Perspica MetriX is here. Learn More >>>

31 May

Introducing Perspica MetriX for Full Stack Observability

We’re excited to introduce Perspica MetriX, helping companies learn which metrics matter so that they can detect and resolve anomalies faster.

Many of today’s business services and applications are supported by “hyper-scale” architectures that generate massive amounts of monitoring data. Due to the adoption of DevOps and infrastructure monitoring automation, modern companies are generating and streaming millions of data points per second across millions of metrics. With this volume of data, production ops teams are no longer able to keep up with what alarms are valid, much less how to fix them. Perspica uses artificial intelligence to analyze high volumes of application and infrastructure data in real time, using machine learning to establish behavior profiling and understand what is normal behavior while leveraging anomaly detection to reduce alarm storms.

Read more

Read more
12 Apr

Azure and Google Cloud Are Closing the Gap on AWS

It’s clear that Amazon is the vendor to beat in cloud computing platforms. Less than ten years ago Amazon introduced the idea that companies could save both money and time by using Amazon’s cloud services rather than build out their own infrastructure. Amazon’s AWS services are now ubiquitous, with annual revenue topping $12 billion in 2016.

Microsoft and Google were later entrants into the cloud market. Today’s MS Azure revenues are $2.5 billion, and Google’s “non-advertising” revenue ­– including Cloud, hardware, and Google Play – all total $3.4 billion.

As IT departments and developers become more comfortable using this type of infrastructure as a service (IaaS), they are becoming more price-conscious and support-sensitive, providing an opportunity for other cloud players to differentiate themselves from AWS and grow their piece of this market.

The 451 recently published its “Voice of the Enterprise Cloud Transformation Study” which surveyed 700 IT buyers in small, medium, large and very large enterprises. This survey has been conducted quarterly since Q3 2014.

Read more

Read more
30 Mar

5 “Must Haves” of Monitoring Containerized Apps

We recently spoke on a panel called “Artificial Intelligence Powered Analytics in Production Operations” (you can watch the video replay here). Perspica’s CTO, JF Huard, was joined by Justin Fitzhugh, VP of Technical Operations for Instart Logic, and Manoj Choudhary, CTO of Loggly. One of the recurring themes of the panel was that manual thresholds don’t work in today’s big data world. Below is a summary of that thread of the conversation.

I was talking with a customer who runs technical operations in what he refers to as a “hyper-scale” data environment. When the conversation got to managing containers, he made a really interesting comment:

“Containers aren’t the problem – it’s that the infrastructure around them and the tools to support them don’t exist. Containers don’t fit into any of the molds or paradigms we have now.”

He hit the nail on the head. And it got me thinking, what needs to happen to the infrastructure and app monitoring tools so that guys like him can monitor and manage containerized applications?

Read more

Read more
21 Mar

Manual Thresholds are Dead … Long Live Automation

We recently spoke on a panel called “Artificial Intelligence Powered Analytics in Production Operations” (you can watch the video replay here). Perspica’s CTO, JF Huard, was joined by Justin Fitzhugh, VP of Technical Operations for Instart Logic, and Manoj Choudhary, CTO of Loggly. One of the recurring themes of the panel was that manual thresholds don’t work in today’s big data world. Below is a summary of that thread of the conversation.

Traditional monitoring practices require operators to set static thresholds manually. In today’s “hyper-scale” environments, setting manual thresholds has become obsolete, with many operations teams looking for AI solutions to automate the process.

Problem #1: Too many metrics and too much data
Big data technologies like AWS and open source software have made it easy to launch and scale new applications quickly. It’s not unusual for an application to generate tens or hundreds of thousands of metrics and half a million data points per second. At this volume, setting manual thresholds simply isn’t possible.

Read more

Read more
15 Mar

The New Way to Manage KPI’S

We recently spoke on panel a called “Artificial Intelligence Powered Analytics in Production Operations” (you can watch the video replay here). Perspica’s CTO, JF Huard, was joined by Justin Fitzhugh, VP of Technical Operations for Instart Logic, and Manoj Choudhary, CTO of Loggly. One of the recurring themes of the panel was that manual thresholds don’t work in today’s big data world. Below is a summary of that thread of the conversation.

