Please join us for our free webcasts. Many of these sessions were also presented at the SHARE, Computer Measurement Group (CMG), National Systems Programmer Association (NASPA), or SharkFest conferences. Each session will be approximately 40 minutes long. Following the webcast, we will provide an open forum for participants to discuss the topic for 10 minutes. To get a notification to register for our webcasts, please sign our guestbook at Guestbook. You will receive a new email to register for each webcast. Please check back often, more webcasts will be announced soon.

The sessions will be presented by the CEO and Founder of Inside Products, Nalini Elkins, or other invited speakers. Nalini is a frequent speaker and invited guest speaker at many technical conferences. To view her biography, please click on: Nalini's Biography.

You may also wish to join the IPv6 Business Information Exchange. This is a free forum provided by Inside Products. We will have web assisted meetings every quarter. For the meeting schedule, topics to be discussed, and mission of the IPv6 Business Exchange, please click here

IPv6 Businress Information Exchange
Octber 10, 2012: 8:00AM PST, 10:00PM Central, 11:00PM Eastern, 4:00PM London

This month's webcast will be the quarterly meeting of the IPv6 Business Information Exchange. Please check that page for details.

You may be interested in some of the webcasts we have given in the past. If you are interested in the foils and / or audio from these sessions, please email us and let us know what you would like. Audio is available for some of these past webcasts.

September 13, 2012
IPv6 Transitions

Enterprises are moving towards IPv6 addressing for the host of advantages it offers. However, choosing the right transition mechanisms to support co-existence of the incompatible IPv4 and IPv6 protocols requires awareness of such techniques along with their benefits and downfalls.

We reviewed your options for IPv6 adoption, including the different transition techniques such as translation, tunneling and dual-stack mode, and how to handle potential security hazards.

You can download the foils here.

August 21, 2012
Introduction to IPv6 Addressing

IPv6 adoption is the buzz across Enterprises of all sizes across the globe. The World is running out of IPv4 addresses and IPv6 implementation is becoming a major concern.

IPv6 is much more than a larger address. Yes, the IP address has changed from 32 bits to 128 bits but what is more important is that address planning, the kinds of addresses, and how DHCP and DNS are to be used and configured have all been changed. New concepts and protocols such as Stateless Address Autoconfiguration (SLAAC), Neighbor Discovery, Router Discovery, and Multicast Listener Discovery protocols must be understood. Many new ICMP messages have been introduced leading to changes in how firewalls policies are to be done.

The recording of the webcast is available here. Or you can just download the foils here.

May 30, 2012
Complications with Digital Certificates

Secure Sockets Layer (SSL) is being widely implemented to protect TCP application traffic. Part of the security relies on digital certificates. In the past, only server certificates were used. Now, more and more companies are requiring client certificates. This complicates the handshake process and poses a number of new issues in management and control.

This session discussed:

• a SSL handshake with only a server certificate,
• a SSL handshake with both server and client certificates,
• managing certificate authorities and bad certificates, and
• potential performance issues.

We started by understanding the cryptographic concepts that underlie digital certificates. Our new Certificate Checker product was used to illustrate how certificates appear in traces.

You can download the foils at Complications with Digital Certificates.

March 21, 2012

This webcast discussed how the power of the mainframe can be harnessed to analyze TCP packet flow. Our IP Problem Finder products have been turning packet traces into English for years. This set of products imports an IP packet trace (CTrace or WireShark), analyzes and correlates the packet flow, and then tells you in an English language report what went wrong.

We have now made much of the logic of the IP Problem Finders available on the mainframe, too. Customers were telling us that they want to read in many hundreds of thousands packets - even millions of packets! What better place to do that than the mainframe?

In this webcast, we showed how we use the power of the mainframe to quickly analyze trace flow. We will also discuss several problem situations and how IP Problem Finder can help pinpoint problems.

• Comparing capacity before and after tuning changes
• Checking whether TCP segmentation offload is being done
• Checking whether TCP window scaling is being used

You can download the foils at IP Problem Finder DB2.

