DevOps Online Training | Step by step practice

so we love coming to velocity right it's like great to be lost in a sea of best practices everybody in this audience believes in unit tests you believe in load testing load balancing continuous integration all the good stuff you're either doing it or you feel genuinely guilty that you're not doing it so there's nothing really I could tell you about technology or tooling and well automation is key DevOps is more than just automation so one of the things that we've been very fortunate at pager Duty is we've gotten to talk to thousands of the world's best Ops teams best dev teams best DevOps teams and more than a handful of teams that are genuinely struggling and so we've really gotten a chance to look at a lot of the people aspects of the opportunities and challenges facing teams and so one of the things we did to quantify those challenges as we sent out a survey and you know we did all the basic stuff we send it to people in various job titles in companies in various verticals and various sizes and we got a couple of hundred responses and one of the things that jumped out is despite the fact that a hundred percent I say effectively a hundred percent of companies lose money due to downtime we were finding that 85% of companies were honest and they said that they have missed a critical alert or they've been unaware of an outage affecting their their infrastructure and the good news for this room is that by and large DevOps teams performed much better and this is teams that self identify as DevOps whatever that means to them we're finding that they're they're missed alerts were far lower there were about two-thirds of teams that were willing to admit that they had missed some kind of essential notification about their system they were responding faster typically within half an hour often within minutes and while stakeholders are never happy during an outage we were finding the devops teams had less unhappy stakeholders but and now now the scary slide we were finding that the average team was finding it was getting information from at least six monitoring tools right and this means a lot of the tools from the vendors on the exhibition floor today this means custom tools this means legacy tools tools in the middle of a migration and only a quarter of them were willing to say that they were confident that they were doing some kind of aggregation and then here's my scariest statistic of the teams that are trying to migrate from ops to DevOps about 80% of them were failing felt they were failing in some significant way and so we took a step back and we were wondering why is this right it doesn't make sense DevOps just seems like such a natural outcome in the sense of the only infrastructure the only place you make any money is production right and the only thing that makes you any more money typically is more functionality that looks a lot more like more code so you would expect that Devon Ops would be working hand in glove and that everybody would be you know paddling the boat in the same direction and we find this not often the case and so the number one thing we were finding that you didn't see right there's a lot of words on this slide but the one that I want to highlight is empathy right as a developer until you've truly been woken up at 3:00 in the morning putting out a fire that's not your fault in code that you didn't write that's not documented that doesn't have tests where you're losing money and a stakeholder is yelling at you you don't truly understand why there's so much process why there's so much hesitation to just let you upload your PHP script onto the server and test in production it's fun but it's it's scary when it doesn't work and it doesn't work and so one of the things that everybody in this audience knows to do is you build resilient systems right nothing has 100 percent uptime no matter what you tell yourself so you set a target let's say your target is fails one in a million times and your your up time is really the function of your subsystems so you've got one system that fails one in a thousand times well to hit a million one in a million you're gonna need to duplicate that you're gonna need to have an independent backup that fails one two thousand times 1000 times one two thousand one a million but you need to do the same thing as people like we genuinely believe that a person cannot be on call for a service no matter how good right a policy cannot be on call for a service you put something in the wiki that says you must reply in five minutes that's great your wiki can't actually get up and do anything so what you need to do is look at your team right let's say the average person in your team misses an alert one in a hundred times right they sleep through it they're out of service that's that's actually a pretty good number but again to get to that one in a million you need to have three levels in your escalation policy so that the the secondary and the tertiary who are called far less frequently but they they miss their alerts one in a hundred times as well 100 times 100 times 100 you're back to that magic one-in-a-million number that I just arbitrarily picked for you the other thing is I like doing I like working at 2:00 in the afternoon with a coffee in my hand and you know two monitors set up and I know what's going on and it's my code but that's not when stuff breaks stuff breaks when you're not prepared and you don't know what's going on and it's not your code so you want to not only aggregate your alerts so that you know what business system is affected how bad it is and what underlying systems are probably causing it but you want to provide as much situational awareness as you can and one tip that we've been telling people is when you're doing your code reviews and everyone in this room is doing code reviews is you need to look at this and say okay imagine that somebody who has never seen this code before is woken or this infrastructure change is woken up what broke and then put the test in place to solve that so that when they get the alert they know what else is going on so in the end.

