Networking basics Introduction to the Mainframe Networking basics
Networking basics Introduction to the Mainframe: Networking basics © 2006 IBM Corporation
Networking basics Chapter 1 Mainframes and Networks © 2006 IBM Corporation
Networking basics Objectives § Understand the role of the network in your company's business objectives and corporate infrastructure. § After completing this chapter, you will be able to: –Explain the use of data networks in high volume industry transaction processing –List at least three typical advantages of a mainframe in network communications –Describe the role of a network administrator in a large network –List the major software components of the z/OS Communications Server. 3 © 2006 IBM Corporation
Networking basics Key Terms § APPN § Connectivity § Fiber optics § Integrated supply chain § Intranet § Network § OLTP § OSI § Backbone § Extranet § Infrastructure § Internet § Mainframe § Network Protocol § OSA 4 © 2006 IBM Corporation
Networking basics Mainframes, networks, and you 5 § What is a network? § Definition depends on who is using the network – IT Professional – Engineer § Our definition of a network § Where are mainframes used today? – ATMs – Credit Card Payments – Online Purchasing © 2006 IBM Corporation
Networking basics Networks and online systems § Categories of Networks –Internet –Intranet –Extranet § Terminals § Online Transaction processing –Characteristics of OLTP § What activities add to network traffic? 6 © 2006 IBM Corporation
Networking basics The importance of networks § Why are networks important? –Satisfy an objective or need. –Provide a means for transmitting data. –Accuracy and speed is essential to business –Numerous businesses depend on their networks. § Examples of mainframe-based networks – 1. 4. 1 ATM Cash withdrawal – 1. 4. 2 Credit purchase at a retail store – 1. 4. 3 Technology choices abound in network technology 7 © 2006 IBM Corporation
Networking basics Who supports the network? § Separated into hardware and software § Network administrator is responsible for both § Responsibilities of a Network Administrator –Definition, maintenance, and modification of an existing z/Series network –Problem, isolation, and correction –Performance tuning –Capacity planning recommendations –Development of operational procedures –Training of network operators 8 § © 2006 IBM Corporation
Networking basics What are basic elements of a data network? § Basic elements are hardware, software, and protocols. § What is a network infrastructure? § Protocols –“Traffic Rules” of the network –Define how two devices in a network communicate § Layered network architecture model. 9 © 2006 IBM Corporation
Networking basics Overview of System z network capabilities § System z is capable of handling many network nodes that are dispersed over a large area. § Internal network capabilities. –Guest machines –LPARs –Cluster § External network capabilities –TCP/IP applications, protocols, and equipment –Subarea System Network Architecture –Advanced Peer-to-Peer Networking –Integration of SNA into IP networks using EE 10 § © 2006 IBM Corporation
Networking basics z/OS Communication Server § Implements the SNA and TCP/IP protocols. § Provides a set of protocols to support peer-to-peer connectivity. § Performance enhancements that can benefit a variety of TCP/IP applications. 11 © 2006 IBM Corporation
Networking basics SNA and TCP/IP on z/OS § What is SNA? § When were the protocols developed? § What is the advantage of SNA? § Why is TCP/IP more popular? 12 © 2006 IBM Corporation
Networking basics Data integrity, security, and availability in a network § Data Integrity –Modification –Non-repudiation –Error-free transmission § Security –Procedures • Preventing unauthorized disclosure of transmitted data • Detecting unauthorized modification of data • Non-repudiation using proof of origin, receipt, and digital certificates. –Products • IBM Security Server and Commutations Server components include RACF and Farewell. • Communications server components include parameters to encrypt network traffic • Major IBM subsystems have security mechanisms. –Availability • Degree to which a system is ready when needed to process data. • Enhance availability through redundancy. • Parallel Sysplex 13 © 2006 IBM Corporation
Networking basics Summary § A network is the hardware and software that enables computers to share files and resources and exchange data. § To support changing requirements protocols such as SNA and TCP/IP can be combined to optimize performance. § z/OS network capability includes a fully-featured communications server with integration of SNA and TCP/IP protocols. § Many technologies exists to protect data between the server and authorized clients. 14 © 2006 IBM Corporation