February 15, 2012
Selective Acknowledgments (SACK) and Bytes in Flight (BIF)

This webcast looked at some interesting topics in TCP/IP performance: Selective Acknowledgments (SACK) and Bytes in Flight (BIF). RFC2018 defines SACK as follows:

"TCP may experience poor performance when multiple packets are lost from one window of data. With the limited information available from cumulative acknowledgements, a TCP sender can only learn about a single lost packet per round trip time. An aggressive sender could choose to retransmit packets early, but such retransmitted segments may have already been successfully received.

A Selective Acknowledgment (SACK) mechanism, combined with a selective repeat retransmission policy, can help to overcome these limitations. The receiving TCP sends back SACK packets to the sender informing the sender of data that has been received. The sender can then retransmit only the missing data segments."

We first viewed traces with Selective Acknowledgments, out of order and retransmitted packets. Then we moved on to discuss:

• What do Bytes In Flight calculations really tell us?
• How does BIF relate to window size?
• Is BIF a good way to assess and compare the throughput of sessions?

You can download the foils at SACK and Bytes in Flight

December 15, 2011
Data Mining to Tune TCP/IP

This webcast discussed how to turn all that data in SMF or provided by TCP/IP monitors into real information. As more and more applications use TCP/IP, it is important to know how the additional usage impacts the overall system. We discussed What causes overhead? How to find problems and pinpoint traffic that may not be necessary. What can be done about them?
Some real examples of problems we have found include a printer sending an error message once a millisecond or a load balancer starting a connection once a second to four different DB2 systems.

October 13, 2011
IPv6 and the IETF

At the IPv6 Business Information Exchange meeting we discussed how enterprise customers can become more involved with the Internet Engineering Task Force (IETF). We also talked about our participation at the last IETF (IETF 82 in Quebec) and a draft RFC that members of the IPv6 BIE proposed at that meeting.

The decisions made by the IETF impact all enterprise networks. We should be there to present our viewpoints and to make our concerns known. In our discussions with them, the IETF has made it clear that they welcome our participation.

Speakers
Mike Ackermann: Blue Cross Blue Shield Michigan
Nalini Elkins: Inside Products, Inc.

September 7, 2011
Understanding DNS Configuration and Performance

Whether you are using IPv4 or starting the migration to IPv6, DNS performance is critical! Poor DNS configuration can add many seconds of delay to response time! Some DNS servers are set up to do both IPv6 and IPv4 queries with each request. Some DNS servers are using multicast. Do you know the performance of this critical resource at your installation?.

Take a look at some of these DNS analysis reports:
http://www.insidethestack.com/dns.html
In this session, we started by understanding how DNS works and then moved on to discuss performance issues. The topics included:

• DNS configuration files
• DNS commands
• DNS messages
• Changes to DNS for IPv6
• Calculating DNS response times

March 23, 2011
SSL Handshake Analysis

SSL is being widely implemented to protect TCP application traffic. How can you diagnose problems? If the connection setup and negotiation causes poor response time, can you tell? In this session, we will discuss how to:

• Diagnose SSL handshake problems,
• Assess performance implications,
• View certificates, certificate authorities and bad certificates, and
• Decode encrypted packet data.

We will demonstrate how SSL Problem Finder can automatically analyze handshake problems.

Problems may include:

• Incorrect cipher suite,
• Performance issues,
• Bad or incorrect certificates, and more.

For some sample automatically generated recommendations from SSL Problem Finder, please view: Bad Analysis or Timing Analysis

February 23, 2011
Saving Money by Tuning TCP

Many problems occur on TCP/IP networks. Tuning TCP can save mainframe CPU MIPs. At one installation, we were able to drop the CPU usage of the TCP/IP stack by a third.

You may be able to delay upgrades while getting better throughput for your TCP/IP network. The mainframe is one of the most critical and expensive resources in the data center. Unnecessary overhead for the TCP stack is something that can be eliminated.

Nalini Elkins and Jim Ashton show what we have done with our Network Health Check at various companies, as well as the results of these efforts.

