If you think there are a lot of tools to configure your systems you haven't looked at the tools available to monitor your stuff.  The set is so large that it is easy to get overwhelmed.  So again in this article I am going to give you a list that I use to narrow the field.  Then I am going to give you a list of my favorites.  

Is it agent, agentless, or hybrid? As with most configuration management tools this question cuts both ways. The best in this class of tools with agents have well documented deployment paths that use various configuration management tools. For instance, they will have Chef or Puppet packages that cut down your time to deploy them tremendously. The choice on this question is how much time you have to deploy it and how fast a response do you need from the tool. Agent based tools are faster in most cases. Agentless tools rely on some form of remote execution tool like ssh or remote powershell and an SNMP(Simple Network Monitoring Protocol) agent. Because the server in an agentless system has to do all the work polling they tend to have more complex to scale. They also can require more risks to take because you have to allow more ports through your firewall. Hybrids allow you to deploy the tool in different ways dependent on the security requirements. So for medium and large companies they tend be a better choice.
How does reporting work? This is what you need the tool to do so paying attention to it is critical. The tools vary widely with the number and type of standard reports they have. They also vary on how easy it is to do custom reports. In more and more cases monitoring tools are pared with a reporting system to handle this issue. Writing custom reports can be as simple as a gui interface. They can be as difficult as a DSL(Domain Specific Language) or traditional programming language to create the reports. If your business needs reports from your systems be sure to confirm that you can create the reports that your business needs to meet it's needs. For instance, can you easily get a report to tell you how many people failed to sign up for your mailing list? Can you tell where people are stopping or failing to complete an action.
Does it do alerting and if so how can you be alerted? Alerting sounds like a no brainer but not all tools do it. Some tools are just built to display a set of stats for people to analyze. Which means they are normally easier to deploy and configure. At the other end of the scale are tools that will try to predict failures and alert you before the problem happens. This sounds great, and is cool, but you need to know a lot about your environment so that you can set the boundaries around good and bad events. That means that it will require a lot of time to tune properly to remove false positives. Also can an alert trigger an action? If it can then you can automate simple things and free a human to sleep or do something more productive. If it can tell you that the disk is filling up can it do the steps your team would do to free up space? This will help a lot with the work life balance.
What dashboards are available out of the box? Can you customize them simply? Most of the tools come with a set of what are called canned dashboards. When you are starting off with monitoring tools it's best to choose a tool with as more than less as long as they tell you something. If the tool has a great dashboard for monitoring Java Applications but your company writes it's apps in Ruby then what good is it to you? All of them will let you customize these dashboards. You can roll up the stats so you can show the entire set of a stat in an environment(Development, Test, Production) in one chart. Be careful though as Techs we love our data and chart customization can get out of hand. Over time you will want to add custom dashboards to make it easier to troubleshoot your devices.
How resource intensive is it on both the server, network, and clients being monitored? This is another one of those it depends discussion. If you are only monitoring a small number of things(computer, network equipment, etc) then this is less of a concern. You always need to be concerned about this because your first instinct is going to be to monitor everything. We can both from a device and data point perspective. I have seen, and caused, situations where we have monitored ourselves to death. For instance, at one company where we had limited bandwidth to our remote sites, we overwhelmed the network with just monitoring traffic. There was no bandwidth left for little things like file transfers and pulling up the company intranet. The problem with this is often until you do a proof of concept with the tool you may not be able to answer it. As a general rule agent based tools can help a lot because they only need to send changes and not everything. In all cases though you need to be sure that you are going to get something from the data you collect. It also makes it harder to filter when the time comes to create dashboards and reports.
The best way to handle most of these issues is to define a set of things you know you need to monitor, things you think you want to monitor, and things you know you don't need to monitor.  Then apply that list to the questions above.  It's a simple flow of can I get the data, can I report on it, can I make a dashboard for it and finally will we have enough resources for all of it. 

