This document provides an overview of HP Virtual Connect technology which simplifies network infrastructure by virtualizing server-to-network connections. Key features include consolidating network connections onto fewer modules to reduce costs, enabling bandwidth allocation per server as needed, and providing a centralized management console for multiple server enclosures. HP Virtual Connect offers various modules that provide Ethernet, Fibre Channel, and converged connectivity to simplify management and improve flexibility of the network environment.
Presentation on the struggles with traditional architectures and an overview of the Lambda Architecture utilizing Spark to drive massive amounts of both batch and streaming data for processing and analytics
Story of architecture evolution of one project from zero to Lambda Architecture. Also includes information on how we scaled cluster as soon as architecture is set up.
Contains nice performance charts after every architecture change.
Large Scale Lakehouse Implementation Using Structured StreamingDatabricks
Business leads, executives, analysts, and data scientists rely on up-to-date information to make business decision, adjust to the market, meet needs of their customers or run effective supply chain operations.
Come hear how Asurion used Delta, Structured Streaming, AutoLoader and SQL Analytics to improve production data latency from day-minus-one to near real time Asurion’s technical team will share battle tested tips and tricks you only get with certain scale. Asurion data lake executes 4000+ streaming jobs and hosts over 4000 tables in production Data Lake on AWS.
DataOps Automation for a Kafka Streaming Platform (Andrew Stevenson + Spiros ...HostedbyConfluent
DataOps challenges us to build data experiences in a repeatable way. For those with Kafka, this means finding a means of deploying flows in an automated and consistent fashion.
The challenge is to make the deployment of Kafka flows consistent across different technologies and systems: the topics, the schemas, the monitoring rules, the credentials, the connectors, the stream processing apps. And ideally not coupled to a particular infrastructure stack.
In this talk we will discuss the different approaches and benefits/disadvantages to automating the deployment of Kafka flows including Git operators and Kubernetes operators. We will walk through and demo deploying a flow on AWS EKS with MSK and Kafka Connect using GitOps practices: including a stream processing application, S3 connector with credentials held in AWS Secrets Manager.
Modern ETL Pipelines with Change Data CaptureDatabricks
In this talk we’ll present how at GetYourGuide we’ve built from scratch a completely new ETL pipeline using Debezium, Kafka, Spark and Airflow, which can automatically handle schema changes. Our starting point was an error prone legacy system that ran daily, and was vulnerable to breaking schema changes, which caused many sleepless on-call nights. As most companies, we also have traditional SQL databases that we need to connect to in order to extract relevant data.
This is done usually through either full or partial copies of the data with tools such as sqoop. However another approach that has become quite popular lately is to use Debezium as the Change Data Capture layer which reads databases binlogs, and stream these changes directly to Kafka. As having data once a day is not enough anymore for our bussiness, and we wanted our pipelines to be resilent to upstream schema changes, we’ve decided to rebuild our ETL using Debezium.
We’ll walk the audience through the steps we followed to architect and develop such solution using Databricks to reduce operation time. By building this new pipeline we are now able to refresh our data lake multiple times a day, giving our users fresh data, and protecting our nights of sleep.
Deep Dive into the New Features of Apache Spark 3.1Databricks
Continuing with the objectives to make Spark faster, easier, and smarter, Apache Spark 3.1 extends its scope with more than 1500 resolved JIRAs. We will talk about the exciting new developments in the Apache Spark 3.1 as well as some other major initiatives that are coming in the future. In this talk, we want to share with the community many of the more important changes with the examples and demos.
The following features are covered: the SQL features for ANSI SQL compliance, new streaming features, and Python usability improvements, the performance enhancements and new tuning tricks in query compiler.
Presentation on the struggles with traditional architectures and an overview of the Lambda Architecture utilizing Spark to drive massive amounts of both batch and streaming data for processing and analytics
Story of architecture evolution of one project from zero to Lambda Architecture. Also includes information on how we scaled cluster as soon as architecture is set up.
Contains nice performance charts after every architecture change.
Large Scale Lakehouse Implementation Using Structured StreamingDatabricks
Business leads, executives, analysts, and data scientists rely on up-to-date information to make business decision, adjust to the market, meet needs of their customers or run effective supply chain operations.
Come hear how Asurion used Delta, Structured Streaming, AutoLoader and SQL Analytics to improve production data latency from day-minus-one to near real time Asurion’s technical team will share battle tested tips and tricks you only get with certain scale. Asurion data lake executes 4000+ streaming jobs and hosts over 4000 tables in production Data Lake on AWS.
DataOps Automation for a Kafka Streaming Platform (Andrew Stevenson + Spiros ...HostedbyConfluent
DataOps challenges us to build data experiences in a repeatable way. For those with Kafka, this means finding a means of deploying flows in an automated and consistent fashion.
The challenge is to make the deployment of Kafka flows consistent across different technologies and systems: the topics, the schemas, the monitoring rules, the credentials, the connectors, the stream processing apps. And ideally not coupled to a particular infrastructure stack.
In this talk we will discuss the different approaches and benefits/disadvantages to automating the deployment of Kafka flows including Git operators and Kubernetes operators. We will walk through and demo deploying a flow on AWS EKS with MSK and Kafka Connect using GitOps practices: including a stream processing application, S3 connector with credentials held in AWS Secrets Manager.
Modern ETL Pipelines with Change Data CaptureDatabricks
In this talk we’ll present how at GetYourGuide we’ve built from scratch a completely new ETL pipeline using Debezium, Kafka, Spark and Airflow, which can automatically handle schema changes. Our starting point was an error prone legacy system that ran daily, and was vulnerable to breaking schema changes, which caused many sleepless on-call nights. As most companies, we also have traditional SQL databases that we need to connect to in order to extract relevant data.
This is done usually through either full or partial copies of the data with tools such as sqoop. However another approach that has become quite popular lately is to use Debezium as the Change Data Capture layer which reads databases binlogs, and stream these changes directly to Kafka. As having data once a day is not enough anymore for our bussiness, and we wanted our pipelines to be resilent to upstream schema changes, we’ve decided to rebuild our ETL using Debezium.
We’ll walk the audience through the steps we followed to architect and develop such solution using Databricks to reduce operation time. By building this new pipeline we are now able to refresh our data lake multiple times a day, giving our users fresh data, and protecting our nights of sleep.
Deep Dive into the New Features of Apache Spark 3.1Databricks
Continuing with the objectives to make Spark faster, easier, and smarter, Apache Spark 3.1 extends its scope with more than 1500 resolved JIRAs. We will talk about the exciting new developments in the Apache Spark 3.1 as well as some other major initiatives that are coming in the future. In this talk, we want to share with the community many of the more important changes with the examples and demos.
The following features are covered: the SQL features for ANSI SQL compliance, new streaming features, and Python usability improvements, the performance enhancements and new tuning tricks in query compiler.
Lambda Architecture: The Best Way to Build Scalable and Reliable Applications!Tugdual Grall
Lambda Architecture is a useful framework to think about designing big data applications. This framework has been built initially at Twitter. In this presentation you will learn, based on concrete examples how to build deploy scalable and fault tolerant applications, with a focus on Big Data and Hadoop.
This presentation was delivered at the OOP conference, Munich, Feb 2016
Project Ouroboros: Using StreamSets Data Collector to Help Manage the StreamS...Pat Patterson
On a typical day we see hundreds of downloads of StreamSets Data Collector, our open source data integration tool. We used to wrangle our download logs using a combination of the AWS S3 command line, sed, grep, awk and other tools, all run from a shell script (on my laptop!) once a week. This was a classic example of a brittle, hard to maintain, custom data integration. One day it dawned on me, "This is crazy, we have a tool that can do all this!". In this session, I'll explain how I built a dataflow pipeline to stream content delivery network (CDN) logs from S3 to MySQL in real-time, allowing us to gain valuable insights into our open source community. You'll also learn how we use the same techniques to not only gain insights into our community on Slack, but also build tools to better serve them.
Designing the Next Generation of Data Pipelines at Zillow with Apache SparkDatabricks
The trade-off between development speed and pipeline maintainability is a constant for data engineers, especially for those in a rapidly evolving organization
Streaming Data Lakes using Kafka Connect + Apache Hudi | Vinoth Chandar, Apac...HostedbyConfluent
Apache Hudi is a data lake platform, that provides streaming primitives (upserts/deletes/change streams) on top of data lake storage. Hudi powers very large data lakes at Uber, Robinhood and other companies, while being pre-installed on four major cloud platforms.
Hudi supports exactly-once, near real-time data ingestion from Apache Kafka to cloud storage, which is typically used in-place of a S3/HDFS sink connector to gain transactions and mutability. While this approach is scalable and battle-tested, it can only ingest data in mini batches, leading to lower data freshness. In this talk, we introduce a Kafka Connect Sink Connector for Apache Hudi, which writes data straight into Hudi's log format, making the data immediately queryable, while Hudi's table services like indexing, compaction, clustering work behind the scenes, to further re-organize for better query performance.
Simplifying Disaster Recovery with Delta LakeDatabricks
There’s a need to develop a recovery process for Delta table in a DR scenario. Cloud multi-region sync is Asynchronous. This type of replication does not guarantee the chronological order of files at the target (DR) region. In some cases, we can expect large files to arrive later than small files.
Real-Time Data Pipelines with Kafka, Spark, and Operational DatabasesSingleStore
Eric Frenkiel, MemSQL CEO and co-founder and Gartner Catalyst. August 11, 2015, San Diego, CA. Watch the Pinterest Demo Video here: https://youtu.be/KXelkQFVz4E
Apache Pulsar: The Next Generation Messaging and Queuing SystemDatabricks
Apache Pulsar is the next generation messaging and queuing system with unique design trade-offs driven by the need for scalability and durability. Its two layered architecture of separating message storage from serving led to an implementation that unifies the flexibility and the high-level constructs of messaging, queuing and light weight computing with the scalable properties of log storage systems.