December 16, 2010
Monitor PCI Compliance

SSL and AT-TLS are being widely implemented to protect TCP application traffic, and for PCI compliance. What tools do you need to monitor and diagnose problems for such connections? Our Early Warning System product provides alerts in real time for:

• AT-TLS handshake failure
• SSL or AT-TLS connection high network response time
• SSL or AT-TLS congestion window problems

Make sure that the security you have put in place for PCI complaince is always there! Use the Early Warning System to alert you to AT-TLS handshake failures.

Our Connection Log product provides logging so you can find:

• Users who are not using secure handshakes
• Users who are using the wrong encryption, or
• Users who are not using encryption at all!

Then, when you have problems with a client and need to diagnose connection setup and negotiation, you need our SSL Problem Finder product. Sith this product, you can:

• Diagnose SSL handshake problems
• Assess performance implications
• View certificates, certificate authorities, bad certificates, and
• Decode encrypted packet data

August 25, 2010
Continuous TCP Health Check

The Early Warning System provides a continuous monitoring of the health of any or all of the TCP connections for mainframe TCP/IP. This is the foundation of a true self-managing, self-healing network.

Many problems occur on TCP/IP networks. Among them are:

• Connection drops,
• Congestion window,
• Round trip time,
• Round trip variance,
• Retransmissions,
• Duplicate acknowledgments,
• Bad statuses, and
• Hung connections.

The Early Warning System provides a ‘heads-up’ for TCP connections or applications which may be heading for trouble. That is, they may be currently experiencing performance problems or will soon be experiencing performance problems.

May 18, 2010
Automated TCP Problem Analysis

Many networks are becoming primarily TCP/IP. As always, for difficult network diagnostic problems, the only option is to take a trace. Reading and analyzing a trace is quite difficult - it requires many years of training.

What if you could have recommendations and trace analysis provided for you automatically? Is this really possible? In this session, we will demonstrate how TCP Problem Finder analyzes many common and not so common problems including:

• Firewall configuration,
• Performance bottlenecks,
• Congestion window (zero, low), and much more.

We expect the TCP Problem Finder to cut problem diagnosis time for a typical trace by a factor of 8. That is, a trace which might take 8 hours to resolve manually, we expect will take one hour using TCP Problem Finder.

April 20, 2010
IPv6 Business Information Exchange Meeting: Bechtel IPv6 Addressing

Discussion with Fred Wettling, a Bechtel Fellow and co-author of Global IPv6 Strategies, and Dr. Pete Welcher of Chesapeake NetCraftsmen. Bechtel has been implementing IPv6 for a while and Fred is integral to this effort. We will discuss the following:

• Brief history / status of Bechtel IPv6 migration.
• Address planning
• Migration pitfalls
• What were the biggest stumbling blocks?
• What might we do differently now that we know more?
• How do we see IPv6 fitting into the next 5 years of network strategic planning?

Following Fred's presentation, the group will have a general discussion to absorb and ponder the ideas.

March 23, 2010
SSL Handshake Analysis

SSL is being widely implemented to protect TCP application traffic. How can you diagnose problems? If the connection setup and negotiation causes poor response time, can you tell? In this session, we will discuss how to:

• Diagnose SSL handshake problems,
• Assess performance implications,
• View certificates, certificate authorities and bad certificates, and
• Decode encrypted packet data.

We will demonstrate how SSL Problem Finder can automatically analyze handshake problems.

Problems may include:

• Incorrect cipher suite,
• Performance issues,
• Bad or incorrect certificates, and more.

For some sample automatically generated recommendations from SSL Problem Finder, please view: Bad Analysis or Timing Analysis

February 10, 2010
NetView SNA Management Replacement

Many companies are trying to replace the SNA management functionality provided by NetView or NetMaster. Companies may have completed a conversion to Enterprise Extender and eliminated most of the NCPs. SNA sessions still need to be managed but the cost of NetView or NetMaster is very high. Can anything be done?

Inside Products has partnered with AnsyNova of Germany to provide the 2cSNA product in the United States. This product is being used in Germany by large organizations such as Luftansa and BMW.