Networking basics Chapter 2 Network Protocols © 2006 IBM Corporation
Networking basics Objectives § This chapter discusses various networking protocols. 16 © 2006 IBM Corporation
Networking basics Network protocols § Point to point - direct link between two hosts § Point to multipoint - direct link from one point to many others § Broadcast - multi-access with ability to address all hosts on the network § Local Area Network - LAN segment § Wide Area Network (WAN) 17 © 2006 IBM Corporation
Networking basics Local Area Network (LAN) Ethernet § Ethernet is believed to consist of 90% of network installations. –Standard defined in 1985 known as IEEE 802. 3 –CSMA/CD access method –Ethernet 10 mbps –Fast Ethernet 100 mbps –Dual Speed Products § Gigabit Ethernet –Retains the standard 10/100 base-T and CSMA/CD but it uses the fiber channel’s physical layer as the underlying transport mechanism –Full Duplex does not require the CSMA/CD scheme, but retains support for the Ethernet frame format. 18 © 2006 IBM Corporation
Networking basics Wide area networking § Network Interface Card (NIC) § Wide Area Networking (WAN) –The subscriber loop • What is a subscriber loop? • Asymmetric Digital Subscriber Line (ADSL) • Integrated Services Digital Network (ISDN) • ISDN Basic rate interface (BRI) • ISDN primary rate interface (PRI) –WAN Connection type • Point to Point • Circuit switching • T 1/E 1 • Packet switching • Frame relay 19 © 2006 IBM Corporation
Networking basics Network routing § Static - manual entry of routes into a table § Dynamic – populating routing tables by protocols –Routing Information Protocol (RIP) –Open Shortest Path First (OSPF) 20 © 2006 IBM Corporation
Networking basics Security § Firewalls and gateways § Security protocols § Protection 21 © 2006 IBM Corporation
Networking basics Summary § Understand the terminology § Understand the layers –physical layer –data link layer (Ethernet) –network layer (IP and routing) –transport layer (TCP and UDP) 22 © 2006 IBM Corporation
Networking basics Chapter 3 Network Hardware on the Mainframe © 2006 IBM Corporation
Networking basics Objectives § Understand the different types of network links § Explain why OSA-Express is the future direction for network connectivity § Set the different OSA modes § Explain how to associate protocols with the OSA-Express 24 © 2006 IBM Corporation
Networking basics Key Terms § Hiper. Socket § Osd § Escon § Ccl § Ose § Parallel § VLAN § OSA § Qdio § LPAR § Escon Converter § NCP 25 © 2006 IBM Corporation
Networking basics The System z channel subsystem and network links § Each server has a channel subsystem § The CSS allows channel I/O operations to continue independently of other operations. § The purpose is to permit a z 990 system to have more than 256 channels § Components –Logical partition name –Logical partition identifier –MIF Image ID –Physical Channel ID –Control Unit –I/O Device 26 © 2006 IBM Corporation
Networking basics The System z channel subsystem and network links 27 © 2006 IBM Corporation
Networking basics Hardware channels § Parallel Channel –Byte Multiplexer –Block Multiplexer § Enterprise System Connectivity Channel 28 © 2006 IBM Corporation
Networking basics Hardware channels § Open Systems Adapter –Integrates several hardware features and supports many networking transport protocols –Three main versions • OSA-Express 2 • OSA-Express • OSA-2 –QDIO verses non-QDIO –QDIO incorporates a number of features: • LPAR-to-LPAR • DMA (Direct Access Memory) • Priority queuing • Enhanced IP network availability • VLAN Support • ARP Takeover –Communication Controller for LINUX (CCL) 29 © 2006 IBM Corporation
Networking basics Hiper. Sockets § Provides high-speed TCP/IP connectivity within a Central Electronics Complex. § Based on the OSA-Express queued direct input/output protocol. –Hiper. Sockets with CHIPD FC –Hiper. Sockets with CHPID FD –Hiper. Sockets with CHPID FE –Hiper. Sockets with CHIPID FF 30 © 2006 IBM Corporation
Networking basics Summary The mainframe originally relied upon the channel subsystem to offload I/O processing to channel programs. DASD is still accessed using FICON and ESCON channels, but for networking connectivity, OSA-Express cards offer better performance and availability. 31 © 2006 IBM Corporation
Networking basics Chapter 4 Sample Configuration © 2006 IBM Corporation
Networking basics Objectives § The objective of this chapter is to provide an introduction to the type of System z networking components and environment that many organizations deploy § After completing this chapter, you will be able to: –List the components of a typical System z networking infrastructure –Give three reasons why organizations implement this type of configuration. 33 © 2006 IBM Corporation
