/ By communityteam@solarwinds.com / 0 Comments

Garry Schmidt first got involved in IT Service Management almost 20 years ago. Since becoming the manager of the IT Operations Center at SaskPower, ITSM has become one of his main focuses. Here are part 1 and part 2 of our conversation.

 

Bruno: What have been your biggest challenges with adopting ITSM and structuring it to fit SaskPower?

Garry: One of the biggest challenges is the limited resources available. Everybody is working hard to take care of their area of responsibility. Often you introduce new things, like pushing people to invest time in problem management, for example. The grid is full. It’s always a matter of trying to get the right priority. There are so many demands on everybody all day long that even though you think investing time and improving the disciplines, you still have to figure out how to convince people the priority should be placed there. So, the cultural change aspect, the organizational change is always the difficult part. “We’ve always done it this way, and it’s always worked fine. So, what are you talking about Schmidt?”

 

Taking one step at a time and having a plan of where you want to get to. Taking those bite-sized pieces and dealing with it that way. You just can’t get approval to add a whole bunch of resources to do anything. It’s a matter of molding how we do things to shift towards the ideal instead of making the big steps.

 

It’s more of an evolution than a revolution. Mind you a big tool change or something similar gives you a platform to be able to do a fair amount of the revolution at the same time. You’ve got enough funding and dedicated resources to be able to focus on it. Most of the time, you’re not going to have that sort of thing to leverage.

 

Bruno: You’ve mentioned automation a few times in addition to problem resolution and being more predictive and proactive. When you say automation, what else are you specifically talking about?

Garry: Even things like chatbots need to be able to respond to requests and service desk contacts. I think there’s more and more capability available. The real challenge I’ve seen with AI tools is it’s hard to differentiate between those that just have a new marketing spin on an old tool versus the ones with some substance to them. And they’re expensive.

 

We need to find some automation capability to leverage the tools we’ve already got, or it’s an incremental investment rather than a wholesale replacement. The enterprise monitoring and alerting experience, I’m not willing to go down the path of replacing all our monitoring tools and bringing everything into a big, jumbo AI engine again. I’m skittish of that kind of stuff.

 

Our typical pattern has been we buy a big complicated tool, implement, and then use only a tenth of the capability. And then we go look for another tool.

 

Bruno: What recommendations would you make to someone who is about to introduce ITSM to an organization?

Garry: Don’t underestimate the amount of time it takes to handle the organizational change part of it.

 

You can’t just put together job aides and deliver training for a major change and then expect it to just catch on and go. It takes constant tending to make it grow.

 

It’s human nature: you’re not going to get everything the first time you see new processes. So, having a solid support structure in place to be able to continually coach and even evolve the things you do. We’ve changed and improved upon lots of things based on the feedback we’ve gotten from the folks within the different operational teams. Our general approach was to try and get the input and involvement of the operational teams as much as we could. But there were cases where we had to make decisions on how it was going to go and then teach people about it later. In both cases, you get smarter as you go forward.

 

This group operates a little differently than all the rest, and there are valid reasons for it, so we need to change our processes and our tools to make it work better for them. You need to make it work for them.

 

Have the mentality that our job is to make them successful.

 

You just need to have an integrated ITSM solution. We were able to make progress without one, but we hit the ceiling.

 

Bruno: Any parting words on ITSM, or your role, or the future of ITSM as you see it?

Garry: I really like the role my team and I have. Being able to influence how we do things. Being able to help the operational teams be more successful while also improving the results we provide to our users, to our customers. I’m happy with the progress we’ve made, especially over the last couple of years. Being able to dedicate time towards reviewing and improving our processes and hooking it together with the tools.

 

I think we’re going to continue to make more good progress as we further automate and evolve the way we’re doing things.