The problem I have with this set of tools is that they all have a high level of complexity during the implementation phase. Even the simplest of them can talk several person weeks to get setup correctly and start returning on the investment.  Once you do have it setup you will be amazed at how much it will help you become more efficient.  

You should also be open to deploying multiple tools in this class.  Monitoring a multi-tiered applications completely may seem easy at first but it is difficult to do accurately.   Keep in mind that monitoring anything is a complex process.  It is not uncommon for companies to deploy two or three tools to meet all of their monitoring needs at a complexity level that makes sense for their company.  You may have one tool for monitoring the basic information like disk space and cpu utilization, another to monitor application health, and a third to monitor user behavior.

Ok so what are my recommendations?
Here are a few of my favorites.
Nagios(http://www.nagios.org/) This is where a lot of the following tools started.  It is a great tool but as the number of your systems starts to increase configuration can be difficult to manage effectively.  Which is why the next two started trying to do.

Zenoss(http://www.zenoss.com/documents/datasheet_core_commercial_compare.pdf) - This a freemium modeled application with a community version that does the basics well.  The Commercial version adds more analysis and optimization information. Check the linked PDF for a more detailed explanation.

Groundwork (http://www.gwos.com/pricing/enterprise/) This tool takes a lot of the hard edges off of Nagios.  They continue to add features and let you monitor your first 50 hosts for free with their enterprise tool.

We will be adding a link in the coming months to the site listing as many as we can find and giving our opinion.  These are a great starting point but before you make a decision look at other and make sure they don't better serve your monitoring needs.  

Our sister site has a review of Zenoss and Groundwork that are worth a look even if they are somewhat old at this point.

Devops Mastery - Episode 15 Monitoring DevOps Tools

Safar Baki Hai. Literally means “The journey isn’t over, yet.” This beautiful music video is probably the first of its kind in the world. Featuring over 20 countries, more than 25 cities, shot by 20 different cinematographers, all packed into one 4 and a half minute song. The concept was simple. The composer and the lyricist did a song on how life itself is a journey. Then they appealed to their friends and friends of friends, who were travelling allover the world, to shoot a part of the video and be a part of a one-of-its-kind project. They got an overwhelming response, getting video clips from more than 53 countries and 80 cities allover the world. Right from the beaches in America, to the Himalayas, from the Eiffel Tower to the Football Stadium in Brazil, from the shores of Barcelona, to the Cablecars on Mt. Titlis. They had only one condition. The videos must be beautiful. And the videos were just that. The tough part was to select just 25 clips out of these 80. You will be astonished by the end result. Because these aren’t just video clips, they are personal journeys. They are a peek into personal moments of discovering a new place, discovering a new way of life. Watch it for yourself. Plus, the song is available for free download at www.gandhaar.weebly.com .

Safar Baki Hai

This week's podcasts features Yao Yue of Twitter. Yao spent the majority of her career working on caching systems at Twitter. She has since created a performance team that deals with edge performance outliers often exposed by the enormous scale of Twitter. In this podcast, she discusses standing up the performance team, thoughts on instrumenting applications, and interesting performance issues (and strategies for solving them) they’ve seen at Twitter.

Why listen to this podcast:

* Performance problems can be caused by a few machines running slowly causing cascading failure
* Aggregating stats on a minute-by-minute basis can be an effective way of monitoring thousands of servers
* Being able to record second-by-second is often too expensive to centrally aggregate, but can be stored locally
* Distinguishing between request timeout and connection/network timeouts is important to prevent thundering herds
* With larger scale organisations, having dedicated performance teams helps centralise skills to solve performance problems

More on this: Quick scan our curated show notes on InfoQ http://bit.ly/2wnBemB

You can also subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq

Subscribe: www.youtube.com/infoq
Like InfoQ on Facebook: bit.ly/2jmlyG8
Follow on Twitter: twitter.com/InfoQ
Follow on LinkedIn: www.linkedin.com/company/infoq
Want to see extented shownotes? Check the landing page on InfoQ: http://bit.ly/2wnBemB

Show Notes

Twitter's Yao Yue on Latency, Performance Monitoring, & Caching at Scale

In this episode of the O’Reilly Programming Podcast, I talk with Matt Stine, global CTO of architecture at Pivotal. He is the presenter of the O’Reilly live online training course "Cloud-Native Architecture Patterns," and he has spoken about cloud-native architecture at the recent O’Reilly Software Architecture Conference and O’Reilly Security Conference.