2cSNA will provide a replacement for:

• NLDM
• NCCF
• REXX functions

2cSNA cannot provide a replacement for all system automation functions but many companies feel that it is a very cost competitive alternative. For an overview, please click here. For more detailed information, please click here.

January 26, 2010
IPv6 Business Information Exchange: Address Planning

This session will be presented by Dr. Peter Welcher who will discuss IPv6 addressing plans, and related considerations. It will include discussion of the following topics and questions:

• How should an organization, especially a medium to large enterprise, university, or government agency, go about planning its use of IPv6 addresses?
• What considerations go into obtaining one or more IPv6 prefixes for your organization?
• Thinking outside the box: NAT for IPv6 is heretical, but can look attractive; will your /48 give you enough subnet bits; should you stick with /64 prefixes only?
• Do you map your IPv4 addressing to IPv6, or start over with a new plan?
• Some IPv4 techniques that might be applicable to IPv6 address allocation?
• Working ends against the middle, to preserve your options (and potential drawbacks).

September 16, 2009
Inside the Stack Products and Direction

We discussed our products and direction, and did a live demonstration of:

• Inside the Stack - Client / Server (zLinux)
• Inside the Stack - Mainframe DB2
• Connection Log - Mainframe Only
• TN3270 Response Time Reports - Mainframe Only
• Early Warning System - Mainframe Only
• TCP Problem Finder- Client / Server (zLinux)
• SSL Problem Finder- Client / Server (zLinux)
• EE Problem Finder- Client / Server (zLinux)
• TCP Response Time Monitor - Client / Server (zLinux)
• Availability Checker - Client / Server (zLinux)

August 12, 2009
TN3270 Response Time Monitoring on z/OS

The TN3270 server for z/OS implements a built-in TN3270 response time monitor function. The function can be used to get statistical host, network, and end-to-end response times. In this session we will examine:

• How to set up TN3270 response time monitoring,
• What options exist for metrics,
• How to retrieve the statistics using the Network Management Interface (NMI),
• How to retrieve the statistics using commands,
• How to do further investigation if a problem is found, and
• Discuss if response time monitoring poses a large load on the system.

We will look at a running configuration and real connection data accessed via the NMI.

July 15, 2009
IPv6 Business Information Exchange: Penn State University

Our universities are among the institutions in the forefront of the integration to IPv6. At this meeting, Derek Morr of Penn State University will discuss their experiences with IPv6. This is an opportunity to talk with a technician from a real installation facing the challenges of integration.

Derek will discuss:

• Addressing,
• DNS, and
• Operational issues.

Come prepared with questions for Derek! At the following quarterly IPv6 Business Exchange meeting in October, Comcast will discuss their IPv6 status.

May 20, 2009
AT-TLS and IPSec Monitoring / Alerting - Demonstration

AT-TLS and IPSec are being widely implemented to protect application traffic. How can you monitor these kinds of connections? If connections are failing in the handshake process, can you tell? How can you set up automated alerting and monitor all your connections in the background?

In this session, we will demonstrate our latest product features - AT-TLS and IPSec logging and alerting to Connection Monitor and Early Warning System. We will demonstrate:

• Making AT-TLS and IPSec connections,
• Tracing them,
• Seeing them in the Connection Log reports, and
• Creating alerts with the Early Warning System.

April 21, 2009
IPv6 Business Information Exchange Meeting: IPv6 Firewall / IPv6 Roadmap

We are fortunate to have Eric Vyncke of Cisco, a security expert and co-author of the book "IPv6 Security", talk about Cisco's IPv6 Firewall implementation for the first portion of this meeting. You may want to take a look at Eric's book at:

We will spend the rest of the session discussing an IPv6 Roadmap. To many of us, it seems inevitable that IPv6 will need to be integrated into our networks at some point during the next 5 - 10 years. IPv4 will not go away, but IPv6 support will need to be added. This is a task which may be more complex than the migration to TCP/IP from SNA. The key to success may be a detailed migration roadmap. The better the roadmap, the better the chances of success in this very complex endeavor.

Some of us have been working on a roadmap for integrating IPv6 into our networks, we would like to discuss the topic with the membership.