The lifeblood of operations teams is maintaining business KPIs. These KPIs can vary from number of transactions per second, response time from their API, or customer-defined uptime SLAs. For companies whose business IS their application, KPI management has a direct revenue impact. A missed uptime SLA can cost refunded penalties, or a slow API response time can cost customer transactions.

Read more

Read more
28 Feb

[Video Recap] Artificial Intelligence Powered Analytics for Production Operations

Many thanks to The Hive for recently hosting a meetup to discuss AI Powered Analytics. And a special thanks to our speakers and moderator for making it a great evening!

The discussion focused around how production operations teams in companies like Uber and Twitter are handling millions of data points per second and analyzing it in real time to keep their hyper-scale applications up and running. Our expert panel gave their perspective on how specialized big data artificial intelligence provides the analysis and observability needed throughout the application stack to deliver prescriptive remediation and maximize application performance and uptime.

Read more

Read more
08 Feb

The Key to Monitoring Your Hyper-Scale Applications

I seem to be having this conversation with a lot of customers these days: the TechOps team responsible for keeping the company’s applications up and running says their data is growing by 20% or more per month. Even new startups can be processing 100,000 data points per second if they’re successful. Companies like Twitter are processing 3 billion data points per second.

One of these SREs referred to this as “hyper-scale” data. He’s running several monitoring tools, but he’s inundated with alarms – both real and false – and he has no way to know which of these alarms deserves at 2:00am call to someone on his team.

Although his app uses 60,000+ metrics, he can only monitor about 5,000 of them. Once one of the non-tracked metrics sets off an alarm, he’ll add them to his tracking dashboard. He has two full-time team members who spend their day tracking down these alarms, trying to figure out the root cause of the issues. He’s not even looking at over 80% of his metrics. Clearly this method doesn’t scale.

Read more

Read more
14 Dec

Ovum Research 2017 Prediction: Machine Learning Applications Will Be Major Disrupter for Big Data Analytics

‘Tis the season for new years predictions. Ovum Research just published, “2017 Trends to Watch: Big Data” and had some interesting insight into how machine learning analytics will affect companies next year:

“The breakout use case for big data will be fast data. The Internet of Things (IoT) is increasing the urgency for enterprises to embrace real-time streaming analytics, as use cases from mobile devices and sensors become compelling to a wide range of industry sectors. Machine learning, which has garnered its share of hype, will continue to grow; but in most cases, machine learning will be embedded in application infrastructure and services rather than custom-developed because few organizations outside the Global 2000 (or digital online businesses) will have data scientists on their staff.”

Read more

Read more
08 Dec

[Part 3] Automate Your Incident Management: Troubleshoot Incidents using your ITSM tool

Yesterday we talked about how Perspica helps automate performance and IT incident management. Now let’s look at how to use tools like ServiceNow and Perspica’s knowledge base to get you the information you need to fix issues when they do occur.

Problem Analysis

Troubleshoot incidents using your ITSM tool

Perspica classifies anomalies to tell an operator if it is a minor fluctuation or part of a larger performance problem impacting end-users. Perspica ranks anomalies by impact on object health and instantly determines the root cause of performance problems. At the same time, Perspica also finds and delivers recommended corrective actions, from both its own knowledge base of best practices and also from the knowledge bases of the affected systems. This analysis is then sent to ITSM tools like Service Now.

Read more

Read more
07 Dec

[Part 2] Automate Your Incident Management: Perspica’s Approach

Yesterday we talked about how application outages are costing companies more than $500,000 per incident.
How can you minimize this cost, and get applications back up and running faster after you have an issue?
Perspica reduces your MTTR by using AI-powered machine learning analytics to understand the relationships between infrastructure components, learn what behavior is normal, then detect anomalous behavior patterns.