A Journey to Building an Autonomous Streaming Data Platform—Scaling to Trilli...Databricks
We will present the design and evolution of Nvidia's 100% Self-Service Streaming Big-Data Platform (ETL, Analytics, AI Training & Inferencing) powered by Spark and Nvidia GPUs. We will discuss the architecture, major challenges that we faced, and lessons learned along the way. Nvidia's data platform processes 10's of billions of events per day, supporting several Nvidia products like GPU Cloud, GeForce NOW Cloud Gaming, AI Smart Cities, DriveSim for Self Driving cars etc. In this talk, we are going to deep dive on Nvidia's next generation data platform with new custom built frameworks, automation tools, and a monitoring system on top of Spark. Thus empowering our developers to build new Spark-powered applications at the speed of light (SOL) with full self-service unified data flows. We will showcase these new tools : a) Zero-engineering dashboards, b) Out-of-the box Spark Streaming applications with automated schema management, c) Custom Spark Streaming to Elastic search connector with enhanced security, d) GDPR compliant SQL access control and auditing with a new custom token management framework, e) Migration from logstash clusters to Spark Streaming for log parsing, etc. We will discuss how decoupling Data-Platform and Applications helped us achieve the next level of scale, self-service, and, security. Finally, we will demo our Platform's App-Store, where developers can shop for new Apps and deploy them with ease - with automated dashboards, streaming ETL, analytics, monitoring, AI training and inferencing. Extended Description: With structured telemetry events and unstructured logs growing at 1000% rate year-over-year, it is extremely important to handle this scale with strict SLAs and high reliability while maintaining extremely low latency. We will discuss how we handled these scaling & security concerns to solve business requirements. Additionally, we will be open-sourcing some of our custom spark frameworks during the talk.
Speakers: Satish Dandu, Rohit Kulkarni
Delta Lake, an open-source innovations which brings new capabilities for transactions, version control and indexing your data lakes. We uncover how Delta Lake benefits and why it matters to you. Through this session, we showcase some of its benefits and how they can improve your modern data engineering pipelines. Delta lake provides snapshot isolation which helps concurrent read/write operations and enables efficient insert, update, deletes, and rollback capabilities. It allows background file optimization through compaction and z-order partitioning achieving better performance improvements. In this presentation, we will learn the Delta Lake benefits and how it solves common data lake challenges, and most importantly new Delta Time Travel capability.
Near Real-Time Analytics with Apache Spark: Ingestion, ETL, and Interactive Q...Databricks
Near real-time analytics has become a common requirement for many data teams as the technology has caught up to the demand. One of the hardest aspects of enabling near-realtime analytics is making sure the source data is ingested and deduplicated often enough to be useful to analysts while writing the data in a format that is usable by your analytics query engine. This is usually the domain of many tools since there are three different aspects of the problem: streaming ingestion of data, deduplication using an ETL process, and interactive analytics. With Spark, this can be done with one tool. This talk with walk you through how to use Spark Streaming to ingest change-log data, use Spark batch jobs to perform major and minor compaction, and query the results with Spark.SQL. At the end of this talk you will know what is required to setup near-realtime analytics at your organization, the common gotchas including file formats and distributed file systems, and how to handle data the unique data integrity issues that arise from near-realtime analytics.
Change Data Feed is a new feature of Delta Lake on Databricks that is available as a public preview since DBR 8.2. This feature enables a new class of ETL workloads such as incremental table/view maintenance and change auditing that were not possible before. In short, users will now be able to query row level changes across different versions of a Delta table.
In this talk we will dive into how Change Data Feed works under the hood and how to use it with existing ETL jobs to make them more efficient and also go over some new workloads it can enable.
Tangram: Distributed Scheduling Framework for Apache Spark at FacebookDatabricks
Tangram is a state-of-art resource allocator and distributed scheduling framework for Spark at Facebook with hierarchical queues and a resource based container abstraction. We support scheduling and resource management for a significant portion of Facebook's data warehouse and machine learning workloads that equates to running millions of jobs across several clusters with tens of thousands of machines. In this talk, we will describe Tangram's architecture, discuss Facebook's need for a custom scheduler, and explain how Tangram schedules Spark workloads at scale. We will specifically focus on several important features around improving Spark's efficiency, usability and reliability: 1. IO-rebalancer (Tetris) Support 2. User-Fairness Queueing 3. Heuristic-Based Backfill Scheduling Optimizations.
Stream All Things—Patterns of Modern Data Integration with Gwen ShapiraDatabricks
Data integration is a really difficult problem. We know this because 80% of the time in every project is spent getting the data you want the way you want it. We know this because this problem remains challenging despite 40 years of attempts to solve it. All we want is a service that will be reliable, handle all kinds of data and integrate with all kinds of systems, be easy to manage and scale as our systems grow. Oh, and it should be super low latency too. Is it too much to ask?
In this presentation, we’ll discuss the basic challenges of data integration and introduce few design and architecture patterns that are used to tackle these challenges. We will then explore how these patterns can be implemented using Apache Kafka. Difficult problems are difficult and we offer no silver bullets, but we will share pragmatic solutions that helped many organizations build fast, scalable and manageable data pipelines.
Monitoring of GPU Usage with Tensorflow Models Using PrometheusDatabricks
Understanding the dynamics of GPU utilization and workloads in containerized systems is critical to creating efficient software systems. We create a set of dashboards to monitor and evaluate GPU performance in the context of TensorFlow. We monitor performance in real time to gain insight into GPU load, GPU memory and temperature metrics in a Kubernetes GPU enabled system. Visualizing TensorFlow training job metrics in real time using Prometheus allows us to tune and optimize GPU usage. Also, because Tensor flow jobs can have both GPU and CPU implementations it is useful to view detailed real time performance data from each implementation and choose the best implementation. To illustrate our system, we will show a live demo gathering and visualizing GPU metrics on a GPU enabled Kubernetes cluster with Prometheus and Grafana.
Spark is fast and general engine for large-scale data processing which can solve all of your problems.
… Or can it?
This talk will cover real-world issues encountered during migration of the existing product to Spark infrastructure.
Aimed at software engineers that just started to evaluate Spark or those who are already using it.
Growing the Delta Ecosystem to Rust and Python with Delta-RSDatabricks
In this session we will introduce the delta-rs project which is helping bring the power of Delta Lake outside of the Spark ecosystem. By providing a foundational Delta Lake library in Rust, delta-rs can enable native bindings in Python, Ruby, Golang, and more.We will review what functionality delta-rs supports in its current Rust and Python APIs and the upcoming roadmap.
We will also give an overview of one of the first projects to use it in production: kafka-delta-ingest, which builds on delta-rs to provide a high throughput service to bring data from Kafka into Delta Lake.
Enabling Insight to Support World-Class Supercomputing (Stefan Ceballos, Oak ...confluent
The Oak Ridge Leadership Facility (OLCF) in the National Center for Computational Sciences (NCCS) division at Oak Ridge National Laboratory (ORNL) houses world-class high-performance computing (HPC) resources and has a history of operating top-ranked supercomputers on the TOP500 list, including the world's current fastest, Summit, an IBM AC922 machine with a peak of 200 petaFLOPS. With the exascale era rapidly approaching, the need for a robust and scalable big data platform for operations data is more important than ever. In the past when a new HPC resource was added to the facility, pipelines from data sources spanned multiple data sinks which oftentimes resulted in data silos, slow operational data onboarding, and non-scalable data pipelines for batch processing. Using Apache Kafka as the message bus of the division's new big data platform has allowed for easier decoupling of scalable data pipelines, faster data onboarding, and stream processing with the goal to continuously improve insight into the HPC resources and their supporting systems. This talk will focus on the NCCS division's transition to Apache Kafka over the past few years to enhance the OLCF's current capabilities and prepare for Frontier, OLCF's future exascale system; including the development and deployment of a full big data platform in a Kubernetes environment from both a technical and cultural shift perspective. This talk will also cover the mission of the OLCF, the operational data insights related to high-performance computing that the organization strives for, and several use-cases that exist in production today.
Designing and Implementing a Real-time Data Lake with Dynamically Changing Sc...Databricks
Building a curated data lake on real time data is an emerging data warehouse pattern with delta. However in the real world, what we many times face ourselves with is dynamically changing schemas which pose a big challenge to incorporate without downtimes.
Neo4j Graph Streaming Services with Apache Kafkajexp
In this presentation we give an high level overview of the Neo4j-Kafka integration and the Confluent partnership.
Providing change-data-capture and ingestion capabilities as Neo4j Extension and the Kafka Connect Neo4j Sink on Confluent Hub allows you to integrate real-time streaming with graph querying and analytics.
Lambda Architecture: The Best Way to Build Scalable and Reliable Applications!Tugdual Grall
Lambda Architecture is a useful framework to think about designing big data applications. This framework has been built initially at Twitter. In this presentation you will learn, based on concrete examples how to build deploy scalable and fault tolerant applications, with a focus on Big Data and Hadoop.
This presentation was delivered at the OOP conference, Munich, Feb 2016
Project Ouroboros: Using StreamSets Data Collector to Help Manage the StreamS...Pat Patterson
On a typical day we see hundreds of downloads of StreamSets Data Collector, our open source data integration tool. We used to wrangle our download logs using a combination of the AWS S3 command line, sed, grep, awk and other tools, all run from a shell script (on my laptop!) once a week. This was a classic example of a brittle, hard to maintain, custom data integration. One day it dawned on me, "This is crazy, we have a tool that can do all this!". In this session, I'll explain how I built a dataflow pipeline to stream content delivery network (CDN) logs from S3 to MySQL in real-time, allowing us to gain valuable insights into our open source community. You'll also learn how we use the same techniques to not only gain insights into our community on Slack, but also build tools to better serve them.
Designing the Next Generation of Data Pipelines at Zillow with Apache SparkDatabricks
The trade-off between development speed and pipeline maintainability is a constant for data engineers, especially for those in a rapidly evolving organization
Streaming Data Lakes using Kafka Connect + Apache Hudi | Vinoth Chandar, Apac...HostedbyConfluent
Apache Hudi is a data lake platform, that provides streaming primitives (upserts/deletes/change streams) on top of data lake storage. Hudi powers very large data lakes at Uber, Robinhood and other companies, while being pre-installed on four major cloud platforms.