• What design and implementation tasks need to be done?
• What are the business motivators for the integration of IPv6?
• What are the overall goals and issues?

March 10, 2009
TCP/IP Trace Analysis for Dummies

This session is for the systems programmer who wants to learn or refresh their TCP/IP trace reading knowledge. When you understand what the TCP/IP packets and flow in a trace are, then you are really a diagnostician. With TCP/IP, there are protocols within protocols - IP, UDP, HTTP, LDAP, and many others. In this session, the speaker will discuss:

• TCP/IP headers
• Congestion control
• Connection startup / shutdown

We will conclude by reading a trace from Enterprise Extender with embedded HPR/UDP packets.

February 10, 2009
Decrypting and Visualizing Data

SSL and AT-TLS are being widely implemented to protect TCP application traffic from the IBM mainframe. When diagnosing problems, we often want to see the data packets so we can find problems with the application. What if we could see the data in a screen format similar to how the user saw it? How can we do this?

In this session, we will discuss:

• The differences between SSL and AT-TLS
• Decrypting options for the data packets
• Viewing application data

January 20, 2009
IPv6 Business Information Exchange Meeting: Bechtel IPv6 Migration

Presentation by Fred Wettling, a Bechtel Fellow and co-author of Global IPv6 Strategies. Bechtel has been implementing IPv6 for a while and Fred is integral to this effort. He will discuss the following:

• Brief history / status of Bechtel IPv6 migration.
• What were the business motivators for the migration to IPv6?
• What migration strategy did we use?
• What were the biggest stumbling blocks?
• What might we do differently now that we know more?
• How do we see IPv6 fitting into the next 5 years of network strategic planning?

Following Fred's presentation, the group will have a general discussion to absorb and ponder the ideas.

December 2, 2008
SSL Problem Finder - Demonstration

SSL is being widely implemented to protect TCP application traffic. How can you diagnose problems? If the connection setup and negotiation causes poor response time, can you tell? In this session, we will demonstrate our latest product - SSL Problem Finder. We will discuss how to:

• Diagnose SSL handshake problems,
• Assess performance implications,
• View certificates, certificate authorities and bad certificates,
• Decode encrypted packet data?

As always, the packets and protocol will be shown visually using our Visual Diagnostic Language (VDL) technology!

November 4, 2008
Controlling TCP Connections

Controlling and trending TCP workload is critical to managing the TCP/IP network. This is more complicated than one might imagine. Connection monitoring is also the intersection of security and network management.

In this session, we will discuss:

• Managing listeners which:
• must be up,
• must not be up,
• must have connections (from certain subnets),
• must not have connections (from certain subnets),
• have not had connections (in a certain timeframe),
• have lost many connections (in a certain timeframe),
• Long lasting connections which start to degrade,
• Hanging connections, and
• Trending workload

September 9, 2008
Secure Sockets Layer (SSL)

SSL is being widely implemented to protect TCP application traffic. How does it really work? How is different from Application Transparent TLS? How do you use it?

In this session, we will discuss:

• SSL / TSL protocol fundamentals,
• Packets which form the SSL handshake,
• Performance implications,
• SSL and certificates,
• Server and client authentication

As always, we will look at the packets!

August 12, 2008
TCP Problem Finder

TCP Problem Finder is the tool we use most for finding and fixing problems for TCP/IP. What you can do is to feed a trace through, either from the mainframe or from anywhere on the network, then TCP Problem Finder will go to work. The product will try to tell you all the problems it sees. TCP Problem Finder also creates an environment which allows the diagnostician to find patterns more easily. As anyone working on network problems knows, this is how you really find problems.

In this session, you will see:

• The results of feeding in a trace
• How to drill down to a problem
• How to find patterns

We will take a few sample problems and see what steps you would take to resolve them.

July 8, 2008
Ten Commandments of TCP/IP Performance

The Ten Commandments of TCP/IP Performance are a distillation of hard-won experience. Monitoring and tuning TCP networks on the mainframe is complex for the basic reason that each network is a mixture of many applications and pieces of hardware. Each connection contains layers of protocols and subprotocols which must be decoded to make sense of the traffic patterns. Making sense of it all is the first step to tuning and improving performance.