This analysis is used in several ways:

Read more

Read more
06 Dec

[Part 1] Automate Your Incident Management

Application downtime is expensive. Gartner estimates the cost of average application outage is $505,500. These costs include lost revenue, staff resources to get the application working again, and subsequent damage to brand reputation. How do these numbers add up so quickly?

PROBLEM #1: VOLUME OF ALARMS HAS OUTSTRIPPED HUMAN CAPACITY

Let’s look at the application itself. Todays’s applications have become more complex than ever before as they have moved from physical to virtual, cloud-based environments. “Big data” has become “bigger data”, making it exceedingly difficult to track application topologies as well as collect performance, events, and log data.

Read more

Read more
20 Sep

Connect the Dots Between Your IT Operations Tools

Once people see a demo of how Perspica brings a new level of intelligence to their infrastructure technology services, they often ask us if Perspica replaces some of their existing tools, or adds to their toolset.

Our customers have invested in a variety of tools to manage their application infrastructure, but often they’re still struggling to maintain the service levels their businesses require. The average application runs 10-12 monitoring tools – and we’ve seen IT departments with a complex infrastructures running up to 100 tools. These tools have created a new problem of “monitoring overload.”

Why? Because even though they provide effective monitoring for their “slice” of the application stack, these tools aren’t connected.
Read more

Read more
30 Aug

Why Monitoring Tools Aren’t Enough

We recently announced how Perspica’s IT incident management system is using AI to leapfrog traditional monitoring tools using advanced analytics. Let’s take a closer look at why monitoring tools aren’t enough for today’s complex applications.

Knowing whether a resource is “up or down” doesn’t help you fix the problem. Monitoring tools were originally created to help us understand when there was a problem with any of our infrastructure elements. Letting us know whether a server needed to be rebooted or that you’d run out of memory was helpful then, but in today’s complex applications knowing that your application has had alert because of a service level violation still leaves a long way to identify and understand how to fix the issue at hand.
Read more

Read more
26 Aug

Perspica’s Jonathan Creasy Assists with Flood Cleanup in Baton Rouge, Louisiana

I’ve been a volunteer firefighter since 2001, and started doing disaster relief with the Red Cross in 2005. I got involved with Disaster Tech Lab in 2010 as a way to continue to use my IT skills to support relief teams after my time with the Red Cross ended.

American_redcross_2012_logo

Last week I received an email from Distaster Tech Lab CEO Evert Bopp asking me if I was available and interested in doing the recon for possible deployments to Louisiana.

I got on the first plane.

Louisiana’s latest hurricane hasn’t received nearly as much press attention as Hurricane Katrina – in fact, it’s been called The “Storm Without a Name” (SWN) even though it reportedly dropped nearly 3 times more rain than Hurricane Katrina did. Katrina let loose 2.3 trillion gallons of rainwater. SWN dropped 7.1.

Read more

Read more
22 Aug

Operational Intelligence for The Hybrid Cloud

Perspica recently announced the integration of its operational intelligence offering with Amazon Web Services to provide a complete view into application infrastructure across hybrid cloud environments. Perspica already provides these advanced big data analytics tools for VMware and Docker powered private cloud environments.

Enterprises are increasingly using hybrid cloud deployment whereby some applications may be deployed in private cloud and others in public cloud guided by latency, security, privacy and availability concerns. The more complex scenarios result when these applications are federated across private and public clouds where some of the application components are running in private and others in public cloud environments.

Read more

Read more
10 Aug

3 Benefits of I.T. Infrastructure Monitoring Tools That Can Improve Your Company’s Bottom Line

As companies embrace continuous delivery paradigms, specialized TechOps and DevOps teams are emerging across the organization.

These agile, cross-functional groups include TechOps teams, concerned with keeping production environments performant, working in tandem with DevOps teams who work on pre-production and staging environments. As the lines between their organizations have blurred, they now often share a common set of concerns:

Read more

Read more