Hudi supports exactly-once, near real-time data ingestion from Apache Kafka to cloud storage, which is typically used in-place of a S3/HDFS sink connector to gain transactions and mutability. While this approach is scalable and battle-tested, it can only ingest data in mini batches, leading to lower data freshness. In this talk, we introduce a Kafka Connect Sink Connector for Apache Hudi, which writes data straight into Hudi's log format, making the data immediately queryable, while Hudi's table services like indexing, compaction, clustering work behind the scenes, to further re-organize for better query performance.
Simplifying Disaster Recovery with Delta LakeDatabricks
There’s a need to develop a recovery process for Delta table in a DR scenario. Cloud multi-region sync is Asynchronous. This type of replication does not guarantee the chronological order of files at the target (DR) region. In some cases, we can expect large files to arrive later than small files.
Real-Time Data Pipelines with Kafka, Spark, and Operational DatabasesSingleStore
Eric Frenkiel, MemSQL CEO and co-founder and Gartner Catalyst. August 11, 2015, San Diego, CA. Watch the Pinterest Demo Video here: https://youtu.be/KXelkQFVz4E
Apache Pulsar: The Next Generation Messaging and Queuing SystemDatabricks
Apache Pulsar is the next generation messaging and queuing system with unique design trade-offs driven by the need for scalability and durability. Its two layered architecture of separating message storage from serving led to an implementation that unifies the flexibility and the high-level constructs of messaging, queuing and light weight computing with the scalable properties of log storage systems.
A Journey to Building an Autonomous Streaming Data Platform—Scaling to Trilli...Databricks
We will present the design and evolution of Nvidia's 100% Self-Service Streaming Big-Data Platform (ETL, Analytics, AI Training & Inferencing) powered by Spark and Nvidia GPUs. We will discuss the architecture, major challenges that we faced, and lessons learned along the way. Nvidia's data platform processes 10's of billions of events per day, supporting several Nvidia products like GPU Cloud, GeForce NOW Cloud Gaming, AI Smart Cities, DriveSim for Self Driving cars etc. In this talk, we are going to deep dive on Nvidia's next generation data platform with new custom built frameworks, automation tools, and a monitoring system on top of Spark. Thus empowering our developers to build new Spark-powered applications at the speed of light (SOL) with full self-service unified data flows. We will showcase these new tools : a) Zero-engineering dashboards, b) Out-of-the box Spark Streaming applications with automated schema management, c) Custom Spark Streaming to Elastic search connector with enhanced security, d) GDPR compliant SQL access control and auditing with a new custom token management framework, e) Migration from logstash clusters to Spark Streaming for log parsing, etc. We will discuss how decoupling Data-Platform and Applications helped us achieve the next level of scale, self-service, and, security. Finally, we will demo our Platform's App-Store, where developers can shop for new Apps and deploy them with ease - with automated dashboards, streaming ETL, analytics, monitoring, AI training and inferencing. Extended Description: With structured telemetry events and unstructured logs growing at 1000% rate year-over-year, it is extremely important to handle this scale with strict SLAs and high reliability while maintaining extremely low latency. We will discuss how we handled these scaling & security concerns to solve business requirements. Additionally, we will be open-sourcing some of our custom spark frameworks during the talk.
Speakers: Satish Dandu, Rohit Kulkarni
Delta Lake, an open-source innovations which brings new capabilities for transactions, version control and indexing your data lakes. We uncover how Delta Lake benefits and why it matters to you. Through this session, we showcase some of its benefits and how they can improve your modern data engineering pipelines. Delta lake provides snapshot isolation which helps concurrent read/write operations and enables efficient insert, update, deletes, and rollback capabilities. It allows background file optimization through compaction and z-order partitioning achieving better performance improvements. In this presentation, we will learn the Delta Lake benefits and how it solves common data lake challenges, and most importantly new Delta Time Travel capability.
Near Real-Time Analytics with Apache Spark: Ingestion, ETL, and Interactive Q...Databricks
Near real-time analytics has become a common requirement for many data teams as the technology has caught up to the demand. One of the hardest aspects of enabling near-realtime analytics is making sure the source data is ingested and deduplicated often enough to be useful to analysts while writing the data in a format that is usable by your analytics query engine. This is usually the domain of many tools since there are three different aspects of the problem: streaming ingestion of data, deduplication using an ETL process, and interactive analytics. With Spark, this can be done with one tool. This talk with walk you through how to use Spark Streaming to ingest change-log data, use Spark batch jobs to perform major and minor compaction, and query the results with Spark.SQL. At the end of this talk you will know what is required to setup near-realtime analytics at your organization, the common gotchas including file formats and distributed file systems, and how to handle data the unique data integrity issues that arise from near-realtime analytics.
Change Data Feed is a new feature of Delta Lake on Databricks that is available as a public preview since DBR 8.2. This feature enables a new class of ETL workloads such as incremental table/view maintenance and change auditing that were not possible before. In short, users will now be able to query row level changes across different versions of a Delta table.
In this talk we will dive into how Change Data Feed works under the hood and how to use it with existing ETL jobs to make them more efficient and also go over some new workloads it can enable.
Tangram: Distributed Scheduling Framework for Apache Spark at FacebookDatabricks
Tangram is a state-of-art resource allocator and distributed scheduling framework for Spark at Facebook with hierarchical queues and a resource based container abstraction. We support scheduling and resource management for a significant portion of Facebook's data warehouse and machine learning workloads that equates to running millions of jobs across several clusters with tens of thousands of machines. In this talk, we will describe Tangram's architecture, discuss Facebook's need for a custom scheduler, and explain how Tangram schedules Spark workloads at scale. We will specifically focus on several important features around improving Spark's efficiency, usability and reliability: 1. IO-rebalancer (Tetris) Support 2. User-Fairness Queueing 3. Heuristic-Based Backfill Scheduling Optimizations.
Stream All Things—Patterns of Modern Data Integration with Gwen ShapiraDatabricks
Data integration is a really difficult problem. We know this because 80% of the time in every project is spent getting the data you want the way you want it. We know this because this problem remains challenging despite 40 years of attempts to solve it. All we want is a service that will be reliable, handle all kinds of data and integrate with all kinds of systems, be easy to manage and scale as our systems grow. Oh, and it should be super low latency too. Is it too much to ask?
In this presentation, we’ll discuss the basic challenges of data integration and introduce few design and architecture patterns that are used to tackle these challenges. We will then explore how these patterns can be implemented using Apache Kafka. Difficult problems are difficult and we offer no silver bullets, but we will share pragmatic solutions that helped many organizations build fast, scalable and manageable data pipelines.
Monitoring of GPU Usage with Tensorflow Models Using PrometheusDatabricks
Understanding the dynamics of GPU utilization and workloads in containerized systems is critical to creating efficient software systems. We create a set of dashboards to monitor and evaluate GPU performance in the context of TensorFlow. We monitor performance in real time to gain insight into GPU load, GPU memory and temperature metrics in a Kubernetes GPU enabled system. Visualizing TensorFlow training job metrics in real time using Prometheus allows us to tune and optimize GPU usage. Also, because Tensor flow jobs can have both GPU and CPU implementations it is useful to view detailed real time performance data from each implementation and choose the best implementation. To illustrate our system, we will show a live demo gathering and visualizing GPU metrics on a GPU enabled Kubernetes cluster with Prometheus and Grafana.
Spark is fast and general engine for large-scale data processing which can solve all of your problems.
… Or can it?
This talk will cover real-world issues encountered during migration of the existing product to Spark infrastructure.
Aimed at software engineers that just started to evaluate Spark or those who are already using it.
Growing the Delta Ecosystem to Rust and Python with Delta-RSDatabricks
In this session we will introduce the delta-rs project which is helping bring the power of Delta Lake outside of the Spark ecosystem. By providing a foundational Delta Lake library in Rust, delta-rs can enable native bindings in Python, Ruby, Golang, and more.We will review what functionality delta-rs supports in its current Rust and Python APIs and the upcoming roadmap.
We will also give an overview of one of the first projects to use it in production: kafka-delta-ingest, which builds on delta-rs to provide a high throughput service to bring data from Kafka into Delta Lake.
Enabling Insight to Support World-Class Supercomputing (Stefan Ceballos, Oak ...confluent
The Oak Ridge Leadership Facility (OLCF) in the National Center for Computational Sciences (NCCS) division at Oak Ridge National Laboratory (ORNL) houses world-class high-performance computing (HPC) resources and has a history of operating top-ranked supercomputers on the TOP500 list, including the world's current fastest, Summit, an IBM AC922 machine with a peak of 200 petaFLOPS. With the exascale era rapidly approaching, the need for a robust and scalable big data platform for operations data is more important than ever. In the past when a new HPC resource was added to the facility, pipelines from data sources spanned multiple data sinks which oftentimes resulted in data silos, slow operational data onboarding, and non-scalable data pipelines for batch processing. Using Apache Kafka as the message bus of the division's new big data platform has allowed for easier decoupling of scalable data pipelines, faster data onboarding, and stream processing with the goal to continuously improve insight into the HPC resources and their supporting systems. This talk will focus on the NCCS division's transition to Apache Kafka over the past few years to enhance the OLCF's current capabilities and prepare for Frontier, OLCF's future exascale system; including the development and deployment of a full big data platform in a Kubernetes environment from both a technical and cultural shift perspective. This talk will also cover the mission of the OLCF, the operational data insights related to high-performance computing that the organization strives for, and several use-cases that exist in production today.
Designing and Implementing a Real-time Data Lake with Dynamically Changing Sc...Databricks
Building a curated data lake on real time data is an emerging data warehouse pattern with delta. However in the real world, what we many times face ourselves with is dynamically changing schemas which pose a big challenge to incorporate without downtimes.
Neo4j Graph Streaming Services with Apache Kafkajexp
In this presentation we give an high level overview of the Neo4j-Kafka integration and the Confluent partnership.
