really happy about that. And from the DevOps perspective, Crimson Clear is probably the most critical service we’re supporting and monitoring right now. It’s our wellness attestation tool that all faculty staff and students are going to have to attest to their health daily before they come on campus. And we’re using Fargate containers to dynamically scale that on demand. And what we needed from LogicMonitor is support for that out of the box. LogicMonitor’s automation automatically found those Fargate containers and began pulling in all the cloud watch metrics that come with them. So absolutely critical monitoring going on right now for our most critical tool, and LogicMonitor’s absolutely knocking it out of the park on that front.
Visibility into Compliance and Security Risks
Mark Banfield – Alright. So you talked about supporting the medical school, but I’m sure there’s many others that across the organization that require compliance. How’s LogicMonitor helping you give you visibility and manage some of that risk?
Ken Perry – Yeah, so on the visibility front, we have applications that deal with student data, financial data, you know, medical research data and things like that. And so recently our security team for Harvard University mandated a core set of security tools, and that was a big push across the organization, and what one of our developers was able to do is he built a data source that monitors all those critical security requirements. And so we were able to, with the extensibility of the LogicMonitor data sources and the tool, build out this real-time monitoring thing where we have Dashboard saying, you know, “These 27 systems are healthy and these 27 systems are not,” and we could give each of the product owners and the teams a real time view of, you know, you need to install the tool here. You need to make sure that it starts up on running time here. And so that was a huge win. And at the core of that was a LogicMonitor tool.
Breaking Down Silos in Monitoring
Mark Banfield – That’s great. So again, you have a very large organization with many different types of specialists use cases. How has LogicMonitor helped you break down silos in monitoring across our very large organization?
Ken Perry – So last week, for instance, I had 37 Zooms with different product owners and things like that, and that was last week alone. And so the demand is as high as ever. I’ve found that the key to winning over faculty staff, product owners and administrators is really giving them the data visualizations and the insight into their data they really want. The product owners are ranged from technicality, but what they all want is a way to gain insight into their system performance and SLAs and data. So I’m able to say, “Here’s your website performance broken down by DNS resolution time, SSL certificate health, here’s the authentication process and how long that takes, and here’s your website healthy, and here’s when you patched it.” And so I’m able to win them over with, you know, the data visualization that LogicMonitor provides. Empowering the users is really one of those, one of the reasons why we’ve succeeded is it’s there. You give them that power, and the word spreads.
Monitoring Remote Learning Tools and Infrastructures
Mark Banfield – So earlier this year in March, obviously the coronavirus pandemic hits, and universities in particular had to move very quickly to start to shift to online learning. What was that transition like for Harvard?
Ken Perry – So, it was definitely an interesting transition in that immediately everybody went to 100% work, remote. All the students went off campus to learn remotely. Everybody in Harvard IT was working day and night for that transition. And so we have to have 100% uptime of our VPN. You have to have 100% uptime of Zoom, and we have to have a 100% uptime of Canvas. And we needed to monitor that in real time and respond to any potential service interruptions before that. And so all lights started to shine on our monitoring and performance tool. Daily meetings with leadership, the development of a essential services monitoring dashboard, which contained all of our most critical tools, including Zoom and Canvas and ServiceNow and VPN and all of our core infrastructure. And so in that, LogicMonitor reached out to us saying, “You know, how can we help? We know this is a time of crisis.” And we’re like, “We need extensive Zoom monitoring and performance monitoring because all of our classes just instantly went remote.” And so LogicMonitor went above and beyond in that they developed a custom Zoom data source. I think that was a huge boost for all of academia. ‘Cause we were all using Zoom all of a sudden, and you import that data source and you input the dashboard and Zoom pops up and you can see how many minutes you have, the health of the tool and things like that. And so not only do we have amazing idea leadership, grabbing everybody to train everybody in Zoom and Canvas, we have LogicMonitor saying, “You know, how can we help?” And so that was absolutely fantastic in facilitating that transition to online learning.
Building Confidence in Operational Performance and Resiliency
Mark Banfield – Yeah, I can imagine it was an incredibly difficult and turbulent time. And you’ve talked a little bit about how LogicMonitor has helped. How has LogicMonitor helped build more confidence in your operational performance as well as with your leadership across Harvard?
Ken Perry – So it goes back to the insight on what’s happening in those tools. And so our executives, when they all went remote, and they were feeling like they had no insight, and LogicMonitor was able to bridge that gap where we can say here’s Zoom and here’s the cycle of meetings per day and here’s VPN and the performance and utilization. And so LogicMonitor, when we hooked it up to the VPN, for instance, identified that our 10-gigabit links where no longer enough. And so we had to upgrade those 10 to 100-gigabit links, and we were doing that as fast as we could and making them redundant as fast as we could to keep up with the demand for things like the VPN. Nothing instills operational confidence in a tool more than the ability to see that data real time and to see that your systems are healthy. And so some of our key leadership, they like that NOC dashboard where it’s green or it’s yellow or it’s red. And we needed to know if, you know, we’re having a performance degradation beforehand, so we can communicate that out to our faculty and staff and students so that they can compensate.
Future of Education and Remote Learning
Mark Banfield – So, Ken, lastly, what do you think the future of education looks like? Will we see more online classes and other remote technologies? How is Harvard planning for this?
Ken Perry – I think that we’ll continue to see a surge in the uses of remote technology and a rapid evolution of those technologies, but I don’t think that there’ll ever replace in-person learning. And so I think that in-person learning experience is central to the role of K-12 and upper education. So I think that we have to remain agile for the online learning and the evolution and the move back to in-person learning. But what I ultimately envisioned as a teaching model is, you know, freeing our students from the classroom and freeing our professors from the classrooms in a lot of ways. What I would like to see us evolve to is take that Zoom meeting and show them an archeological site, show them how you’re studying cancer cells in the lab, you know, show them how to run a centrifuge, you know, real time in Zoom, and it freed us to do so many new things. So you’re not only have that on campus learning experience, but you also have, you know, the students can’t go to a level three laboratory, but that researcher, that educator can take them there virtually with Zoom. And so I think that Harvard IT is planning for that by enhancing our infrastructure and facilitating resolution of any type of interruptions in the service.
Mark Banfield – Great. Excellent. Well, Ken, really appreciate you taking the time today. We’ve learned a ton about the innovative use cases and innovative ways that you are getting great value out LogicMonitor across the Harvard organization. And we wish you the best of luck in the future. And for any of you that want to learn more about LogicMonitor or any of the things we talked about today, please visit our website at www.logicmonitor.com.