/ By communityteam@solarwinds.com / 0 Comments

In a recent episode of a popular tech podcast, I heard the hosts say, “If you can’t automate, you and your career are going to get left behind.” Some form of this phrase has been uttered on nearly every tech podcast over the last few years. It’s a bit of hyperbole, but has some obvious truth to it. IT isn’t slowing down. It’s going faster and faster. How can you possibly keep up? By automating all the things.

 

It sounds great in principle, but what if you don’t have any experience with automation, coding, or scripting? Where do you get started? Here are three things you can do to start automating your career.

 

1. Pick-Tool

As an IT practitioner, your needs are going to be different from the people in your development org, and different needs (may) require different tools. Start by asking those around you who are already writing code. If a Dev team in your organization is leveraging Ruby in Jenkins, it makes sense to learn Ruby over something like Java. By taking this approach, there are a couple of benefits to you. First, you’re aligning with your organization. Secondly and arguably most importantly, you now have built-in resources to help you learn. It’s always refreshing when people are willing to cross the aisle to help each other become more effective. I mean, isn’t that ultimately what the whole DevOps movement is about—breaking down silos? Even if you don’t make a personal connection, by using the tools your organization is already leveraging, you’ll have code examples to study and maybe even libraries to work from.

 

What if you’re the one trying to introduce automation to your org or there are no inherent preferences built in? Well, you have a plethora of automation tools and languages to choose from. So, how do you decide? Look around your organization. Where do you run your apps? What are your operating systems? Cloud, hybrid, or on-premises, where do you run your infrastructure? These clues can help you. If your shop runs Windows or VMware, PowerShell would be a safe bet as an effective tool to start automating operations. Are you in information security or are you looking to leverage a deployment tool like Ansible? Then Python might be a better choice.

 

2. Choose-Task

Time after time, talk after talk, the most effective advice I’ve seen for how to get started automating is: pick something. Seriously, look around and pick a project. Got a repetitious task with any sort of regularity? Do you have any tasks frequently introducing human error to your environment, where scripting might help standardize? What monotony drives you crazy, and you just want to automate it out of your life? Whatever the case may be, by picking an actionable task, you can start learning your new scripting skill while doing your job. Also, the first step is often the hardest one. If you pick a task and set it as a goal, you’re much more likely to succeed vs. some sort of nebulous “someday I want to learn x.”

 

3. Invoke-Community

The people in your organization can be a great resource. If you can’t get help inside your org for whatever reason, never fear—there are lots of great resources available to you. It would be impossible to highlight the community and resources for each framework, but communities are vital to the vibrancy of a language. Communities take many forms, including user groups, meetups, forums, and conferences. The people out there sharing in the community are there because they’re passionate, and they want to share. These people and communities want to be there for you. Use them!

 

First Things Last

When I was first getting started in my career, a mentor told me, “When I’m hiring for a position and I see scripting (automation) in their experience, I take that resume and I move it to the top of my stack.” Being the young, curious, and dumb kid I was, I asked “Why?” In hindsight, it appears his answer was somewhat prophetical in talking about the DevOps movement to come: “If you can script, you can make things more efficient. If you make things more efficient, you can help more people. If you can help more people, you’re more valuable to the organization.”

 

The hardest and most important step you’ll make is to get started. If reading this article is part of your journey to automating your career, I hope you’ve found it helpful.

 

PS: If you’re still struggling to pick a language/framework, in my next post I’ll offer up my opinions on a “POssible SHortcut" to start automating effectively right away.

/ By communityteam@solarwinds.com / 0 Comments

In a recent episode of a popular tech podcast, I heard the hosts say, “If you can’t automate, you and your career are going to get left behind.” Some form of this phrase has been uttered on nearly every tech podcast over the last few years. It’s a bit of hyperbole, but has some obvious truth to it. IT isn’t slowing down. It’s going faster and faster. How can you possibly keep up? By automating all the things.

 

It sounds great in principle, but what if you don’t have any experience with automation, coding, or scripting? Where do you get started? Here are three things you can do to start automating your career.