Providing change-data-capture and ingestion capabilities as Neo4j Extension and the Kafka Connect Neo4j Sink on Confluent Hub allows you to integrate real-time streaming with graph querying and analytics.
A tecnología é uma ferramenta muito favoravel para o desenvolvimentos dos trabalhos na Educaçao. Com certeza o trabalho pedagógico tem mais fluencia ao contribui para o proceso de aprendizagem.
Blade Server I/O and Workloads of the Future (report)IT Brand Pulse
At the Intel Xeon E5-2600 v3 inflection point, this technology brief looks at how the latest Cisco UCS and HP BladeSystem blade servers match-up to workloads of the future.
Marvell QLogic 2600 Series 16Gb Gen 5 FC HBAs Double Performance and FlexibilityMarvell
Accelerate Virtualization and Cloud Deployments While Eliminating I/O Bottlenecks
KEY FINDINGS
Support for increased workloads, acceleration of application performance, and meeting the growing demands placed on the enterprise data center is key in the selection and deployment of a 16Gb Gen 5 Fibre Channel Adapter with the right architecture to scale
and support robust databases, mail servers, and secondary storage.
• Performance: QLogic® adapters from Cavium™ deliver double the performance of previous-generation adapters with up to 1.2 million IOPS and 3200 MBps bidirectional throughput.
• Superior Virtual Scalability and Lower Costs: Greater performance, VM density, and cost savings compared to Emulex adapters in VMware® vSphere® 5 and Microsoft® Hyper-V®environments.
• Unparalleled Flexibility: QLogic I/OFlex™ technology—any I/O, any network.
• Integrated Brocade Fabric Features: QLogic adapters deliver improved availability, streamlined deployment, and increased network performance.
To know more visit @ https://www.marvell.com/fibre-channel-adapters-and-controllers/
Blade Server I/O and Workloads of the Future (slides)IT Brand Pulse
At the Intel Xeon E5-2600 v3 inflection point, this technology brief looks at how the latest Cisco UCS and HP BladeSystem blade servers match-up to workloads of the future.
Your priorities are clear: meet the challenges of today’s dynamic world, contain costs, deal with IT skill shortages and take full advantage of new technologies. In short, manage your IT organization and infrastructure for business success. With its industry-leading flexibility, BladeCenter is the right choice for your dynamic business...
Learn about the IBM RackSwitch G8264CS which simplifies deployment with its innovative IBM Omni Port technology. For more information on IBM Systems, visit http://ibm.co/RKEeMO.
Visit the official Scribd Channel of IBM India Smarter Computing at http://bit.ly/VwO86R to get access to more documents.
Turbocharge the NFV Data Plane in the SDN Era - a Radisys presentationRadisys Corporation
On October 8, 2014, Karl Wale (Director of Product Management) and James Radley (Architect) presented: Turbocharge the NFV Data Plane in the SDN Era. This expert duo discussed the evolution of the network and service provider objectives around the challenges of deploying SDN/NFV solutions. They take you through some application use cases and introduce the new Radisys FlowEngine data plane software technology.
IT Brand Pulse industry brief describing a new approach to configuring virtual networks for virtual machines...layering hypervisor-based virtual networking services on top of hardware based virtual networking services. The result is more efficient management and lower costs.
Reduce complexity and costs associated with appliance sprawl by consolidating on an IBM® BladeCenter® platform with a high-performance, low-latency statistical load balancer integrated into a 10 Gigabit Ethernet (GbE) blade switch: IBM iFlow Director. iFlow Director’s traffic load...
Similar to Family data sheet HP Virtual Connect(May 2013) (20)
1. May 2013
Family data sheet
HP Virtual Connect
Connect systems to any network in a simple and flexible way
HP Virtual Connect Flex-10/10D Module
HP Virtual Connect 8 Gb 24-Port Fibre Channel Module HP Virtual Connect 8 Gb 20-Port Fibre Channel Module
HP Virtual Connect FlexFabric 10 Gb/24-Port Module
2. 2
Family data sheet | HP Virtual Connect
Move to a future-ready data center with a converged
infrastructure
What is a converged infrastructure?
Converged infrastructure helps you remove silos in your data center and integrates multiple
technologies into pools of interoperable resources to deliver operational flexibility for the
future. Built on leading innovation and industry standards, a converged infrastructure helps
you tackle the problem of IT sprawl by simplifying, integrating, and automating technology.
It gives you the ability to create and virtualize the compute, storage, and network environment
for any workload.
HP Converged Infrastructure has four pillars: HP Virtual Resource Pools, HP Data Center
Smart Grid, HP Matrix Operating Environment, and HP FlexFabric. HP Virtual Connect (VC) is a
cornerstone of the HP FlexFabric architecture. It is a common, virtualized network fabric that
connects servers to networking and storage while simplifying and increasing flexibility from the
data center to the network edge.
Challenges such as network sprawl at the server edge, explosive growth of virtual machines,
and increase in networking infrastructure costs can slow down your ability to respond to
change. HP VC simplifies the server edge by addressing the network sprawl. HP VC Flex-10/10D
and HP VC FlexFabric modules enable you to allocate up to four connections per server network
port and provide management and fine-tuning of bandwidth.
Overprovisioning of network capacity consumes precious budget, while underprovisioning
creates bottlenecks that can lower virtual machine performance. You are no longer locked-in
to either 1 Gb or 10 Gb allocations for a workload. Allocating too much or too little bandwidth
creates inefficiencies that can cost you money and time.
VC Flex-10/10D and VC FlexFabric modules along with HP Flex-10/FlexFabric adapters allow you
to adjust bandwidth to your changing workload—whenever you need to.
Figure 1: Business challenges of today
Traditionally, there were only two ways to connect server blades to outside networks and
storage—pass-thru modules or switches. The pass-thru approach though simple, leaves you
with countless cables and associated cabling costs, reliability concerns, and the risk of human
errors. Blade switches reduce the number of required cables but can lead to more switches
and increased management overheads. And with either approach, it’s hard to separate server
management from network and storage management.
HP VC technology changes all of this. It is the next step in virtualization—helping IT fully realize
the benefits of a converged infrastructure. It extends the benefit of virtualization beyond the
server to the rest of your IT infrastructure. VC simplifies your IT infrastructure by virtualizing
server-to-network connections.
HP offers the Virtual Connect FlexFabric 10 Gb/24-Port Module and three FlexFabric
adapters—the HP NC551m and HP NC551i Dual Port FlexFabric 10 Gb Converged Network
adapters, and HP NC553i 10 Gb Dual Port FlexFabric Converged Network Adapter (CNA) and
VC FlexFabric integrated adapters and mezzanine adapters. These include the HP NC551i
Integrated FlexFabric adapters for the AMD-based HP ProLiant G7 blade servers and the
HP NC553i Integrated FlexFabric adapters for the Intel®
-based ProLiant G7 blade servers.
Are these your challenges too?
IT sprawl Limited
budget
Restricted
bandwidth
3. 3
We also offer the HP NC551m and HP NC553m FlexFabric mezzanine adapters to provide
VC FlexFabric module connectivity for ProLiant G6 blade servers or to provide ProLiant
G7 blade servers with additional I/O ports. HP VC FlexFabric modules and FlexFabric adapters
extend Flex-10 technology to include Fibre Channel over Ethernet (FCoE) and accelerated iSCSI.
We continue to advance simpler network infrastructure and server connection management
with HP VC FlexFabric modules and adapters. By applying industry-standard convergence
technology, you can now address issues of network sprawl, without disrupting your LAN and
SAN setup.
Delivering low TCO and superior ROI
The HP Virtual Connect FlexFabric 10 Gb/24-Port Module allows you to select a single
interconnect that connects servers to data and storage networks, while reducing the time to
evaluate, purchase, install, update, and manage network infrastructure at the server edge.
It helps you break down silos that prevent you from gaining more value from your physical
IT resources. As more IT resources are virtualized, you can increase utilization—and reduce
your TCO.
HP VC solutions: VC Converged Networking
Key features and benefits
Ease of operations and deployment
• Server blades are change ready—you can add, move, replace, or upgrade server blades, and
move workloads without affecting your LAN and SAN.
• You can wire once, then add, move, and change network connections in minutes instead of
days, managing connections to thousands of servers from one pane of glass.
• Our modules are completely compatible with existing data and storage networks, protocols,
switches, and procedures.
• We simplify your network by providing one module for all data and storage connection needs.
• VC FlexFabric and Flex-10/10D modules are compatible with all other standards-based switch
products. They provide high-performance and end-to-end optical or copper connections with
HP Networking or other brands of core switches.
• VC FlexFabric and Flex-10/10D modules appear as pass-thru devices to the network. Any
change to the server is transparent to its associated network. This clearly separates the
server blades from your LAN and relieves LAN administrators from server maintenance.
• FlexFabric adapters support hypervisor best-practice configurations with six HP FlexNIC and
two HP FlexHBA standard connections on each blade.
• HP VC FlexFabric supports direct attach to FC storage with HP 3PAR Storage Systems, thereby
removing SAN fabric between servers and HP 3PAR Storage Arrays. This brings in operational
simplicity and a reduction in your TCO while connecting to FC storage.
Enterprise-class performance and availability
• Manage with 95 percent1
fewer network cards, switches, and cables to buy, install, qualify,
and maintain. Additionally, achieve significant reduction in power and cooling costs as well as
equipment cost with VC Converged Networking.
• Extend the life of your network by only using the capacity you need.
• Enable increased uptime with high-availability features, such as NIC teaming, trunk failover,
and dual redundant VC FlexFabric modules.
• Get direct uplink connections to LAN and SAN to maintain your current IT practices, unlike
other solutions that reroute SAN traffic to the LAN.
1
HP Internal: Based on HP analysis of
networking equipment (adapters and enclosure
interconnects), as shown here: A traditional
server blade configuration would typically require
a dual port LAN on Motherboard (LOM), an extra
quad-port NIC mezzanine, and a dual port host
bus adapter (HBA), along with switch modules
(six Ethernet and two Fibre Channel).