This paper was sited on the IBM website for z/OS Best Practices. You may download the paper from our site by signing our guestbook at Guestbook

June 10, 2008
IPv6 Address Management

One of the new challenges with IPv6 is structuring addresses will be allocated in your enterprise. In IPv4, the variable length subnet masks (VLSM) is often used to allocate addresses to routers and links between routers. In IPv6, there is no such concept. Or is there? In IPv6, the prefix (/48, /56, etc) is often used. We need to learn to become quite familiar with the usage of the prefix and see how that may help us to manage our addresses.

This webcast will discuss:

• IPv6 prefixes
• How to get an IPv6 address from ARIN
• A simple addressing scheme
• Possible "gotchas" in address allocation

May 13, 2008
Inside the Stack - Monitoring TCP/IP

This webcast will be a little different from the webcasts we usually do. We have had a number of people ask us to tell them about how our products work, so once a quarter, we will do a webcast on one of our products. The other webcasts will be technology webcasts as we usually do.

The first webcast will be on our flagship product: Inside the Stack. We will discuss:

• Connection monitoring
• Trending application usage
• Monitoring activity of business partners
• Finding tuning opportunities

Inside the Stack is the fundamental base for your TCP/IP monitoring needs. We will end with a live demo of many of the product features.

March 11, 2008
Visual Diagnostic Language (VDL) for TCP/IP Trace Analysis

Reading and understanding a TCP/IP packet trace is difficult, even for diagnosticians with many years of experience. What if we could 'see' the trace flow in visual symbols? Finding a problem on the network is a matter of pattern analysis. Where does the good pattern fail? Is a bad pattern established? Humans are quite adept at pattern analysis. In particular, the part of the brain which does visual pattern analysis is far older than the part of the brain which analyzes written symbols such as numbers and text.

In this session, the speaker will show how to use the VDL for diagnosing and seeing the patterns for:

• Normal data flow
• TCP start up and shut down
• TCP errors (dup acks, out of sequence, fragments, retransmissions)
• Congestion window
• Timing problems

We expect that VDL will allow you to find the problem in a trace in far less time than the normal methods used today. We have applied for a patent for this technology.

February 12, 2008
Transition to IPv6

The American Registry for Internet Numbers (ARIN) announced in May, 2007 that ISPs should use IPv6. ARIN has also implemented a pricing policy to encourage use of IPv6 for ISPs. Microsoft Vista is enabled for IPv6 'out of the box'. The transition to IPv6 seems inevitable in the next few years. What transition techniques may be used for IPv6? What might the migration issues be?

In this session, the speaker will discuss:

• Dual stack mode (IPv6 / IPv4)
• Tunneling (Manual, GRE, GRE with IPSec, 6to4, Teredo)
• Translation (NAT-PT, SIIT)
• Application changes

January 15, 2008
Network Cryptography Implementations

More and more companies are wondering whether to implement IPsec for security. One of the most critical portions of IPsec is the initial negotiation - ISAKMP. As the migration to IPv6 begins, Cryptographically Generated Addresses (CGA) may be one of the ways to protect IPv6 addresses. Both IPsec negotiation and IPv6 rely on public / private key encryption. IPsec also uses Diffie-Hellman key exchange. How do these really work?

In this session, the speaker will explain:

• Asymmetric encryption / symmetric algorithms
• Public key / private key encryption
• Diffie-Hellman key exchange / groups
• ISAKMP phase 1 and phase 2 negotiation
• Packet flow for ISAKMP
• IPv6 Cryptographically Generated Addresses (CGA)

November 6, 2007
Baselining and Security Alerting for TCP/IP

Earlier this year, we discussed what should be monitored in real time to find if connections or applications are headed for trouble. We also discussed the performance metrics to monitor:

• Congestion window
• Round trip time
• Round trip variance
• Retransmissions
• Duplicate acknowledgments
• Bad status
• Hung connections

In this session, we will discuss baselining for TCP/IP. Setting appropriate thresholds may be the most critical portion of this whole task. You may have 'Red letter days' in your industry. These are very high volume days. On such days, the thresholds should be different. In this session, we will discuss all the complexities of baselining and how to do this automatically.