 

1. Pick-Tool

As an IT practitioner, your needs are going to be different from the people in your development org, and different needs (may) require different tools. Start by asking those around you who are already writing code. If a Dev team in your organization is leveraging Ruby in Jenkins, it makes sense to learn Ruby over something like Java. By taking this approach, there are a couple of benefits to you. First, you’re aligning with your organization. Secondly and arguably most importantly, you now have built-in resources to help you learn. It’s always refreshing when people are willing to cross the aisle to help each other become more effective. I mean, isn’t that ultimately what the whole DevOps movement is about—breaking down silos? Even if you don’t make a personal connection, by using the tools your organization is already leveraging, you’ll have code examples to study and maybe even libraries to work from.

 

What if you’re the one trying to introduce automation to your org or there are no inherent preferences built in? Well, you have a plethora of automation tools and languages to choose from. So, how do you decide? Look around your organization. Where do you run your apps? What are your operating systems? Cloud, hybrid, or on-premises, where do you run your infrastructure? These clues can help you. If your shop runs Windows or VMware, PowerShell would be a safe bet as an effective tool to start automating operations. Are you in information security or are you looking to leverage a deployment tool like Ansible? Then Python might be a better choice.

 

2. Choose-Task

Time after time, talk after talk, the most effective advice I’ve seen for how to get started automating is: pick something. Seriously, look around and pick a project. Got a repetitious task with any sort of regularity? Do you have any tasks frequently introducing human error to your environment, where scripting might help standardize? What monotony drives you crazy, and you just want to automate it out of your life? Whatever the case may be, by picking an actionable task, you can start learning your new scripting skill while doing your job. Also, the first step is often the hardest one. If you pick a task and set it as a goal, you’re much more likely to succeed vs. some sort of nebulous “someday I want to learn x.”

 

3. Invoke-Community

The people in your organization can be a great resource. If you can’t get help inside your org for whatever reason, never fear—there are lots of great resources available to you. It would be impossible to highlight the community and resources for each framework, but communities are vital to the vibrancy of a language. Communities take many forms, including user groups, meetups, forums, and conferences. The people out there sharing in the community are there because they’re passionate, and they want to share. These people and communities want to be there for you. Use them!

 

First Things Last

When I was first getting started in my career, a mentor told me, “When I’m hiring for a position and I see scripting (automation) in their experience, I take that resume and I move it to the top of my stack.” Being the young, curious, and dumb kid I was, I asked “Why?” In hindsight, it appears his answer was somewhat prophetical in talking about the DevOps movement to come: “If you can script, you can make things more efficient. If you make things more efficient, you can help more people. If you can help more people, you’re more valuable to the organization.”

 

The hardest and most important step you’ll make is to get started. If reading this article is part of your journey to automating your career, I hope you’ve found it helpful.

 

PS: If you’re still struggling to pick a language/framework, in my next post I’ll offer up my opinions on a “POssible SHortcut" to start automating effectively right away.

/ By leon.adato@solarwinds.com / 0 Comments

The 2019 Writing Challenge got off to an amazing start and I’m grateful to everyone who contributed their time, energy, and talent both as the lead writers and commenters. The summary below offers up just a sample of the amazing and insightful ways in which IT pros break down difficult concepts and relate them—not just to five-year-olds, but to folks of any age who need to understand something simply and clearly.

 

Day 1: Monitoring—Leon Adato

I had the privilege of kicking off the challenge this year, and I felt there was no word more appropriate to do so with than “monitoring”

 

Jeremy Mayfield  Dec 1, 2019 8:27 AM

 

Great way to start the month. I was able to understand it. You spoke to my inner child, or maybe just me as I am now...... Being monitored is like when the kids are at Grandma’s house playing in the yard, and she pretends to be doing dishes watching everything out the kitchen window.

 

Rick Schroeder  Dec 2, 2019 2:30 AM

Are there cases when it’s better NOT to know? When might one NOT monitor and thereby provide an improvement?