The total traditional configuration components
come up to 40 vs. the VC FlexFabric module
solution. The VC FlexFabric module solution
requires only embedded dual port FlexFabric
adapters on servers (no mezzanine cards) and
two VC FlexFabric modules. Here’s a simple
calculation to understand this better:
(40-2)/40 = 95 percent.
Family data sheet | HP Virtual Connect
4. 4
• Benefit from built-in standards-based data center connectivity with features such as
port-based VLANs, VLAN tagging, Internet Group Management Protocol (IGMP) Snooping,
N_Port ID Virtualization (NPIV), and uplink port aggregation with up to 1,000 VLANs per shared
uplink set.
• Fine-tune the performance of each data and storage connection to meet the needs of each
virtual machine and workload.
Security and management
• The embedded Virtual Connect Manager (VCM) Web-based console is present in each VC
FlexFabric and Flex-10/10D module. With this, you can define available LANs, SANs, server
connections, and manage server connection profiles for individual HP BladeSystem enclosures.
• In keeping with its use of industry-standard protocols, we are expanding the SNMP capabilities
of HP VC. Specifically, the latest VC firmware now supports both SNMP v1 and SNMP v2 traps;
traps for key, predefined threshold conditions; and per-destination configuration of traps.
• Role-based privileges for the administrator account are defined by default and can be modified
by the BladeSystem administrator and integrated with Lightweight Directory Access Protocol
(LDAP) servers.
• Role-based customization for server, network, and storage admin to tailor privileges
for administrators.
• QoS features allow administrators to configure traffic queues for different priority network
traffic, categorize and prioritize ingress traffic, and adjust DOT1P priority settings on
egress traffic.
• For environments that have implemented TACACS+ and RADIUS protocols for security,
VC supports these protocols in addition to LDAP.
• Additional role-based privileges for user accounts can be created by domain, server blade,
networking, and storage.
• Multi-enclosure stacking enables all of the VC modules (up to four connected enclosures) to
function as a single VC domain.
HP Virtual Connect FlexFabric 10 Gb/24-Port Module
The first interconnect built for the demands of a converged infrastructure, HP Virtual Connect
FlexFabric 10 Gb/24-Port Module, provides server administrators with a simple, flexible way to
connect servers and virtual machines to any LAN and SAN. You can now add up to four times
more data and storage connections per blade with full 10 Gb Ethernet and 8 Gb Fibre Channel
performance. It extends VC Flex-10 to include one Fibre Channel or accelerated iSCSI storage
connection (FlexHBA) with three Ethernet connections (FlexNICs) per 10 Gb port. With this
VC module, you can replace four interconnect modules with one module to support Ethernet,
Fibre Channel, and iSCSI connections. What’s more, reduce NICs and HBAs in each server by
using HP NC551i Integrated FlexFabric Adapter or HP NC553i Integrated FlexFabric Adapter. It is
also possible to increase the number of FlexNICs and available bandwidth that can be used
by virtual machines by adding HP NC551m or HP NC553m FlexFabric mezzanine adapters for
ProLiant G6 or G7 blade servers.
HP VC Flex-10/10D Module
HP VC Flex-10/10D Module is a 30-Port, next-generation Flex-10 Module with 600 Gb
full-duplex bandwidth capacity. It has 10 dedicated SFP+ based uplinks that can operate
at 1/10GbE and four inter-stacking links. VC Flex-10/10D Module provides up to four times
the number of connections (FlexNICs) per server port, without increasing the number of NICs
or managed switches required to connect them, by using Flex-10 technology. By virtue of
the dual-hop feature, it extends VC Flex-10 to include one Fibre Channel or accelerated iSCSI
storage connection (FlexHBA) with three Ethernet connections (FlexNICs) per 10 Gb port.
With this VC module, you can replace four interconnect modules with one module to support
Ethernet, Fibre Channel, and iSCSI connections. This reduces overheads—cards, switches, and
cables—by 95 percentage parts over traditional Ethernet and SAN switches.
Family data sheet | HP Virtual Connect
5. 5
HP Virtual Connect Fibre Channel Module
Key features and benefits
Ease of operations and deployment
• With NPIV and HP Virtual Connect Fibre Channel technology, storage management is no longer
limited to a single HBA World Wide Name (WWN) on the physical server. NPIV provides the
ability to share a single physical Fibre Channel HBA port among multiple virtual ports, each
with its own unique identifiers, allowing control of virtual machines access to LUNs on per
virtual machine basis.
• Standards-based Virtual Connect Fibre Channel Module is compatible with other NPIV
standards-based switch products. NPIV allows you to scale, gaining immediate benefits
without having to add domain IDs. Consider a blade server environment, such as an
HP c7000 chassis, where there are Fibre Channel switches located at the back of the chassis.
By using NPIV on these switches, you can add them to your fabric without having to assign a
domain ID individually and that provides high-performance, end-to-end connections with your
available options of core switches.
• The Virtual Connect Fibre Channel modules appear as pass-thru devices to the network.
Any changes to the server blade are transparent to the associated network. This clearly
separates the blade servers from your SAN and relieves your SAN administrators from
server maintenance.
Enterprise-class performance and availability
• Storage resources can be provisioned and associated directly to a specific virtual machine in a
virtualized server environment.
• Separate storage resources for up to 128 virtual machines per blade.
• High-availability features such as dual modules and automatic fail-over provide
increased uptime.
• VC server blade profiles are shared and continually updated between high-availability pairs.
• Enhanced NPIV capability supports multiple virtual machines per server blade and provides a
separate storage resource to each virtual machine—up to 128 per server blade.
Security and management
Role-based privileges for the administrator account are defined by default and can be modified
by the BladeSystem administrator and integrated with LDAP servers.
HP Virtual Connect 8 Gb 20-Port Fibre Channel Module
HP Virtual Connect 20-Port Fibre Channel Module offers next-generation 8 Gb technology that
includes backward compatibility with 2 Gb and 4 Gb networks, to help you get the most out of
your technology investment.
While it appears as a pass-thru device to the network, it provides all the key benefits of
integrated switching including the high-performance 8 Gb uplinks to the SAN. The 8 Gb
Fibre Channel interconnect facilitates superior performance and server consolidation.
This Virtual Connect Fibre Channel Module combines the simplicity of a pass-thru module
with the high performance of an 8 Gb switch. Additionally, four external 8 Gb ports reduce
oversubscription when using 4 Gb HBAs.
HP Virtual Connect 8 Gb 24-Port Fibre Channel Module
You can consider HP Virtual Connect 8 Gb 24-Port Fibre Channel Module for the highest
port density in the Virtual Connect Fibre Channel line-up. This standards-based module is
compatible with all other NPIV standards-based switch products. This compatibility accounts
for high-performance and end-to-end connections with your available options of core switches.
The 8 Gb Fibre Channel interconnect enables greater performance and server consolidation.
Eight SAN-facing ports and 16-server ports help reduce oversubscription for high throughput
applications. Also, separate storage resources are available to each virtual machine—up to 255 per
server blade.
Family data sheet | HP Virtual Connect
6. 6
HP Virtual Connect Enterprise Manager
HP Virtual Connect Enterprise Manager (VCEM) is a management option for multiple
BladeSystem enclosures configured with VC. VCEM provides a central console that aggregates
VC resources, improves productivity, and enables faster response to changing data center
workload demands.
The optional VCEM provides a single console to manage up to 1,000 BladeSystem enclosures.
With the VCEM for BladeSystem environments, you can complete many tasks across the
data center quickly and reliably. Some of such tasks are:
• Adding new VC domains, blade servers, and BladeSystem enclosures
• Recovering problematic servers and associated workloads to spare blades
• Migrating and repurposing servers to address changing workload requirements quickly
• Administering LAN and SAN addresses more efficiently and avoiding costly duplications
• Defining configurations for groups of VC domains
• Reducing repetitive, labor-intensive tasks
• Responding faster to changing workload needs across the data center
• Performing more efficient system maintenance
Together, VC and VCEM can create a change-ready infrastructure that enables system
administrators to add, replace, and recover servers across the data center without impacting
network configurations and availability.
Role-based privileges for the administrator account are defined by default and can be modified
by the BladeSystem administrator. VCEM is now available, bundled with either a pair of VC
Flex-10/10D modules or VC FlexFabric modules.
Ethernet, Flex-10, and FlexFabric network adapters
The 1 Gb and 10 Gb Ethernet network adapters embedded on the server blades, and
the optional Ethernet mezzanine adapters are supported by the VC Ethernet modules.
Multiple mezzanine adapters can be added per server blade.
The following network adapters are supported by the VC Ethernet modules for
Gen8 blade servers:
• HP Ethernet 10 Gb 2-Port 560M Adapter: The HP Ethernet 10 Gb 2-Port 560M mezzanine
adapter features the latest generation 10GbE Intel 82599 controller. The HP 560M is
a low cost, low-power dual port PCIe v.2.0 x8 lane mezzanine adapter designed for in
use HP ProLiant BL-c servers. The HP 560M addresses the demanding needs of the
next-generation cloud and data center by providing unmatched features for virtualization,
scalability to keep up with the Gen8 server platform, with proven and reliable performance
HP customers have come to expect. The 560M supports high performance networking
features such as VLAN tagging, Low latency interrupts, TCP and UDP checksum offloading,
MSI-X, NIC teaming (bonding), Receive Side Scaling (RSS), WOL, jumbo frames, PXE boot,
intelligent offloads, and advanced virtualization features (Intel VT-c NetQ, VMQ) and is
SR-IOV ready.
Family data sheet | HP Virtual Connect
7. 7
• HP Flex-10 10 Gb 2-Port 530M Adapter: HP Flex-10 10 Gb 2-Port 530M Adapter is a
dual port, 10GbE adapter, featuring the next-generation, single-chip 10 Gb Ethernet
solution from Broadcom in a mezzanine form factor designed for HP BladeSystem c-Class
Gen8 servers. This adapter provides high-performance Ethernet connectivity with support
for HP Virtual Connect Flex-10 interconnect technology, allowing the benefits of virtualization
to extend beyond the server and into the rest of the infrastructure. This adapter supports
enterprise-class features such as VLAN tagging, adaptive interrupt coalescing, MSI-X, NIC
teaming (bonding), Receive Side Scaling (RSS), jumbo frames, and PXE boot; and virtualization
features such as VMware NetQueue and Microsoft®
VMQ. Support for HP Sea of Sensors 3D
Technology enhances server performance while reducing energy use and expense.