Matt Stine on Cloud-Native Architecture

On this week’s podcast, Danny Yuan, Uber’s Real-time Streaming/Forecasting Lead, lays out a thorough recipe book for building a real-time streaming platform with a major focus on forecasting. In this podcast, Danny discusses everything from the scale Uber operates at to what the major steps for training/deploy models in an iterative (almost Darwinistic) fashion and wraps with his advice for software engineers who want to begin applying machine learning into their day-to-day job.  

Why listen to this podcast:

* Uber processes 850,000 - 1.3 million messages per second in their streaming platform with about 12 TB of growth per day. The system’s queries scan 100 million to 4 billion documents per second.
* Uber’s frontend is mobile. The frontend talks to an API layer. All services generate events that are shuffled into Kafka. The real-time forecasting pipeline taps into Kafka to processes events and stores the data into Elasticsearch. * There is a federated query layer in front of Elasticsearch to provide OLAP query capabilities.
* Apache Flink’s advanced windowing features, programming model, and checkpointing convinced Uber to move away from the simplicity of Apache Samza.
* The forecasting system allows Uber to remove the notion of delay by using recent signals plus historical data to project what is happening now and what will happen into the future.
* Uber’s pipeline for deploying ML models: HDFS, feature engineering, organizing into data structures (similar to data frames), deploy mostly offline training models, train models, & store into a container-based model manager. 
* A model serving layer is used to pick which model to use, forecasting results are stored in an OLAP data store, a validation layer compares real results against forecast results to verify the model is working as desired, and a rollback feature enables poor performing models to be automatically replaced by previous one.
* “Without output, you don’t have input.” If you want to start leveraging machine learning, developers just need to start doing. Start with intuition and practice. Over time ask questions and learn what you need, then apply a laser focus to gain that knowledge.

You can also subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq

Subscribe: www.youtube.com/infoq
Like InfoQ on Facebook: bit.ly/2jmlyG8
Follow on Twitter: twitter.com/InfoQ
Follow on LinkedIn: www.linkedin.com/company/infoq
Check the landing page on InfoQ: https://bit.ly/2GJQbUo

Streaming: Danny Yuan on Real-Time, Time Series Forecasting @Uber

In this week’s podcast, we speak with Justin Garrison, co-author of Cloud Native Infrastructure (CNI).  

•	Behind scenes for O’Reilly book and choice of cover animal
•	Infrastructure and CNI approach
•	State and Immutability / Immutable VM  
•	Terraform and Kubernetes 
•	Observability 
•	The Why of Immutability 
•	Infrastructure as Software (Netflix) 
•	Site Reliability Engineering and DevOps

Justin Garrison on Cloud Native Infrastructure, Immutability, Observability and Much More

Marathi song mashup

In this podcast Wes Reisz is talking to Anastasiia Voitova, known as @vixentael in the security communities. She started her career as a mobile application developer, and in recent years has moved to focus mainly on designing and developing graphics software. We’re going to talk about cryptography, how to design libraries to be usable by developers, and designing cryptographic libraries. We’ll also discuss about her talk from the recent QCon New York , called “Making Security Usable”.

