Join AITSM Co-Op, A community for those who ♥️ AITSM (ITSM, ITIL, AI and Automation for ITSM)

A Productive Rant About Automation (Part-1) - The Biggest Challenges in IT Infra Automation

June 14, 2022

A Productive Rant About Automation (Part-1) - The Biggest Challenges in IT Infra Automation

A Productive Rant About Automation (Part-1) - The Biggest Challenges in IT Infra Automation. How Conversational IT & Automation Can Help in Faster Resolutions?


A Productive Rant About Automation (Part-1) - The Biggest Challenges in IT Infra Automation. How Conversational IT & Automation Can Help in Faster Resolutions?

"I can see there's a myth about the Automation. Once we do it, it should do some magic. And to be frank, I think we have to really understand - How we can change the thought process, the way of working and then bring the automation to enable that working model." - Says Sapna Gupta, our esteemed women thought leader and guest on the AITSM show.

Sapna Gupta is Infra Automation Advisor at Global IT Automation CoE, Cargill, having 20+ years of experience.

She put forth interesting analogies and forthright explanations to understand the anatomy of Infrastructure Automation. What are the most pervasive problems & biggest challenges? How to use Conversational IT/AI and Automation to get faster resolutions? (More insights and exploration in the next part/episode)

From her plenty of examples, enthusiastic spirit, and point-by-point conversation from 2+ decades of rich experience, we'd learn a lot from her.

Don't miss the end, where she connects a relatable example of LinkedIn connections to reinforce an effortless future state.

"Do we really make efforts to maintain our LinkedIn? Not really. It just happens. Make it so seamless that people are literally not putting the effort, but it is just happening. In the IT environment, all the different devices are connected to each other in CMDB, we can use conversational AI or automation."

Learn and Grow, With AITSM Show!!!

Transcript

Anand Tambey: Welcome back to AITSM show. Today we have our woman thought leader special season, and we are having Sapna Gupta with us. She had a rich experience in IT service management, and she is currently working in Cargill as global IT Information Advisor. And we will get to know more about her during the interview also.

 Proud to have you and welcome to AITSM show. Please let us know about your experience, what is your story? What is your experience about, and then we will move towards our topic.

Sapna Gupta: Yeah. Thank you. Thank you, Anand. Thanks for giving me the opportunity and And invite me on this particular platform. And really glad to connect with you and speaking over here and sharing my experience. Overall I have almost 20 plus years of experience in IT industry in multiple different profiles.

Okay. So I started my career with a hardcore technical person into the infrastructure space with a windows administrator, then database administrator, Unix administrator storage administrator, and with all the different experience on the different technologies, which was emerging over the period of time where we were having standalone servers to the virtualization world and all, so like a different technology experience. So over the period of time and into the management area where I got a chance to work on the different migration project, which was from the on-prem to the cloud and then the data center consolidation projects.

Okay. So it's kind of like a mix of the different kind of a landscape. And the technologies, which was emerging with the time as I was going in my career as well. And today we see that there's quite a mature IT infrastructure that we have which is spread across, the hybrid cloud and a lot of organizations they're adopting this as strategy and strategically on the cloud and it's not like a craze now that being on the cloud, rather it's become has become a kind of a need.

Anand Tambey: Right.

Sapna Gupta: I have never been dedicatedly working on a specific profile on the ITSM, but as it is kind of a very integrated with the whole IT management part. So yes, I can very well relate it with that and currently in Cargill I'm working on the automation advisor profile, which is very specifically focusing on how we can improve the automation maturity across the different infrastructure technologies.

The infrastructure is kind of a base layer, which enable the next IT system, which is like the application space. To support them, to adopt the DevOps model, the agile methodologies. Okay. So it is not possible until that the application team has a handshake with the infrastructure team.

Okay. So this is where I'm more focusing on. And connecting back with the topic, what we have today definitely that the way the ITSM layer is a horizontal layer connect with the, all the different technical towers. Okay. And create a kind of a process or a kind of a methodology. And how we can ensure that we are adhering to the process guidelines.

And handling that incidents, the request, the problems, the change across the entire IT landscape. In the same way the automation is also a kind of a horizontal layer. We can not see in the isolation. And it is not limited to one particular technologic tower or the process. Again, it is also quite well integrated with each of the process, each of the technology. And once they work together, that's how we can see the benefit, the real benefit out of it.

Anand Tambey: So actually, when you discussed about the theme I was really very excited to hear about it. Say we are working in a area of ITSM and Automation and AI, but there are some things which needs to be changed, they are kind of myths, our beliefs, whatever our beliefs about automation.

When you are topic said like it's a productive rant about automation. It really intrigued me and interested me. So what are your views about automation here?

Sapna Gupta: Okay. That's a very good question. And I can see there's a myth about the Automation. Once we do it, it should do some magics. Okay. And to be frank I think we have to really understand - How we can change the thought process? That changed the way of working and then bring the automation to enable that working model. Probably I'll ellaborate it more. If we see that years back. Okay. The way we were monitoring our IT land scape is based on the threshold.