 

I’m not talking about not over-monitoring, nor about monitoring unactionable items, nor alerting inappropriately.

 

Sometimes standing something on its head can provide new insight, new perspective, that can move one towards success.

 

Being able to monitor doesn’t necessarily mean one should monitor—or does it?

 

When is it “good” to not know the current conditions? Or is there ever a time for that, assuming one has not over-monitored?

 

Mathew Plunkett Dec 2, 2019 8:35 AM

rschroeder asked a question I have been thinking about since THWACKcamp. It started with the question “Am I monitoring elements just because I can or is it providing a useful metric?” The answer is that I was monitoring some elements just because it was available and those were removed. The next step was to ask “Am I monitoring something I shouldn’t?” This question started with looking for monitored elements that were not under contract but evolved into an interesting thought experiment. Are there situations in which we should not be monitoring an element? I have yet to come up with a scenario in which this is the case, but it has helped me to look at monitoring from a different perspective.

 

Day 2: Latency –Thomas LaRock

Tom’s style, and his wry wit, is on full display in this post, where he shows he can explain a technical concept not only to five-year-olds, but to preteens as well.

 

Thomas Iannelli  Dec 2, 2019 12:02 PM

In graduate school we had an exercise in our technical writing class where we took the definition and started replacing words with their definitions. This can make things simpler or it can cause quite a bit of latency in transferring thoughts to your reader.

 

Latency -

  • The delay before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something is late or postponed before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first scheduled before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving data to another place begins following an instruction for its moving of data to another place.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place begins following an instruction for its moving of the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place begins following a code or sequence in a computer program that defines an operation and puts it into effect for its moving of the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place.

 

.....and so on

 

Juan Bourn Dec 2, 2019 1:39 PM

I think I am going to enjoy these discussions this month. I have a hard time explaining things without using technical terms sometimes. Not because I don’t understand them (i.e., Einstein’s comment), but because I sometimes think only in technical terms. It’s honestly what I understand easiest. For me, latency is usually associated as a negative concept. It’s refreshing to hear it discussed in general terms, as in there’s latency in everything. Like many things in IT, it’s usually only brought to light or talked about if there’s a problem with it. So latency gets a bad rep. But it’s everywhere, in everything.

 

Jake Muszynski  Dec 2, 2019 10:42 AM

Hold on, I will reply to this when I get a chance.

 

Day 3: Metrics–Sascha Giese

One of my fellow Head Geek’s passions is food. Of course he uses this context to explain something simply.

 

Mark Roberts  Dec 4, 2019 9:49 AM

The most important fact in the first line is that to make a dough that will perform well for a pizza base a known amount of flour is necessary. This is the baseline, 1 pizza = 3.5 cups. If you needed to make 25 pizzas you now know how to determine how much flour you need 25 x 3.5 = A LOT OF PIZZA

 

Dale Fanning Dec 4, 2019 11:54 AM

Why metrics are important—those who fail to learn from history are doomed to repeat it. How can you possibly know what you need to be able to do in the future if you don’t know what you’ve done in the past?

 

Ravi Khanchandani  Dec 4, 2019 8:08 AM

Are these my School Grades

Metrics are like your Report cards—giving you grades for the past, present & future (predictive grades).

Compare the present ratings with your past and also maybe the future

Different subjects rated and measured according to the topics in the subjects

 

Day 4: NetFlow—Joe Reves

What is remarkable about the Day 4 entry is not Joe’s mastery of everything having to do with NetFlow, it’s how he encouraged everyone who commented to help contribute to the growing body of work known as “NetFlowetry.”

 

Dale Fanning Dec 5, 2019 9:27 AM

I think the hardest thing to explain about NetFlow is that all it does is tell you who has been talking to who (whom? I always forget), or not, as the case may be, and *not* what was actually said. Sadly when you explain that they don’t understand that it’s still quite useful to know and can help identify where you may need to look more deeply. If it was actual packet capture as something you’d be buried in data in seconds.