Networking basics Key Terms 34 § CF § RAS § CICS § Service Level Agreement (SLA) § CPC § Stub Area § DB 2 § Sysplex § LPAR § Switch § OSA § VIPA § OMPROUTE § VTAM © 2006 IBM Corporation
Networking basics Example case § Requirements for a reliable network § The ZOS Company data center • Processing is divided up physically by central processor complexes and logically by logical partitions. • Production LPAR • Development LPAR • Systems programming LPAR • Fencing off the Production LPARs –Key mainframe network availability aspects • Reliability, availability, serviceability (RAS) –Component Failure –Dual and Diverse Paths –Performance –Failure Process • Security • Scalability • Continuing compatibility • Evolving architecture 35 © 2006 IBM Corporation
Networking basics Continued § Hardware Availability –Switches –OSA Cards –CPC –Coupling Facility § Software Availability –OSPF –TCP/IP –VTAM 36 © 2006 IBM Corporation
Networking basics Summary § Organizations run many of their mission-critical applications on System z and system availability is a key factor in maintain an organization’s business. § Key Points –Most organizations will have two central processor complexes (CPCs) to allow for scheduled and unscheduled outages –Most organizations will have a geographically isolated site to allow for a disaster recover situations –OSA card can be shared among LPARs on a CPC. –TCP/IP VIPAs are not associated with a physical interface and assist in maintaining availability for applications and users. 37 © 2006 IBM Corporation
Networking basics Chapter 5 TCP/IP on z/OS © 2006 IBM Corporation
Networking basics Objectives § After completing this chapter, you will be able to: –Explain the usage of the TCP/IP profile configuration file –Explain the basics of FTP and telnetd server configuration –Discuss the options for resolver configuration –List some of the most common client applications 39 © 2006 IBM Corporation
Networking basics Key Terms 40 § TCP/IP Profile § Inetd § Send buffer size § Trace Route § Otelnetd § Receive buffer size § ASCII § FTPD § Datagram Forwarding § EBCDIC § Window Size § netstat © 2006 IBM Corporation
Networking basics The heart of the matter § The TCP/IP daemon implements the IP protocol tack and runs a huge number of IP applications to the same specifications as any other operating system might do. § Control issues and the stack –A TCP/IP started task when started as a program using JCL uses a level of configurability that can’t be matched in a daemon environment. –Example: You can support more than one instance of a TCP/IP started task. § An IPv 6 apology –This book uses IPv 4, but System z fully supports IPv 6. 41 © 2006 IBM Corporation
Networking basics The TCP/IP Profile § Sample JCL for TCP/IP task //TCPIP PROC //TCPIP EXEC PGM=EZBTCPIP, //PROFILE DD DISP=SHR, DSN=SYS 1. PARMLIB(PROFILE) //SYSTCPD DD DISP=SHR, DSN=SYS 1. PARMLIB(TCPDATA) 42 © 2006 IBM Corporation
Networking basics The TCP/IP Profile § Profile Statements –Link Configuration Defining links, LPAR 1 DEVICE OSAEDEV 1 MPCIPA PRIROUTER LINK OSAELNK 1 IPAQENET OSAEDEV 1 DEVICE OSAEDEV 2 MPCIPA PRIROUTER LINK OSAELNK 2 IPAQENET OSAEDEV 2 DEVICE VIPADEV 1 VIRTUAL 1 LINK VIPALNK 1 VIRTUAL 1 VIPADEV 1 HOME 201. 2. 11. 9 VIPALNK 1 201. 2. 11. 1 OSAELNK 1 201. 2. 11. 2 OSAELNK 2 43 © 2006 IBM Corporation
Networking basics The TCP/IP Profile – IP Configuration – TCP Configuration • TCPMAXRCVBUFRSIZE • TCPSENDBFRSIZE – Static Routing Information – Automated IP Application Monitoring 44 © 2006 IBM Corporation
Networking basics The FTP Server § Information that can be controlled in the FTP. DATA set. –Banner Page –Anonymous Configuration –Data Set Defaults –Tracing and Logging –File System –SSL/TLS –JES and DB 2 environments § Character Sets § MVS and its UNIX subset 45 © 2006 IBM Corporation
Networking basics The telnet daemon § Two telnet Servers available in the z/OS environment: –TN 3270 – supports line mode telnet, but is primarily used to support the TN 3270 Enhanced Protocol –Z/OS UNIX Telnet server is a line mode server only. –What is inetd? § Character sets 46 © 2006 IBM Corporation
Networking basics A good resolver is hard to find § The resolver configuration file defines the operating characteristics of IP applications. – Sample resolver configuration file DOMAIN XYZ. COM § HOSTNAME MAINFRAME § NAMESERVER 200. 1. 1. 1 200. 1. 1. 2 § TCPIPJOBNAME TCPIP § Searching for resolver configuration information –Resolver configuration parameters can be placed in the obvious location: /etc/resolv. conf –Resolver configuration parameter can be placed in a file allocated to STSTCPD –Resolver address space § The multi-stock environment 47 © 2006 IBM Corporation