Okay. So if any threshold is crossing, okay, let's send an alert, let the analyst be aware about that alert, maybe through the emails, through the incident tickets, whatever is a different medium, and then take an action on it. Okay. So that was the initial level of the maturity the day when we were working sometime back and we feel like proud that we are monitoring everything, but if we see in today's contest, okay.

Everything every configuration settings. we can define the threshold and we can monitor and that result in the thousands or the lacs of the emails or the like a ticket volume, which is not possible to handle. Okay. So now the need comes how we can have the integrated view. Okay. So which can remove the duplicates and can bring a single event information for example, Like, if something has broken it will create multiple tickets, network layer, database layer, OS layer, application layer. Maybe the event is one.

Okay. So we have to work how those different incidents and the others can be clubbed together. Okay. Now that's what we are calling out the correlation of the events. Okay. Now the next thing is, once we know this is. And most of the time, like 70, 80% of times we know these are the specific set of the tasks that we have to execute.

Even if it is going to the L1 person, he's having a SOP(Statement Of Procedures) defined where, he has to check following things. And it takes some time because for each of the layer there's a different analyst. Okay. So checking the server. There's a different analyst, for database, different analyst, and everyone is taking their own time to check the first initial test.

Okay. What has gone rogue, so now, then we know this is the repetitive task. Can we look to automate and check later, all the checks happened in the parallel. Once this test are done, we have the observations that this is where some fault is being observed and rest of the test has been passed. All has an interdependency on each other, so that kind of outcome.

So the human interactions on our engineered efforts is going more than troubleshooting, finding rather than creating that finding. Okay. So that's where the automation is actually helping it out to integrate monitoring findings and then executing certain checks and balances, and then resulting to some of the findings which can help.

And another is to narrow down the problem statement and that can drastically reduce the overall resolution time for such kind of events. Another example which I would like to quote from a process point of view, maybe some of the tasks like the disaster recover y. we used to create replica of the VMs that we see like protecting the VMs.

 But how many times we do the testing. Okay. But whether the DR environment is really working. Can we use. certain automations to actually test that whether the DR execution is happening seamlessly. Because if it is not seamless, I think we are losing that the reasons behind that, why we are creating the DR.

So I hope it explains.

Anand Tambey: Yeah. From a VMs perspective if it is a web services based environment load balancing on VMs, or there could be like personal working machines that are also there. So for infrastructure automation, how do you see IT process automation can help or kind of conversational IT or likewise that can help.

Sapna Gupta: I will take a very small example of life of CMDB updates. We all know it's like one of the core that we need to have the right information in our repository, which can enable us for many things, all of our processes, the integrations, the automations there's a lot of dependency on CMDB It is very easy to say that everyone should have the good CMDB right CMDB, right information, correct information, up to date. It's very challenging. Okay. Now, my thought process is, do we really take efforts, make efforts, okay to update our Facebook list, to maintain our LinkedIn. Not really. Like we had never been so thoughtful that each one of these friends should be in my Facebook list.

It just happen. Okay. We just comes and they just click and it just happened. We literally don't take the deliberate efforts. Okay. I think in the same way the way our friends circle, our connected people. We want to make a linkage in the same way. In the IT environment CMDB is kind of the same thing. Like all the different devices are connected to each other, how we can use the conversational AI or the automation.

Okay. Make it so seamless that people are literally not putting the efforts, but it is just happening. Okay. Likewise, we can have the chat bot. Which is programmed to some of the checks and the balances, like if some of the CI is having, having this particular IP and this IP in the different database is mapped to a different VLAN and this VLAN belongs to that site, but in your CMDB server is located maybe in India, but this VLAN belongs to maybe Singapore, that means there is some mismatch.

Okay. So why, so we can actually program in the backend to highlight such kind of a good discrepancy and can use the chat bot to connect with a specific owner residing in Bangalore because the IP indicate this is from Singapore. Would you like to revalidate, but you want to redirect this conversation to someone else.

 So in this process, actually, we have validated the owner person also. Okay. And we know who should be the accountable person.

Okay. We are validating multiple things and it is actually become very effortless and auto update, which can help entire organizations or the different processes in a long run.

So the conversation AI definitely can help in many ways actually.

Anand Tambey: Right, right. So it can also be like just as you said it could be like to identify the proper CI owner and If there is kind of some kind of automated process runs behind the same, say RPA or likewise, and which can actually detect that these IPs from different area, than whatever the owner is on.

And then it puts an alert onto that person, some IT Service Manager. There is this anomaly detected what to do?

Sapna Gupta: Yeah. Absolutely!.

Anand Tambey: Right. Thanks a lot we will continue this insightful discussion in the next episode. Please stay tuned. Bye-bye.

Sapna Gupta Profile Photo

Sapna Gupta

Infra Automation Advisor at Global IT Automation CoE, Cargill

As Infra Automation Advisor and member of Global IT Automation CoE, Sapna is accountable for developing automation framework, automation maturity model, strategies and automation roadmap for global IT.