 

Farhood Nishat Dec 5, 2019 8:43 AM

They say go with the flow

but how can we get to know what is the current flow

for that we pray to god to lead us towards the correct flow

but when it comes to networks and tech

we use the netflow to get into that flow

cause a flow can be misleading

and we cant just go with the flow

 

George Sutherland Dec 4, 2019 12:23 PM

NetFlow is like watching the tides. The EBB and flow, the high and low.

 

External events such as the moon phases and storms in tides are replaced by application interactions, data transfers, bandwidth contention and so on.

 

Know what is happening is great, but the real skill is creating methods that deal with the anomalies as they occur.

 

Just another example of why our work is never boring!

 

Day 5: Logging–Mario Gomez

Mario is one of our top engineers and every day, he finds himself explaining technically complex ideas to customers of all stripes. This post shows he’s able to do it with humor as well.

 

Mike Ashton-Moore  Dec 6, 2019 10:08 AM

I always loved the Star Trek logging model.

If the plot needed it then the logs had all the excruciating detail needed to answer the question.

But security was so lax (what was Lt Worf doing all this time?)

So if the plot needed it, Lt Worf and his security detail were on vacation and the logs only contained no useful information.

 

However, the common thread was that logs only ever contained what happened, never why.

 

Michael Perkins Dec 5, 2019 5:14 PM

What’s Logging? Paul Bunyan and Babe the Blue Ox, lumberjacks and sawmills, but that’s not important now.

 

What do we do with the heaps of logs generated by all the devices and servers on our networks? So much data. What do we need to log to confirm attribution, show performance, check for anomalies, etc., and what can we let go? How do we balance keeping logs around long enough to be helpful (security and performance analyses) with not allowing them to occupy too much space or make our tools slow to unusability?

 

George Sutherland Dec 5, 2019 3:18 PM

In the land of internal audit

The edict came down to record it

 

Fast or slow good or bad

It was the information that was had

 

Some reason we knew most we did not

We collected in a folder, most times to rot

 

The volume was large, it grew and grew

Sometimes to exclusion of everything new

 

Aggregation was needed and to get some quick wins

Thank heavens we have SolarWinds

 

Day6: Observability—Zack Mutchler (MVP)

THWACK MVP Zack Mutchler delivers a one-two punch for this post—offering an ELI5 appropriate explanation but then diving deep into the details as well, for those who craved a bit more.

 

Holly Baxley Dec 6, 2019 10:36 AM

To me—monitoring and observability can seem like they do the same thing, but they’re not.

 

Monitoring -

“What’s happening?”
Observability -

“Why is this happening?”

“Should this be happening?”

“How can we stop this from happening?”

“How can we make this happen?”

 

The question is this...can we build an intelligent AI that can actually predict behavior and get to the real need behind the behavior, so we can stop chasing rabbits and having our customers say, “It’s what I asked for, but it’s not what I want.”

 

If we can do that—then we’ll have mastered observability.

 

Mike Ashton-Moore  Dec 6, 2019 10:15 AM

so—alerting on what matters, but monitor as much as you’re able—and don’t collect a metric just because it’s easy, collect it because it matters

 

Juan Bourn Dec 6, 2019 9:16 AM

So observability is only tangible from an experience stand point (what is seen by observing its behavior)? Or will there always be metrics (like Disney+ not loading)? If there are always metrics, then are observability and metrics two sides of the same coin?

 

/ By leon.adato@solarwinds.com / 0 Comments

The 2019 Writing Challenge got off to an amazing start and I’m grateful to everyone who contributed their time, energy, and talent both as the lead writers and commenters. The summary below offers up just a sample of the amazing and insightful ways in which IT pros break down difficult concepts and relate them—not just to five-year-olds, but to folks of any age who need to understand something simply and clearly.