We will end by discussing an interesting security requirement from the U.S. Federal Government's Office of Management and Budget. The memo OMB 06-16 states that mobile and remote access devices should be reauthenticated after 30 minutes of inactivity. We will discuss how this may be accomplished and some of the pitfalls. To view this memo, please go to: OMB 06-16.

October 2, 2007
VTAM / CSM Buffer Pools and TCP/IP

The VTAM buffer pools and Communication Storage Manager (CSM) pools are used by both VTAM and TCP/IP. How are they defined? Which buffer pools may impact TCP/IP performance? What happens when Enterprise Extender is used?

In this session, we will discuss:

• Function of VTAM and CSM buffer pools for TCP/IP
• Commands to monitor
• Some problem scenarios

September 11, 2007
Proactive Monitoring for TCP/IP

What should be monitored in real time to find if connections or applications are headed for trouble? This may be mainframe web applications, FTP or other critical applications which are currently experiencing problems, or be getting close to doing so.

In this session, we will discuss:

• Performance metrics to monitor
• Congestion window
• Round trip time
• Round trip variance
• Retransmissions
• Duplicate acknowledgments
• Bad status
• Hung connections
• Why each of these metrics is important
• How to set thresholds
• The complexities of baselining

Aug 7, 2007
Topology Updates: Advanced Enterprise Extender Trace Analysis

Many companies are implementing Enterprise Extender. Diagnosing problems and reading a trace with EE packets is a challenge! One of the important functions to understand is topology updates. In this session, the speaker will discuss:

• Topology services
• Local and network topology databases
• When TDU's are sent
• TDU packet structure
• Control vectors and fields in a TDU

We will then spend time breaking out a TDU packet and the control vectors (TG characteristics, node characteristics, etc.) At the end of this session, you should feel quite comfortable with the important TDU packet!

July 10, 2007
Introduction to ICMPv6

The American Registry for Internet Numbers (ARIN) announced in May that applications which require 'large' blocks of contiguous numbers should use IPv6. ARIN is also considering a policy to encourage use of IPv6 where ever possible. It would seem wise to begin to understand this new protocol.

One of the major changes within IPv6 is for ICMP. ICMPv6 contains messages not just for error recovery but for informational exchange. In fact, the new Neighbor Discovery Protocol uses ICMPv6. The session will cover:

• New ICMPv6 message structure
• Neighbor Discovery Protocol
• Path MTU in IPv4 and IPv6

June 5, 2007
IPSec Implementation

IPSec is becoming the protocol of choice for end-to-end security. How does it really work? The implementation of IPSec on z/OS after version 1.7 requires implementation of Policy Agent, the Traffic Regulator Daemon, and the IKE daemon. We will look at how these tasks work together. We will do some problem diagnostics and tracing of IPSec sessions. The session will cover:

• Security associations (setup: manual, dynamic, modes)
• IP Authentication Header
• Internet Key Exchange (ISAKMP)
• Main and Quick modes
• Implementation of IPSec on z/OS

May 8, 2007
Inside the RU: Advanced Enterprise Extender Trace Analysis

Many companies are implementing Enterprise Extender. Diagnosing problems and reading a trace with EE packets is a challenge! Many of the important fields for directory search and topology updates are inside the SNA RU. In this session, the speaker will discuss:

• FM headers used in EE
• GDS variables used in EE
• Control vectors used in EE
• Sample Locates
• Types of directory search: broadcast, directed, one-hop

April 3, 2007
Network Cryptography for Dummies

Many companies are starting to use IPSec, SSL and other types of encryption and security protocols. To truly understand these protocols, fundamental concepts of cryptography such as public / private key encryption, cipher strengths, asymmetric and symmetric encryption are important.