Why listen to this podcast:

 - Choosing a good encryption algorithm isn’t enough - the parameters need to be chosen carefully as well
 - Algorithms like MD5 should not be used for hashing any more
 - Security is not just the encryption layer - it is the design of the whole system
 - Backups should be encrypted as well
 - Logs may contain sensitive GDPR data and need to be processed accordingly

More on this: Quick scan our curated show notes on InfoQ https://bit.ly/2yRWdQc

You can also subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq

Subscribe: www.youtube.com/infoq
Like InfoQ on Facebook: bit.ly/2jmlyG8
Follow on Twitter: twitter.com/InfoQ
Follow on LinkedIn: www.linkedin.com/company/infoq
Check the landing page on InfoQ: https://bit.ly/2yRWdQc

InfoQ Podcast

Anastasiia Voitova on Cryptography and the Design of Cryptographic Libraries

Arun Gupta discusses with Wes Reisz some of the container-focused services that AWS offers, including differentiating ECS and EKS. Arun goes into some detail the role that Amazon Fargate plays and goals behinds EKS. Arun wraps ups discussing some of the open source work that AWS has recently been doing in the container space.

Why liste to this podcast:

- ECS & EKS are both managed control planes; Amazon Fargate is a technology used to provision clusters.
 - ECR is the Amazon Container registry (similar to the Docker Registry).
 - EKS is an opinionated why of running a Kubernetes cluster on AWS. It is a highly available managed control plane available on US East 1 and US West 2 
- EKS uses a split account. The control plane runs in an Amazon account and the workers run in customer’s account.
- Upstream compatibility is a core tenant of EKS.

You can subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq

Subscribe: www.youtube.com/infoq
Like InfoQ on Facebook: bit.ly/2jmlyG8
Follow on Twitter: twitter.com/InfoQ
Follow on LinkedIn: www.linkedin.com/company/infoq
Check the landing page on InfoQ: https://bit.ly/2tWT8t9

Arun Gupta on Managed Container Control Planes on AWS

Today on The InfoQ Podcast, Wes talks with Joe Beda. Joe is one of the co-creators of Kubernetes. What started in the fall of 2013 with Craig McLuckie, Joe Beda, and Brendan Burns working on cloud infrastructure has become the default orchestrator for cloud native architectures. Today on the show, the two discuss the recent purchase of Heptio by VMWare, the Kubernetes Privilege Escalation Flaw (and the response to it), Kubernetes Enhancement Proposals, the CNCF/organization of Kubernetes, and some of the future hopes for the platform.

Why listen to this podcast:

 - Heptio, the company Joe and Craig McLuckie co-founded, viewed themselves as not a Kubernetes company, but more of a cloud native company. Joining VMWare allowed the company to continue a mission of helping people decouple “moving to cloud/taking advantage of cloud” patterns (regardless of where you’re running).

 - Re:Invent 2017 when EKS was announced was a watershed moment for Kubernetes. It marked a time where enough customers were asking for Kubernetes that the major cloud providers started to offer first-class support.

 - Kubernetes 1.13 included a patch for the Kubernetes Privilege Escalation Flaw Patch. While the flaw was a bad thing, it demonstrated product maturity in the way the community-based security response.

 - Kubernetes has an idea of committees, sigs, and working groups. Security is one of the committees. There were a small group of people who coordinated the security response. From there, trusted sets of vendors validated and test patches. Most of the response is based on how many other open source projects handle security response.

 - Over the last couple of releases, Kubernetes has introduced a Sig Architecture special interest group. It’s an overarching review for changes that sweep across Kubernetes. As part of Sig Architecture, the Kubernetes community has introduced Kubernetes Enhancement Proposal process (or KEPs). It’s a way for people to propose architectural changes to Kubernetes.

 - The goal of the CNCF is to curate and provide support to a set of projects (of which Kubernetes is one). The TOC (Technical Oversight Committee) decides which projects are going to be part of the CNCF and how those projects are supported.

 - Kubernetes was always viewed by the creators as something to be build on. It was never really viewed as the end goal.

You can also subscribe to the InfoQ newsletter to receive weekly updates on the hottest topics from professional software development. bit.ly/24x3IVq

Subscribe: www.youtube.com/infoq
Like InfoQ on Facebook: bit.ly/2jmlyG8
Follow on Twitter: twitter.com/InfoQ
Follow on LinkedIn: www.linkedin.com/company/infoq

Marathi song mashup

Avinash Ravindra Patil

Related tracks