 

Day 1: Monitoring—Leon Adato

I had the privilege of kicking off the challenge this year, and I felt there was no word more appropriate to do so with than “monitoring”

 

Jeremy Mayfield  Dec 1, 2019 8:27 AM

 

Great way to start the month. I was able to understand it. You spoke to my inner child, or maybe just me as I am now...... Being monitored is like when the kids are at Grandma’s house playing in the yard, and she pretends to be doing dishes watching everything out the kitchen window.

 

Rick Schroeder  Dec 2, 2019 2:30 AM

Are there cases when it’s better NOT to know? When might one NOT monitor and thereby provide an improvement?

 

I’m not talking about not over-monitoring, nor about monitoring unactionable items, nor alerting inappropriately.

 

Sometimes standing something on its head can provide new insight, new perspective, that can move one towards success.

 

Being able to monitor doesn’t necessarily mean one should monitor—or does it?

 

When is it “good” to not know the current conditions? Or is there ever a time for that, assuming one has not over-monitored?

 

Mathew Plunkett Dec 2, 2019 8:35 AM

rschroeder asked a question I have been thinking about since THWACKcamp. It started with the question “Am I monitoring elements just because I can or is it providing a useful metric?” The answer is that I was monitoring some elements just because it was available and those were removed. The next step was to ask “Am I monitoring something I shouldn’t?” This question started with looking for monitored elements that were not under contract but evolved into an interesting thought experiment. Are there situations in which we should not be monitoring an element? I have yet to come up with a scenario in which this is the case, but it has helped me to look at monitoring from a different perspective.

 

Day 2: Latency –Thomas LaRock

Tom’s style, and his wry wit, is on full display in this post, where he shows he can explain a technical concept not only to five-year-olds, but to preteens as well.

 

Thomas Iannelli  Dec 2, 2019 12:02 PM

In graduate school we had an exercise in our technical writing class where we took the definition and started replacing words with their definitions. This can make things simpler or it can cause quite a bit of latency in transferring thoughts to your reader.

 

Latency -

  • The delay before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something is late or postponed before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first scheduled before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before a transfer of data begins following an instruction for its transfer.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving data to another place begins following an instruction for its moving of data to another place.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place begins following an instruction for its moving of the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place.
  • The period of time by which something causes or arranges for something to take place at a time later than that first arranged or planned to take place at a particular time before an act of moving the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place begins following a code or sequence in a computer program that defines an operation and puts it into effect for its moving of the quantities, characters, or symbols on which operations are performed by a computer, being stored and transmitted in the form of electrical signals and recorded on magnetic, optical, or mechanical recording media to another place.

 

.....and so on

 

Juan Bourn Dec 2, 2019 1:39 PM

I think I am going to enjoy these discussions this month. I have a hard time explaining things without using technical terms sometimes. Not because I don’t understand them (i.e., Einstein’s comment), but because I sometimes think only in technical terms. It’s honestly what I understand easiest. For me, latency is usually associated as a negative concept. It’s refreshing to hear it discussed in general terms, as in there’s latency in everything. Like many things in IT, it’s usually only brought to light or talked about if there’s a problem with it. So latency gets a bad rep. But it’s everywhere, in everything.

 

Jake Muszynski  Dec 2, 2019 10:42 AM

Hold on, I will reply to this when I get a chance.

 

Day 3: Metrics–Sascha Giese

One of my fellow Head Geek’s passions is food. Of course he uses this context to explain something simply.

 

Mark Roberts  Dec 4, 2019 9:49 AM

The most important fact in the first line is that to make a dough that will perform well for a pizza base a known amount of flour is necessary. This is the baseline, 1 pizza = 3.5 cups. If you needed to make 25 pizzas you now know how to determine how much flour you need 25 x 3.5 = A LOT OF PIZZA

 

Dale Fanning Dec 4, 2019 11:54 AM

Why metrics are important—those who fail to learn from history are doomed to repeat it. How can you possibly know what you need to be able to do in the future if you don’t know what you’ve done in the past?