• DES, 3DES, AES
• Asymmetric encryption / symmetric encryption
• Certificate authority
• Diffie-Hellman key exchange / groups
• Message authentication code (MAC)
• Message digest algorithm 5 (MD5)
• Rivest Shamir Adleman (RSA)
• Secure hash algorithm 1 (SHA1)
• Hashed message authentication codes (HMAC, HMAC MD5, HMAC_SHA)
• X.500 distinguished name
• X.509 digital certificate

We will also discuss key strengths and which encryption algorithms can be broken and how easily.

March 13, 2007
TCP/IP Network Health Check

Many companies are not optimizing their TCP/IP networks. This is a critical task and can save hard dollars in both CPU time taken by the stack and in the bytes transferred over the network. In our work, we have seen from 20% - 80% overhead at many companies which can be eliminated fairly easily.

• Unneeded traffic - where is it generated?
• Unnecessary TCP sessions - why?
• TCP errors which can be eliminated

We will conclude by looking at productivity on Enterprise Extender networks.

February 6, 2007
TCP/IP Analysis for Dummies

This session is for the systems programmer who is beginning to dive into TCP/IP. When you understand what the TCP/IP packets and flow in a trace are, then you are really a diagnostician. With TCP/IP, there are protocols within protocols - IP, UDP, HTTP, LDAP, and many others. In this session, the speaker will discuss:

• TCP/IP headers
• Congestion control
• Connection startup / shutdown

We will conclude by reading a trace from Enterprise Extender with embedded HPR/UDP packets.

January 9, 2007
Enterprise Extender Trace Analysis for Dummies

Many companies are implementing Enterprise Extender. Diagnosing problems and reading a trace with EE packets is a challenge! There are now multiple headers involved - both the SNA, HPR, UDP, and IP. In this session, the speaker will discuss:

• EE trace headers
• How they impact performance
• Adaptive rate-based (ARB) headers and flow control
• ARB slowdown
• EE keep alive

December 12, 2006
How Do TCP/IP Profile Parameters Impact Performance?

Do you know how to properly set the TCP profile parameters? These can impact the throughput and security of your TCP network. In this session, the speaker will discuss:

• Parameters which affect TCP, UDP, and IP
• How they impact performance
• What can go wrong?
• Case studies to illustrate settings
• Setting TCP parameters in other platforms such as Windows or Linux

We will end by discussing VTAM parameters specific to Enterprise Extender and how they may affect performance and overhead.

November 14, 2006
TCP/IP Response Time Monitoring for z/OS

Do you want to know how to get end-to-end response time for your TCP applications for z/OS AND the wide-area network? Wouldn't it be nice to break it out into CPU vs. network time?

In the first part of this session, we will show how you can get response time for various TCP applications on z/OS such as TN3270, CICS, FTP, and web server. We will show sample statistics gathered by the TCP Response Time Monitor product.

We will also discuss how to think about response time monitoring for Enterprise Extender. How do we define, match and break out a transaction into its components when so many headers are involved? We will again see if we can break out the time into host and network time, this time for packets flowing over HPR over UDP.

October 10, 2006
Hands on with IPv6

You know the world is moving to IPv6, but what is really involved? We will look at how to enable the z/OS implementation. As always, there is nothing like hands-on experience! We will look at trace packets on the IPv6 networks, both public and z/OS based.

We will look at:

• How has addressing changed?
• Categories of addresses, address states, address model
• Implementation of IPv6 on z/OS (dual mode stacks, migration)
• Problem determination in an IPv6 network (ICMPv6, SNMP, Netstat, SMF)
• Path MTU Discovery and Multicast Listener Discovery
• DNS and Resolver changes
• Troubleshooting IPv6
• Taking and reading an IPv6 packet trace
• Using new IPv6 Netstat commands

September 12, 2006
Ten Commandments of TCP/IP Performance

The Ten Commandments of TCP/IP Performance are a distillation of hard-won experience. Monitoring and tuning TCP networks on the mainframe is complex for the basic reason that each network is a mixture of many applications and pieces of hardware. Each connection contains layers of protocols and subprotocols which must be decoded to make sense of the traffic patterns. Making sense of it all is the first step to tuning and improving performance.

This paper was sited on the IBM website for z/OS Best Practices. You may download the paper from our site by signing our guestbook at Guestbook