• HP Flex-10 10 Gb 2-Port 552M Adapter: HP Flex-10 10 Gb 2-Port 552M Adapter, a mezzanine
Type A card designed for the HP BladeSystem c-Class Gen8 platform, delivers the performance
benefits of 10 Gb Ethernet with the scalability and flexibility of HP Virtual Connect and
Flex-10. This adapter supports up to 8 FlexNICs per adapter, with each FlexNIC with a unique
MAC address and identity on the LAN. Bandwidth can be uniquely allocated for virtual server
functions and applications. Hardware acceleration and offloads for stateless TCP/IP and TCP
Offload Engine (TOE) enable ideal use of CPU resources.
• HP FlexFabric 10 Gb 2-Port 554M Adapter: HP FlexFabric 10 Gb 2-Port 554M Adapter
combines the functionality of Ethernet and Fibre Channel (FC) onto a single mezzanine
Type A form factor designed for the HP BladeSystem c-Class Gen8 platform. It optimizes
network and storage traffic with hardware acceleration and offloads for stateless TCP/IP, TOE,
and Fibre Channel. This adapter utilizes next-generation technology, based on the Emulex
BE3 chipset that brings together Ethernet, Fibre Channel, and iSCSI on one adapter.
• HP Ethernet 10 Gb 2-Port 560FLB Adapter: The HP 560FLB is a dual port 10 Gb Ethernet
FlexibleLOM mezzanine adapter for select HP ProLiant BladeSystem c-Class Gen8 servers.
Based on the Intel 82599 controller, it provides customers an upgrade path. Customers
already using a 1 Gb FlexibleLOM can upgrade to the 10 Gb 560FLB, protecting their server
investment. New features supported with the HP 560FLB include enhanced Virtualization
(VMQ, NetQueue, VMDq, SRIOV), new I/O innovations with Intel Integrated I/O and Data Direct
I/O, and IEEE 1588. The 560FLB ships with advanced server features including support for
failover and load balancing, TCP/IP checksum offloading, large send offloading, Wake-on-LAN,
PXE, jumbo frames, VLAN tagging, and EFI/uEFI.
• HP Flex-10 10 Gb 2-Port 530FLB Adapter: HP Flex-10 10 Gb 2-Port 530FLB Adapter is a
dual port, 10GbE adapter, featuring the next-generation, single-chip 10 Gb Ethernet solution
from Broadcom in a FlexibleLOM form factor designed for BladeSystem c-Class Gen8 servers.
This adapter provides high-performance Ethernet connectivity with support for HP Virtual
Connect Flex-10 interconnect technology, allowing the benefits of virtualization to extend
beyond the server and into the rest of the infrastructure. Flex-10 10 Gb 2-Port 530FLB Adapter
supports enterprise-class features such as VLAN tagging, adaptive interrupt coalescing,
MSI-X, NIC teaming (bonding), Receive Side Scaling (RSS), jumbo frames, and PXE boot; and
virtualization features such as VMware NetQueue and Microsoft VMQ. Support for HP Sea of
Sensors 3D Technology enhances server performance while reducing energy use and expense.
• HP FlexFabric 10 Gb 2-Port 554FLB Adapter: HP FlexFabric 10 Gb 2-Port 554FLB Adapter
provides Ethernet functionality (Flex-10), Ethernet and iSCSI, or FCoE or FlexFabric connectivity
in a FlexibleLOM form factor designed for BladeSystem c-Class Gen8 platform. This adapter
delivers Ethernet functionality (Flex-10), or Ethernet and iSCSI, or FCoE or FlexFabric—at
the price of 10GbE. This adapter utilizes next-generation technology, based on the Emulex
BE3 chipset that brings together Ethernet, Fibre Channel, and iSCSI on one adapter.
Family data sheet | HP Virtual Connect
8. 8
The following network adapters are supported by the VC Ethernet modules for
G7 blade servers:
• HP NC532m Dual Port Flex-10 10GbE Multifunction BL-c Adapter: HP NC325m PCI Express
Quad Port Gigabit Server Adapter for c-Class BladeSystem provides up to eight discrete
physical NICs per mezzanine card.
• HP NC325m PCI Express Quad Port Gigabit Server Adapter for c-Class BladeSystem: The
NC325m mezzanine option features four Gigabit Ethernet ports providing the highest port
density available for BladeSystem server blades in a single adapter. The NC325m PCI Express
Quad Port Gigabit Server Adapter for c-Class BladeSystem is ideal for virtualization, security,
server consolidation, network segmentation, and other ProLiant server applications requiring
high throughput and port density.
• HP NC542m Dual Port Flex-10 10GbE Multifunction BL-c Adapter: HP NC542m Dual Port
Flex-10 10GbE Multifunction BL-c Adapter is based on the latest version of the Mellanox
ConnectX family. The HP NC542m delivers the high throughput and low latency perfect for
mission-critical applications such as financial services and virtualized environments.
• HP NC550m 10GbE 2-Port PCIe x8 Flex-10 Ethernet Adapter: HP NC550m 10 Gb 2-Port
PCIe x8 Flex-10 Ethernet Adapter is a member of the c-Class 10GbE adapter portfolio. This
adapter is a Flex-10 designed Ethernet Adapter based on the Blade Engines 2 controller
and delivers superior throughput in the 10GbE portfolio. This adapter is perfect for
virtualized environments.
• HP NC552m 10 Gb 2-Port Flex-10 Ethernet Adapter: HP NC552m 10 Gb 2-Port
Flex-10 Ethernet Adapter is a new member of the c-Class 10GbE adapter portfolio. This
adapter is a Flex-10 designed Ethernet Adapter based on the Blade Engines 3 controller
that delivers superior throughput compared to adapters in the 10GbE portfolio. Meant
for virtualized environments, this adapter works with all BladeSystem c-Class G7 servers.
When paired with any of our VC modules or Ethernet switches, the combination provides a
high-bandwidth, high-performance network environment.
• HP NC551m Dual Port FlexFabric 10 Gb Converged Network Adapter: HP NC551m Dual
Port FlexFabric 10 Gb Converged Network Adapter delivers the performance benefits and cost
savings of converged network connectivity for BladeSystem servers. This dual port mezzanine
adapter helps fine-tune network and storage traffic with hardware acceleration and offloads
for stateless TCP/IP, TOE, FCoE, and iSCSI. When connected to VC FlexFabric Module, FC, and
Ethernet, I/O are separated and routed to the corresponding network. For iSCSI storage,
the NC551m Dual Port FlexFabric 10 Gb Converged Network Adapter supports full-protocol
offload providing better CPU efficiency when compared to software initiators enabling the
server to handle increased virtualization workloads and compute-intensive applications.
This mezzanine adapter also supports VC Flex-10 that allows each 10 Gb port to be divided
into four physical NICs to help fine-tune bandwidth management for virtualized servers.
This combination of high-performance network and storage connectivity reduces cost and
complexity and provides the flexibility and scalability for BladeSystem servers.
• HP NC553m 10 Gb 2-Port FlexFabric Adapter: HP NC553m FlexFabric 10 Gb 2-Port
FlexFabric Adapter delivers the performance benefits and cost savings of converged network
connectivity for HP BladeSystem servers. This dual port adapter, based on the Blades
Engine 3 controller, helps fine-tune network and storage traffic with hardware acceleration
and offloads for stateless TCP/IP, TOE, FCoE, and iSCSI. When connected to VC FlexFabric
Module FC, and Ethernet, I/O are separated and routed to the corresponding network. For iSCSI
storage, this mezzanine adapter supports full protocol offload providing better CPU efficiency
when compared to software initiators enabling the server to handle increased virtualization
workloads and compute-intensive applications.
Family data sheet | HP Virtual Connect
9. 9
Integrated adapters
• NC553i: 10 Gb, Dual Port, FlexFabric Converged Network Adapter
• NC551i: 10 Gb FlexFabric embedded adapter
• NC532i: 10 Gb Flex-10 enabled Ethernet adapter
Fibre Channel host bus adapters
Seven add-on host bus adapters (HBAs) (mezzanine or daughter cards) are currently offered for
these BladeSystem c-Class server blades:
• HP QMH2572 8 Gb FC HBA for HP BladeSystem c-Class: This adapter is a dual port,
8 Gb, Fibre Channel HBA in a mezzanine form factor designed for HP BladeSystem c-Class
Gen8 servers. This adapter provides reliable SAN and is ideal for virtualized environments.
With 200,000 IOPS per port and leading-edge FC connectivity, users can enjoy the benefits of
increased storage networking bandwidth and high I/O performance to meet the requirements
of demanding applications. Support for HP Sea of Sensors 3D Technology enhances server
performance while reducing energy use and expense.
• HP LPe1205A 8 Gb FC HBA for HP BladeSystem c-Class: HP LPe1205A 8 Gb FC HBA for
BladeSystem c-Class delivers 8 Gb Fibre Channel (FC) connectivity and is compatible with
the Gen8 server blades for BladeSystem c-Class. With backward compatibility to 4 Gb FC
and 2G FC technology, this adapter offers flexibility in supporting existing FC infrastructure
while offering features and functionality designed to meet and exceed the business
requirements of next-generation data centers. With features such as comprehensive
virtualization support, efficient power usage, high reliability and availability features, and
robust manageability, this adapter is ideal for mission-critical applications that rely on high
availability and connectivity.
• Emulex LPe1105-HP 4 Gb FC HBA for HP BladeSystem c-Class: Emulex LPe1105-HP dual
port Fibre Channel HBA for HP BladeSystem c-Class provides reliable, high-performance 4 Gb
connectivity, enabling high availability to scalable storage. Based on the same field-proven
ASIC, firmware, and driver technology as Emulex’s renowned LPe1150 HBA, the Emulex
LPe1105-HP 4 Gb FC HBA for HP BladeSystem c-Class is fully driver compatible with all
Emulex HBAs.