 DevOps does make you stronger or strong components but really start with your people look at your process and then finally look at your tools and while we're talking about tools there's one tool I have to recommend it's called page or duty we're easy to find on the expo floor come talk to us we'd love to hear what you're up to Thanks so we love coming to velocity right it's like great to be lost in a sea of best practices everybody in this audience believes in unit tests you believe in load testing load balancing continuous integration all the good stuff you're either doing it or you feel genuinely guilty that you're not doing it so there's nothing really I could tell you about technology or tooling and well automation is key DevOps is more than just automation so one of the things that we've been very fortunate at pager Duty is we've gotten to talk to thousands of the world's best Ops teams best dev teams best DevOps teams and more than a handful of teams that are genuinely struggling and so we've really gotten a chance to look at a lot of the people aspects of the opportunities and challenges facing teams and so one of the things we did to quantify those challenges as we sent out a survey and you know we did all the basic stuff we send it to people in various job titles in companies in various verticals and various sizes and we got a couple of hundred responses and one of the things that jumped out is despite the fact that a hundred percent I say effectively a hundred percent of companies lose money due to downtime we were finding that 85% of companies were honest and they said that they have missed a critical alert or they've been unaware of an outage affecting their their infrastructure and the good news for this room is that by and large.

 DevOps teams performed much better and this is teams that self identify as DevOps whatever that means to them we're finding that they're they're missed alerts were far lower there were about two-thirds of teams that were willing to admit that they had missed some kind of essential notification about their system they were responding faster typically within half an hour often within minutes and while stakeholders are never happy during an outage we were finding the devops teams had less unhappy stakeholders but and now now the scary slide we were finding that the average team was finding it was getting information from at least six monitoring tools right and this means a lot of the tools from the vendors on the exhibition floor today this means custom tools this means legacy tools tools in the middle of a migration and only a quarter of them were willing to say that they were confident that they were doing some kind of aggregation and then here's my scariest statistic of the teams that are trying to migrate from ops to DevOps about 80% of them were failing felt they were failing in some significant way and so we took a step back and we were wondering why is this right it doesn't make sense DevOps just seems like such a natural outcome in the sense of the only infrastructure the only place you make any money is production right and the only thing that makes you any more money typically is more functionality that looks a lot more like more code so you would expect that Devon Ops would be working hand in glove and that everybody would be you know paddling the boat in the same direction and we find this not often the case and so the number one thing we were finding that you didn't see right there's a lot of words on this slide but the one that I want to highlight is empathy right as a developer until you've truly been woken up at 3:00 in the morning putting out a fire that's not your fault in code that you didn't write that's not documented that doesn't have tests where you're losing money and a stakeholder is yelling at you you don't truly understand why there's so much process why there's so much hesitation to just let you upload your PHP script onto the server and test in production it's fun but it's it's scary when it doesn't work and it doesn't work and so one of the things that everybody in this audience knows to do is you build resilient systems right nothing has 100 percent uptime no matter what you tell yourself so you set a target let's say your target is fails one in a million times and your your up time is really the function of your subsystems so you've got one system that fails one in a thousand times well to hit a million one in a million you're gonna need to duplicate that you're gonna need to have an independent backup that fails one two thousand times 1000 times one two thousand one a million but you need to do the same thing as people like we genuinely believe that a person cannot be on call for a service no matter how good right a policy cannot be on call for a service you put something in the wiki that says you must reply in five minutes that's great your wiki can't actually get up and do anything so what you need to do is look at your team right let's say the average person in your team misses an alert one in a hundred times right they sleep through it they're out of service that's that's actually a pretty good number but again to get to that one in a million you need to have three levels in your escalation policy so that the the secondary and the tertiary who are called far less frequently but they they miss their alerts one in a hundred times as well 100 times 100 times 100 you're back to that magic one-in-a-million number that I just arbitrarily picked for you the other thing is I like doing I like working at 2:00 in the afternoon with a coffee in my hand and you know two monitors set up and I know what's going on and it's my code but that's not when stuff breaks stuff breaks when you're not prepared and you don't know what's going on and it's not your code so you want to not only aggregate your alerts so that you know what business system is affected how bad it is and what underlying systems are probably causing it but you want to provide as much situational awareness as you can and one tip that we've been telling people is when you're doing your code reviews and everyone in this room is doing code reviews is you need to look at this and say okay imagine that somebody who has never seen this code before is woken or this infrastructure change is woken up what broke and then put the test in place to solve that so that when they get the alert they know what else is going on so in the end .

DevOps does make you stronger or strong components but really start with your people look at your process and then finally look at your tools and while we're talking about tools there's one tool I have to recommend it's called page or duty we're easy to find on the expo floor come talk to us we'd love to hear what you're up to Thanks .

Visualpath is the greatest Institute For DevOps Training, Our Institute Providing quality DevOps Online Training education to the students In Hyderabad, Ameerpet,
India, United States. They Are Offering Real-Time Projects with Hands-On Experience and Providing 100% DevOps Placement Oriented Training