 

Ravi Khanchandani  Dec 4, 2019 8:08 AM

Are these my School Grades

Metrics are like your Report cards—giving you grades for the past, present & future (predictive grades).

Compare the present ratings with your past and also maybe the future

Different subjects rated and measured according to the topics in the subjects

 

Day 4: NetFlow—Joe Reves

What is remarkable about the Day 4 entry is not Joe’s mastery of everything having to do with NetFlow, it’s how he encouraged everyone who commented to help contribute to the growing body of work known as “NetFlowetry.”

 

Dale Fanning Dec 5, 2019 9:27 AM

I think the hardest thing to explain about NetFlow is that all it does is tell you who has been talking to who (whom? I always forget), or not, as the case may be, and *not* what was actually said. Sadly when you explain that they don’t understand that it’s still quite useful to know and can help identify where you may need to look more deeply. If it was actual packet capture as something you’d be buried in data in seconds.

 

Farhood Nishat Dec 5, 2019 8:43 AM

They say go with the flow

but how can we get to know what is the current flow

for that we pray to god to lead us towards the correct flow

but when it comes to networks and tech

we use the netflow to get into that flow

cause a flow can be misleading

and we cant just go with the flow

 

George Sutherland Dec 4, 2019 12:23 PM

NetFlow is like watching the tides. The EBB and flow, the high and low.

 

External events such as the moon phases and storms in tides are replaced by application interactions, data transfers, bandwidth contention and so on.

 

Know what is happening is great, but the real skill is creating methods that deal with the anomalies as they occur.

 

Just another example of why our work is never boring!

 

Day 5: Logging–Mario Gomez

Mario is one of our top engineers and every day, he finds himself explaining technically complex ideas to customers of all stripes. This post shows he’s able to do it with humor as well.

 

Mike Ashton-Moore  Dec 6, 2019 10:08 AM

I always loved the Star Trek logging model.

If the plot needed it then the logs had all the excruciating detail needed to answer the question.

But security was so lax (what was Lt Worf doing all this time?)

So if the plot needed it, Lt Worf and his security detail were on vacation and the logs only contained no useful information.

 

However, the common thread was that logs only ever contained what happened, never why.

 

Michael Perkins Dec 5, 2019 5:14 PM

What’s Logging? Paul Bunyan and Babe the Blue Ox, lumberjacks and sawmills, but that’s not important now.

 

What do we do with the heaps of logs generated by all the devices and servers on our networks? So much data. What do we need to log to confirm attribution, show performance, check for anomalies, etc., and what can we let go? How do we balance keeping logs around long enough to be helpful (security and performance analyses) with not allowing them to occupy too much space or make our tools slow to unusability?

 

George Sutherland Dec 5, 2019 3:18 PM

In the land of internal audit

The edict came down to record it

 

Fast or slow good or bad

It was the information that was had

 

Some reason we knew most we did not

We collected in a folder, most times to rot

 

The volume was large, it grew and grew

Sometimes to exclusion of everything new

 

Aggregation was needed and to get some quick wins

Thank heavens we have SolarWinds

 

Day6: Observability—Zack Mutchler (MVP)

THWACK MVP Zack Mutchler delivers a one-two punch for this post—offering an ELI5 appropriate explanation but then diving deep into the details as well, for those who craved a bit more.

 

Holly Baxley Dec 6, 2019 10:36 AM

To me—monitoring and observability can seem like they do the same thing, but they’re not.

 

Monitoring -

“What’s happening?”
Observability -

“Why is this happening?”

“Should this be happening?”

“How can we stop this from happening?”

“How can we make this happen?”

 

The question is this...can we build an intelligent AI that can actually predict behavior and get to the real need behind the behavior, so we can stop chasing rabbits and having our customers say, “It’s what I asked for, but it’s not what I want.”