• Emulex LPe1205-HP dual-channel PCI Express 8 Gb/s FC HBA for HP BladeSystem
c-Class: Emulex LPe1205-HP dual port PCI Express Fibre Channel mezzanine card HBA
provides high-performance SAN connectivity for BladeSystem c-Class to meet the needs
of most-demanding applications. Leveraging eight generations of design, this adapter
seamlessly attaches to 2, 4, and 8 Gb/s SAN devices. It provides superior reliability, enhanced
security, and scalable management capabilities—making it well-suited for virtual server and
enterprise-class applications.
• QLogic QMH2462 4 Gb FC HBA for HP BladeSystem c-Class: QLogic QMH2462 4 Gb FC
HBA for HP BladeSystem c-Class provides BladeSystem c-Class with two 4 Gb Fibre Channel
ports for fast and reliable SAN connectivity. Through the implementation of a SAN with
BladeSystem, you can achieve improved data availability, easily scale capacity, and realize
management cost savings from consolidating disk resources.
• QLogic QMH2562 8 Gb FC HBA for HP BladeSystem c-Class: Designed for the
HP BladeSystem c-Class, the QLogic QMH2562 8 Gb FC HBA for HP BladeSystem c-Class
provides reliable SAN connectivity and is ideal for virtualized environments. It delivers
twice the data throughput compared to the previous generation 4 Gb mezzanine card.
With 200,000 IOPS per port and leading-edge FC connectivity, you can enjoy the benefits of
increased storage networking bandwidth and I/O performance to meet the requirements of
demanding applications.
Family data sheet | HP Virtual Connect
10. 10
Technical specifications
HP Virtual Connect Fibre Channel
HP Virtual Connect FlexFabric
10 Gb/24-Port Module
HP Virtual Connect Flex-10
10 Gb Ethernet Module
HP Virtual Connect Flex-10/10D Module
Blade type Single bay Single bay Single bay
Network connections 16 x 10 Gb downlinks to servers
2 x 10 Gb cross connects
4 x 10 Gb external SR, LR fiber, and copper
uplinks SFP+ (Enet/FC)
4 x 10 Gb external SR, LRM, and LR fiber and
copper uplinks SFP+ (Enet)
1 internal interface to c-Class Onboard
Administrator Module
16 x 10 Gb downlinks midplane
2 x 10 Gb cross connect
1 x 10 Gb copper uplinks CX-4
8 x 10 Gb SR, LR, or LRM fiber uplinks SFP+
1 internal interface to c-Class Onboard
Administrator Module
16 x 10 Gb downlinks midplane
4 x 10 Gb cross connect
10 x 10 Gb SR, LR, or LRM fiber uplinks SFP+
1 internal interface to c-Class Onboard
Administrator Module
Media types FCSFP/SFP+
2/4/8 Gb short wave up to 500 m
1/2/4 Gb long wave up to 10 km
Ethernet SFP/SFP+
10GbE SR, LR, and LRM
10GbE copper direct-attached cable
1GbE SX
1GbE 1000BASE-T copper
HP 7 m C-series Active Copper SFP+ Cable
HP 10 m C-series Active Copper SFP+ Cable
HP X242 SFP+ 15 m DAC Cable
HP X242 SFP+ 7 m DAC Cable
SFP+ SR, LR, LRM SFP SX, R J-45
SFP+ Copper
Twinax CX-4 (IB4x)
HP 7 m C-series Active Copper SFP+ Cable
HP 10 m C-series Active Copper SFP+ Cable
HP X242 SFP+ 15 m DAC Cable
HP X242 SFP+ 7 m DAC Cable
SFP+ SR, LR, LRM SFP SX, R J-45
SFP+ Copper
HP 7 m C-series Active Copper SFP+ Cable
HP 10 m C-series Active Copper SFP+ Cable
HP X242 SFP+ 15 m DAC Cable
HP X242 SFP+ 7 m DAC Cable
Performance Line rate, full-duplex 480 Gb/s bridging fabric
1.2 µs on Ethernet only ports
1.7 µs Ethernet/FC ports
Maximum Ethernet frame size 9216 (Jumbo Frame)
Maximum FC frame size 2148 bytes
(2112 byte payload)
Buffer-to-buffer flow control management
Packet prioritization
Line rate, full-duplex 480 Gb/s bridging fabric
Less than 1.2 µs latency
Line rate, full-duplex 600 Gb/s bridging fabric
Less than 0.9 µs with Ethernet only ports
Maximum Ethernet frame size 9216 (Jumbo Frame)
Protocol support IEEE 802.1Qbb (preliminary), 802.1Qaz
(preliminary), 802.1AB, 802.1D, 802.1Q, IEEE
802.2, 802.3ad INCITS FC-BB5 Rev 2.00
INCITS T11 NPIV, FC-BB5
802.1AB, 802.1D, 802.1Q, IEEE 802.2,
802.3ad
802.1AB, 802.1D, 802.1Q, IEEE 802.2,
802.3ad, FC-BB5
Management Simple and intuitive GUI and setup wizards
embedded SNMP v1, v2; SMI-S port
mirroring—Any uplink port can be used as a
dedicated mirrored port from the server port(s)
Simple and intuitive GUI and setup wizards
embedded SNMP v1, v2; SMI-S port
mirroring—Any uplink port can be used as a
dedicated mirrored port from the server port(s)
Simple and GUI and setup wizards embedded SNMP
v1, v2; SMI-S CLI port mirroring—Any uplink port
can be used as a dedicated mirrored port from the
server port(s)
Extended management
features
Virtual Connect Manager supports PXE, WOL, port
VLAN, VLAN Tagging, VLAN pass through, IGMP
Snooping, NIC Teaming Integrated with Onboard
Administrator, HP Systems Insight Manager,
HP Storage Essentials (FC Management MIB)
Telnet, SNMP, FC port telemetry via GUI,
Telemetry support for port utilization including
memory and CPU performance measurement
Virtual Connect Manager supports PXE, WOL, port
VLAN, VLAN Tagging, VLAN pass through, IGMP
Snooping, NIC Teaming Integrated with Onboard
Administrator, HP Systems Insight Manager
Telnet, SNMP. Telemetry support for port
utilization including memory and CPU
performance measurement
Virtual Connect Manager supports PXE, WOL, port
VLAN, VLAN Tagging, VLAN pass through, IGMP
Snooping, NIC Teaming Integrated with Onboard
Administrator, HP Systems Insight Manager Telnet,
SNMP. Telemetry support for port utilization
including memory and CPU performance
measurement
High-availability
features
Link Aggregation Protocol Automatic
loop protection
Mirrored profile database
Multipath heartbeat between redundant modules
Link Aggregation Protocol Automatic
loop protection
Mirrored profile database
Multipath heartbeat between redundant modules
Link Aggregation Protocol Automatic
loop protection
Mirrored profile database
Multipath heartbeat between redundant modules
Security LDAP, SSL, TACACS+ and Radius, role-based
management, and GUI and CLI session timeout
LDAP, SSL, TACACS+ and Radius, role-based
management, and GUI and CLI session timeout
LDAP, SSL, TACACS+ and Radius, role-based
management, and GUI and CLI session timeout
Diagnostics Troubleshoot network performance and monitor
health in terms of CPU and memory, FlexNIC and
LAG stats
Troubleshoot network performance and monitor
health in terms of CPU and memory, FlexNIC and
LAG stats
Troubleshoot network performance and monitor
health in terms of CPU and memory, FlexNIC and
LAG stats
Maximum per enclosure 8 8 8
Warranty
(parts/labor/onsite)
1 year parts, 1 year labor, 1 year onsite 1 year parts, 1 year labor, 1 year onsite 1 year parts, 1 year labor, 1 year onsite
Part number 571956-B21 455880-B21
591973-B21 (dual module with VCEM)
638526-B21
662048-B21 (dual module with VCEM)
Rollback from upgrade Minimizes maintenance window Minimizes maintenance window Minimizes maintenance window
Direct attach with
FC storage
With HP 3PAR T, V, and F series Not applicable Not applicable
Family data sheet | HP Virtual Connect
11. 11
HP Virtual Connect Fibre Channel Module
HP Virtual Connect 8 Gb 20-Port Fibre Channel Module HP Virtual Connect 8 Gb 24-Port Fibre Channel Module
Blade type Single bay Single bay
Network connections 16 internal 8 Gb downlinks presented as F-Ports
4 external 8 Gb uplinks presented as N-Ports
16 internal 8 Gb downlinks presented as F-Ports
8 external 8 Gb uplinks presented as N-Ports
Media types Small form-factor pluggable (SFP) laser
2/4/8 Gb short wave up to 500 m (1,640 ft) 1/2/4 Gb
long wave up to 10 km
Small form-factor pluggable (SFP) laser
1/2/4 Gb short wave, long wave
SFP+ 2/4/8 Gb short wave, long wave
Performance 8 Gb/s line speed, full-duplex
1.2 μs latency
Maximum frame size 2112 byte payload
Buffer-to-buffer flow control management packet
prioritization
8 Gb/s line speed, full-duplex
.74 μs latency
Maximum frame size 2148 bytes (2112 byte payload)
Protocol support NCITS T11 NPIV NCITS T11 NPIV
Management Simple and intuitive GUI and setup wizards accessible
through VC Ethernet module
CLI accessible through VC Ethernet module
Embedded SNMP v1 and v2
SMI-S
Simple and intuitive GUI and setup wizards accessible
through VC Ethernet module
CLI accessible through VC Ethernet module
Embedded SNMP v1 and v2
SMI-S
Extended management features Virtual Connect Manager supports
HP Storage Essentials (FC Management MIB)
Virtual Connect Manager supports
HP Storage Essentials (FC Management MIB)
High availability features Link Aggregation Protocol Automatic loop protection
Mirrored profile database
Multipath heartbeat between redundant modules
Link Aggregation Protocol Automatic loop protection
Mirrored profile database
Multipath heartbeat between redundant modules
Security LDAP, SSL, and role-based management LDAP, SSL, and role-based management
Maximum per enclosure 6 6
Warranty (parts/labor/onsite) 1 year parts, 1 year labor, 1 year onsite 1 year parts, 1 year labor, 1 year onsite
Part number 572018-B21 466482-B21
Family data sheet | HP Virtual Connect
12. 12
Mezzanine and FlexibleLOM for HP BladeSystem c-Class
HP Ethernet 10 Gb 2-Port
560M Adapter
HP Flex-10 10 Gb 2-Port
530M Adapter
HP Flex-10 10 Gb 2-Port
552M Adapter
HP FlexFabric 10 Gb 2-Port
554M Adapter
IEEE compliance 802.3, 802.1ab, 802.3x, 802.3ad,
802.3p/802.1q, 802.3ae,
802.1qau, 802.3ap, 802.1as,
802.1qaz, 802.1qbb
802.3, 802.3ab, 802.3u,
802.3x, 802.3ad, 802.3p,
802.1q, 802.3ae, 802.3ap
802.1p, 802.1q, 802.1qau,
802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x
802.1p, 802.1q, 802.1qau,
802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x
Bus type x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0
Form factor Type A mezzanine Card Type A mezzanine card Type A mezzanine card Type A mezzanine card
Ports and
transfer rate
20,000 Mb/s full-duplex
Ethernet transfer rate per port
20,000 Mb/s full-duplex
Ethernet transfer rate per port
20,000 Mb/s full-duplex
Ethernet transfer rate per port
20,000 Mb/s full-duplex
Ethernet transfer rate per port
Network controller Intel 82599 Broadcom 57810S Emulex BE3 Emulex BE3
ProLiant adapter teaming* • • • •
ProLiant essentials upgrade • • • •
Pre-boot execution
environment (PXE)
• • •
Wake-on-LAN (WOL) • (1 Gb) • • •
Jumbo frames • • • •
TCP checksum and segmentation • • • •
Receive Side Scaling (RSS) • • • •
TCP/IP offload engine (TOE) • • •
Accelerated iSCSI •
Remote direct memory
access (RDMA)
Not supported Not supported Not supported
Warranty (parts/labor/onsite) 1-year parts** 1-year parts** 1-year parts** 1-year parts**
