Deep Embedded Self Organizing Map (DESOM), a hybrid Deep Neural Network based Autoencoder-Decoder (AE-DE) with an embedded Self Organizing Map (SOM), is applied successfully for the first time to detect anomaly in the performance metrics of mobile network entities with over 94% accuracy. SOM has been widely used in many areas for anomaly detection such as fraud detection, intrusion detection, etc. DESOM is a recent enhancement of SOM but not evaluated as practical solution for real problems prior to this work. Several novel methods to detect concept drift using the intrinsic features of DESOM have been incorporated in the complete solution pipeline.

Speaker
Jayanta Choudhury
Senior Data Scientist
Ericsson Inc.
Santa Clara, California United States

Anila Joshi
Sr. Data Science Manager
Ericsson Inc.
Santa Clara, California United States

Track
Performance Engineering and DevOps

Schedule
TBA

Speaker
Dr. Eli David
Co-founder, DeepCube
Tel Aviv, Israel

Track
What’s New / Modern Enterprise IT

Schedule
TBA

This presentation will examine one of the myths busted by MythBusters®. The myth had to do with the most efficient way to queue customers checking out of a grocery store; a single line for all customers or one line per checker. Queuing theory tells us that a single line provides the best response time. However, the “load test” performed by MythBusters showed that the single line queuing strategy was fastest. We will start by examining the details of the MythBusters load test. Next, modeling will be used to represent their “application” to understand the performance implications and sensitivity of the two competing architectures. The presentation will conclude with a set of lessons learned for applying performance modeling to real-world applications.

Speaker
Richard Gimartc
Consultant
RG
Austin, Texas United States

Track
Monitoring and Observability

Schedule
TBA

With the growing popularity of logging and distributed tracing for observability into our systems, attention needs to be paid to the overhead and impact of the logging and the tracing itself. In this session, we make the case for attention to logging, tracing, and other low-level services. We also discuss reasons why these low-level services have historically been glossed over and what it would take to facilitate visibility into low-level services.

Speaker:
Danny Chen
Software Engineer
Bloomberg LP
New York, New York United States

Track:
Monitoring and Observability

Scheduled:
TBA

The way we develop and operate software is changing – and performance engineering should change too to remain relevant. It is not happening the first time in history – performance engineering has a long history that can be traced at least to 1960s – and we will discuss what changes we had in the past, what trends exist now, and how performance testing (and performance engineering in general) should adjust to agile / DevOps / Cloud /AI / etc. It is changing as we speak, so no exact recipes – but trends are pretty clear and we will discuss what skills performance engineers need to remain in business.

Speaker:
Alexander Podelko
Staff Performance Engineer
MongoDB
Stamford, Connecticut United States

Track:
Performance Engineering and DevOps

Scheduled:
TBA

As a user, we are very confined to watching different data in tables, charts, and graphs format. But the right visualizations can and should help you enabling you to analyze even larger data sets. While moving from antiquated excel sheets is a step in the right direction, determining the right visualization for your data can still be a challenge. This session will discuss why visualization is more important that the data itself and show what happens when we apply machine learning to data and see automated data visualizations.

Speaker:
Jyotsna Chatradhi
Engineering Lead
Broadcom
Plano, Texas United States

Track:
Monitoring and Observability

Scheduled:
TBA

AI is more of an art than just a technology. Finding the right use cases is key and design thinking is crucial to determine what your customers need. We will walk through how to identify customer- focused use cases and the importance of stakeholder management to bring the product vision to life.

Speaker:
Kanwar Gaurav Paul
Director Product Management Artificial Intelligence
Fidelity Investments
Cary, North Carolina United States

Track:
What’s New

Scheduled:
TBA

While software architects have broken their monoliths into event-driven service architectures, many DevOps architects are still building monolithic inspired release pipelines. This results in complex pipeline code, tight integration of process and tools, lengthy diagnostics sessions to fix broken pipelines and puts strains on the underlying resources that build, deploy, test and validate.

In this session, we introduce a new approach: Event-driven Continuous Delivery and Operation Automation for modern DevOps! This approach decouples the declaration of processes for delivery and remediation from tooling. It uses an “everything as code” approach which includes deployment, testing, quality gate, observability, promotion and remediation definitions.

One open-source tool which implements this new approach is Keptn which was heavily inspired by Google’s SRE practice around SLIs & SLOs. Join us, see Keptn in action and get inspired on what the future of DevOps tooling can look like.

Speaker:
Andreas Grabner
DevOps Activist
Dynatrace
Linz, OO Austria

Track:
Performance Engineering and DevOps

Scheduled:
TBA

With more data and processing occurring outside the data center, edge computing must now be part of the overall capacity management discipline. ITIL V3 added physical data center power, space, and cooling to capacity management. ITIL V4 added Dev/Sec/Ops and containers to capacity planning. This session discusses how edge computing complements cloud computing in being added to the capacity management discipline.

Speaker:
Chris Molloy
Distinguished Engineer
IBM
Raleigh, North Carolina United States

Track:
Data Centers and Cloud Infrastructure

Scheduled:
TBA

Observability is most often accomplished with open source technologies to handle logs, metrics, and traces. Most vendor tools collect lots of data across the various data sets. The challenge is that very few tools provide good guidance and capabilities around SLA management. When the tools provide this, they are often missing additional data types. In this session, there will be an outline of what it means to manage SLAs and SLOs with various methodologies. There will be examples of how this is done with various open-source tools including Grafana, Jaeger, and Prometheus.

Speaker:
Jonah Kowall
CTO
Logz.io
Miami, Florida United States

Track:
Monitoring and Observability

Scheduled:
TBA