Networking basics TCP/IP clients § IP applications supported by z/OS –FTP –telnet –Ping –Tracerte or traceroute –Snmp –Netstat § Character sets 48 © 2006 IBM Corporation
Networking basics Summary § The TCP/IP started task is the engine that drives all IP based activity on z/OS. § The TCP/IP profile data set controls the configuration of the TCP/IP environment. § The FTP server implements the FTP standard and can communicate with any FTP clients on the network. § IP applications running on z/OS use a resolver configuration file for environmental values. § TCP/IP on z/OS support all of the well known server and client applications 49 © 2006 IBM Corporation
Networking basics Chapter 6 TCP/IP in a sysplex © 2006 IBM Corporation
Networking basics Objectives § After completing this chapter, you will learn: –The key features of a z/OS sysplex –How dynamic VIPA functions in a sysplex –How dynamic XCF functions in a sysplex –How sysplex distributor functions in a sysplex –How the sysplex performs problem detection and recovery –What are the routing needs of a sysplex 51 © 2006 IBM Corporation
Networking basics Key Terms § SYSPLEX § WLM § Multiple application-instance DVIPA § OSPF § Sysplex distributor § Stub area § RFC § XCF § Dynamic XCF § Unique application-instance DVIPA 52 § © 2006 IBM Corporation
Networking basics Clustering in z/OS § Computer cluster –A cluster is a group of computers coupled together so they are working as one unit. • Loosely coupled: Computers that are running with a minimum amount of communication and cooperation. Efficient use of the individual computer but limits the amount of coordination and sharing of workloads • Tightly coupled: Very effective workload sharing and communication. § The z/OS sysplex –Coupling facility • Handles the communication between the LPARs. • Information exchanged: Workload, status, and data transmissions. –Workload Manager (WLM) • Can be used to define performance goals for different applications and different types of work. • Can configure z/OS to direct traffic to the LPAR with the lightest workload. 53 © 2006 IBM Corporation
Networking basics Dynamic virtual addressing § DVIPA is activated in one of the two different ways; –An application explicitly issuing a bind() function call to the IP address. This is called unique application-instance DVPIPA –A TCP/IP stack dynamically activating the address. This is called multiple application-instance DVIPA. § Unique application-instance DVPIPA –Bind()function call can do two services. • Associate a socket with a specific IP address • Associate a socket with a specific port number –Enabling DVIPA on a TCP/IP Stack –Moving the DVIPA § Multiple application-instance DVIPA 54 © 2006 IBM Corporation
Networking basics Dynamic cross system coupling § VTAM and XCF § Establishing the links –When a TCP/IP stack becomes active in the sysplex and this stack has DYNAMICXCF coded, the following sequence of events occurs internally within the TCP/IP stack: • A DEVIC statement for this stack’s XCF device is automatically generated • A corresponding LINK statement is automatically generated • A HOME statement entry using the DYNAMICXCF IP address is added to the active HOME list for the stack. • The device is started. § More than just XCF 55 © 2006 IBM Corporation
Networking basics Sysplex distributor § What is a tightly coupled host to do? –Terminology • Distributing Host • Target Host • Backup Host • Combinations § Sysplex distributor in action –TCP/IP definitions –Walking through distributed connections –Controlling distribution § Moving distribution to the network § Problem detection and recovery in the cluster 56 © 2006 IBM Corporation
Networking basics Routing in a sysplex § Routing in a sysplex –Open Shortest Path First (OSPF) • OSPF requirements in a cluster • Rejoining an active sysplex –Network Interface Card § Summary –Sysplex is a cluster of tightly coupled independently functioning z/OS systems –Internal communications are handled by the coupling facility –TCP/IP uses the coupling facility and workload manager to optimize availability and load balancing in the sysplex –With a sysplex environment, OSPF is used to handle routing changes dynamically. 57 © 2006 IBM Corporation
Networking basics Summary § A sysplex is a cluster of tightly-coupled independent instances of the z/OS operating system. The internal communications within a sysplex are facilitated by the Coupling Facility. TCP/IP takes advantage of the Coupling Facility and Workload Manager to optimize availability and load balancing in a sysplex. 58 © 2006 IBM Corporation
Networking basics Chapter 7 Systems Network Architecture (SNA) basics and implementation © 2006 IBM Corporation
Networking basics Objectives § After completing this chapter, you will learn: –The basic concepts of SNA subarea networking –The major components in an SNA network –Routing and addressing in an SNA subarea –SNA flow control –VTAM definitions for subarea nodes 60 © 2006 IBM Corporation