Part number 665246-B21 631884-B21 674764-B21 647590-B21
* Teaming available with Windows®
except Windows 2012
** Or the warranty of the server the adapter is installed in, whichever is greater.
Family data sheet | HP Virtual Connect
13. 13
Ethernet network adapters—mezzanine adapter options
HP Ethernet 10 Gb 2-Port
560FLB Adapter
HP Flex-10 10 Gb 2-Port
530FLB Adapter
HP FlexFabric 10 Gb 2-Port
554FLB Adapter
NC325m Quad Port Gigabit
Ethernet Adapter
NC532m Dual Port
10GbE Multifunction
Adapter
IEEE compliance 802.3, 802.1ab, 802.3x,
802.3ad, 802.3p, 802.1q,
802.3ae, 802.1au, 802.3ap,
802.1as, 802.1qaz, 802.1qbb,
IEEE 1588
802.3, 802.1ab, 802.3x,
802.3ad, 802.3p, 802.1q,
802.3ae, 802.1au, 802.3ap
802.1p, 802.1q, 802.1qau,
802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x
802.1p, 802.1q, 802.3,
802.3ad, and 802.3x
802.3u, 802.3x, 802.3ad,
802.1p, 802.1q, 802.3z,
802.3ap, 802.3ae
Bus type x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0 x4 PCI Express x4 PCIe 2.0
Form factor FlexibleLOM FlexibleLOM FlexibleLOM Type 1 mezzanine card Type 1 mezzanine card
Ports and transfer rate 20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
4 @ 1000 Mb/s 2 @ 10000 Mb/s
Network controller Intel 82599 Broadcom 57810S Emulex BE3 Dual Broadcom 5715S Broadcom 57711
ProLiant adapter teaming* • • • • •
ProLiant essentials
upgrade
• • • • •
Pre-boot execution
environment (PXE)
• • • •
Wake-on-LAN (WOL) • (1 Gb) • • • •
Jumbo frames • • • • •
TCP checksum and
segmentation
• • • • •
Receive Side Scaling (RSS) • • • •
TCP/IP offload engine (TOE) • • •
Accelerated iSCSI •
Remote direct memory
access
Not supported Not supported Add-on upgrade
Warranty (parts/labor/
onsite)
1-year parts** 1-year parts** 1-year parts** 1 year parts** 1 year parts**
Part number 655639-B21 656590-B21 647586-B21 416585-B21 467799-B21
* Note: Teaming available with Windows except Windows 2012
** Note: Or the warranty of the server the adapter is installed in, whichever is greater.
Family data sheet | HP Virtual Connect
14. 14
Ethernet network adapters—mezzanine adapter options
* Teaming available with Windows except Windows 2012
NC542m Dual Port
10GbE Multifunction
BL-c Adapter
NC550m Dual Port
10 Gb Flex-10
Ethernet Adapter
NC551m Dual Port
FlexFabric 10 Gb
Converged Network
Adapter
NC552m Dual Port
10 Gb 2-Port Flex-10
Ethernet Adapter
NC553m Dual Port
FlexFabric 10 Gb
Converged Network
Adapter
IEEE compliance 802.1p, 802.1q, 802.3u,
802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x, and 802.3z
802.1p, 802.1q, 802.1qau,
802.3u, 802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x, and 802.3z
802.1p, 802.1q, 802.1qau,
802.3u, 802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x, and 802.3z
802.1p, 802.1q, 802.1qau,
802.3u, 802.3ad, 802.3ae,
802.3ap (10GBASE-KX4),
802.3x, and 802.3z
802.1p, 802.1q, 802.1qau,
802.3u, 802.3ad,
802.3ae, 802.3ap
(10GBASE-KX4), 802.3x,
and 802.3z
Bus type x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0 x8 PCIe 2.0
Form factor Type 1 mezzanine card Type 1 mezzanine card Type 2 mezzanine card Type 1 mezzanine card Type 1 mezzanine card
Ports and transfer rate 20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
20,000 Mb/s full-duplex
Ethernet transfer rate
per port
Network controller Mellanox ConnectX-2 EN Emulex BE2 Emulex BE2 Emulex BE3 Emulex BE3
ProLiant adapter teaming* • • • • •
ProLiant essentials
upgrade
• • • • •
Pre-boot execution
environment (PXE)
• • • • •
Wake-on-LAN (WOL) • • • • •
Jumbo frames • • • • •
TCP checksum and
segmentation
• • • • •
Receive Side Scaling (RSS) • • • • •
TCP/IP offload engine (TOE) • • • • •
Accelerated iSCSI • • • • •
Remote direct memory
access
Not supported Not supported Not supported Not supported Not supported
Warranty (parts/labor/
onsite)
1 year parts 1 year parts 1 year parts 1 year parts 1 year parts
Part number 539857-B21 581204-B21 580151-B21 610609-B21 613431-B21
Family data sheet | HP Virtual Connect
15. 15
Fibre Channel HBA—mezzanine cards
Emulex LPe1105-HP
4 Gb FC HBA for
HP BladeSystem c-Class
HP QLogic QMH2462
4 Gb FC HBA for
HP BladeSystem c-Class
Emulex LPe1205-HP
dual-channel PCI Express
8 Gb/s FC HBA for
HP BladeSystem c-Class
QLogic QMH2562
8 Gb FC HBA for
HP BladeSystem c-Class
Performance and form factor
Blade type PCI Express Type 1
mezzanine card
PCI Express Type 1
mezzanine card
PCI Express Type 1
mezzanine card
PCI Express Type 1
mezzanine card
Performance 115,000 IOPS per port 150,000 IOPS per port 200,000 IOPS per channel 200,000 IOPS per channel
Port configuration Dual 4 Gb Fibre Channel ports Dual 4 Gb Fibre Channel ports Dual 8 Gb Fibre Channel ports Dual 8 Gb Fibre Channel ports
Media types 62.5/125 multimode fiber optic
cable with LC type connector
62.5/125 multimode fiber optic
cable with LC type connector
62.5/125 multimode fiber optic
cable with LC type connector
62.5/125 multimode fiber optic
cable with LC type connector
Management and protocols
Management features Emulex installation and
management tools automate
installation and provide local and
remote HBA configuration and
management
QLogic SANsurfer FC HBA Manager
for centralized management and
remote control of distributed
HBAs
Emulex installation and
management tools automate
installation and provide local and
remote HBA configuration and
management
QLogic SANsurfer FC HBA
Manager for centralized
management and remote control
of distributed HBAs
High-availability features Multipath support for redundant
HBAs and paths
Multipath support for redundant
HBAs and paths
Multipath support for redundant
HBAs and paths
Multipath support for redundant
HBAs and paths
Protocols supported Full support for both
FC service Class 2 and 3
Full support for both
FC service Class 2 and 3
Full support for both
FC service Class 2 and 3
Full support for both
FC service Class 2 and 3
Deployment
Maximum per enclosure Server dependent Server dependent Server dependent Emulex driver
support for Windows Server
2003, Windows Server 2008,
Linux, and VMware ESX 3.5
Server dependent QLogic driver
support for Windows, Linux,
HP-UX, and VMware
Options available Emulex driver support for x86
and x64 Linux and Microsoft
Windows Server 2000, 2003,
and 2003 x64
QLogic driver support for x86 and
x64 Linux and Microsoft Windows
Server 2000, 2003, and 2003 x64
Emulex driver support for Server
2000, 2003, and 2003 x64
QLogic driver support for Server
2000, 2003, and 2003 x64
Warranty
(parts/labor/onsite)
1 year parts, 1 year labor,
1 year onsite
1 year parts, 1 year labor,
1 year onsite
1 year parts, 1 year labor,
1 year onsite
1 year/1 year/1 year
Part number 403621-B21 403619-B21 456972-B21 451871-B21590647-B21
Family data sheet | HP Virtual Connect