 

If we can do that—then we’ll have mastered observability.

 

Mike Ashton-Moore  Dec 6, 2019 10:15 AM

so—alerting on what matters, but monitor as much as you’re able—and don’t collect a metric just because it’s easy, collect it because it matters

 

Juan Bourn Dec 6, 2019 9:16 AM

So observability is only tangible from an experience stand point (what is seen by observing its behavior)? Or will there always be metrics (like Disney+ not loading)? If there are always metrics, then are observability and metrics two sides of the same coin?

 

/ By PCMag.com Latest Articles / 0 Comments
'With no incentive to innovate or invest, these conglomerates charge sky-high internet prices to reap profits from consumers,' the Democratic Presidential candidate said in unveiling his 'high-speed internet access for all plan' on Friday.
/ By David Linthicum / 0 Comments

Are you the cloud security person, or cloudsec/cloudops, as it’s often called? The person who controls the cloud deployment, including all security such as IAM (identity and access management), encryption services, and application and user authentication? I have some observations about cloud security to share.

First, enterprises typically don’t have a good-fitting security solution. For most of the security solutions I’ve seen in the cloud, it appears that companies have turned on native public cloud security, done some iterations, and called it a day.

Second, security isn’t seen as something to be continuously improved through technology. Indeed, once a cloud system is secured, enterprise IT is reluctant to change or upgrade security. That by itself is a risk. However, by continuously improving your security systems, using best practices, tools, and yes, easy hacks, you’ll be at a much lower risk for a breach.

To read this article in full, please click here

/ By David Linthicum / 0 Comments

Are you the cloud security person, or cloudsec/cloudops, as it’s often called? The person who controls the cloud deployment, including all security such as IAM (identity and access management), encryption services, and application and user authentication? I have some observations about cloud security to share.

First, enterprises typically don’t have a good-fitting security solution. For most of the security solutions I’ve seen in the cloud, it appears that companies have turned on native public cloud security, done some iterations, and called it a day.

Second, security isn’t seen as something to be continuously improved through technology. Indeed, once a cloud system is secured, enterprise IT is reluctant to change or upgrade security. That by itself is a risk. However, by continuously improving your security systems, using best practices, tools, and yes, easy hacks, you’ll be at a much lower risk for a breach.

To read this article in full, please click here

/ By Paul Krill / 0 Comments

Microsoft has found that approximately 70 percent of the security vulnerabilities it addresses are due to memory safety issues. To make it easier to write safer code, the company is developing a language designed for safe infrastructure programming. 

Due to be open-sourced soon, the first version of the new language, called Project Verona, incorporates three core ideas:

  • Data-race freedom, which gives up concurrent, arbitrary mutation to enable scalable memory management with temporal safety without global synchronization.
  • Concurrent owners, which provides a new concurrency model that offers lightweight, asynchronous coordination of resources.
  • Linear regions, with the ownership model based on groups of objects. This differs from the memory-safe Rust language, which is based on a single object, Microsoft explained. In Verona, there are memory management strategies per region. Also featured is compartmentalization for legacy components.

Verona explores compartmentalization at the language design level. With Verona, there are threads that can access regions. Any region can be accessed by only one thread. There is a linear entry point into a region, and regions can be nested within other regions. A shared immutable region can maintain items not being mutated.

To read this article in full, please click here

/ By Scott Carey / 0 Comments

Amazon Web Services has expanded the capabilities of its Amazon SageMaker machine learning toolkit to address a number of challenges that enterprises confront when trying to operationalize machine learning, from model organization, training, and optimization to monitoring the performance of models in production.

Launched at the Amazon’s re:invent conference in 2017, SageMaker aims to make machine learning adoption simpler for customers by bringing together a hosted environment for Jupyter notebooks with built-in model management, automated spin up of training environments in Amazon S3, and HTTPS endpoints for hosting capabilities using EC2 instances.

To read this article in full, please click here