Networking basics Key Terms § Systems services control point (SSCP) § Transmission group § Logical Unit (LU) § Peer network § Virtual route § Physical unit (PU) § Hierarchical Network § Explicit route § APPN network § VTAM § SNA § Subarea Network 61 § © 2006 IBM Corporation
Networking basics § Why you need to know about SNA today § What is SNA? –Systems Network Architecture is a data communication architecture established by IBM. –Different implementations of SNA can occur from product to product. –Recognize and recover from loss of data during transmission § History –Introduced in 1974 –IBM needed to address the issues with the errors in data transmission, and design a protocol to guarantee data integrity § Future –TCP/IP is the dominant protocol today, but it will take a long time for SNA to disappear –Many companies do not want to spend the money to convert from SNA…”If it ain’t broke, don’t fix it” IBM introduced new technologies to preserve the SNA and integrate it into IP. (SNA over IP) 62 © 2006 IBM Corporation
Networking basics SNA basics § 3 Major Implementations of SNA: –Subarea Networking –Advanced Peer to Peer (APPN) –High Performance Routing and SNA/IP (HPR) § Hierarchical Network structure –Aware of all activity on the network –Sub networks have a central point of control § Difference between TCP/IP and SNA –TCP/IP – Connectivity with computers and data sharing –SNA – Developed for central control 63 © 2006 IBM Corporation
Networking basics Subarea Networking - 1 § SNA Nodes –Send data and receive data from the network –Processors, controllers, and workstations –Node “types” – categorization of nodes by hardware and capabilities –Nodes fall into two broad categories • Hierarchical Roles (Subarea Networking) • Peer-oriented roles (APPN –Control point – manages the network resources § System Services Control Point –Activates, controls, and deactivates network resources in a subarea network. –Manages resources on a subarea network level –Coordinates the initiation and termination of sessions between applications. –Coordinates the testing and status monitoring of resources. 64 © 2006 IBM Corporation
Networking basics Subarea Networking - 2 § Subareas and domains –Every z/OS system with VTAM that implements SNA is referred to as a domain. –Explicit route • Forward explicit route • Reverse explicit route –Logical Paths • Virtual Route (VR) • Transmission Priority (TP) • Route Extension § Connecting subarea nodes –Transmission groups 65 © 2006 IBM Corporation
Networking basics Architectural components of the SNA network - 1 § Physical Units –Receives and acts upon request from the system services control point –Manages links and link stations, while accounting for the unique aspects of different link types. § Logical Units –Ports through which users access the SNA network. –Logical Unit types • LU type 1 • LU type 2 • LU type 3 • LU type 6. 2 66 © 2006 IBM Corporation
Networking basics Architectural components of the SNA network - 2 § SNA messages – Message units that contain: • End-User data, called data requests • Network commands, called command requests – Path information unit (PIU) • Transmission Header, Request Unit, Response Header, Response Unit § Initiating LU-to-LU sessions – Class of service (COS) 67 © 2006 IBM Corporation
Networking basics VTAM subarea definitions - 1 § Where are VTAM definitions stored? –Stored in two data sets –One data set stores text definitions of the SNA network and is reference by the VTAMLST DDNAME –The second data set stores load modules and is reference by the VTAMLIB DDNAME. § Start Options –Required Start Options • SSCPID • SSCPNAME • NETID • HOSTSA • HOSTPU 68 © 2006 IBM Corporation
Networking basics VTAM subarea definitions - 2 § Configuration lists – Specifies the resources that are to be activated when VTAM is started. § Defining resources to VTAM § Dynamic reconfiguration – Dynamic definition – Sift-down effect § Sample Network 69 © 2006 IBM Corporation
Networking basics Summary An SNA network implements the IBM proprietary networking architecture. In TCP/IP, the unit that is transferred through the network is called a packet. In SNA the unit that flows in the network is called a path information unit (PIU). It carries the data, the SNA headers, the transmission header (TH), and request header (RH). SNA has robust data flow control that permits constant flow of data to and from the network and prevents congestion in network resources. Although today TCP/IP is the dominant networking protocol in almost all organizations, SNA applications will exist for the foreseeable future due to the massive investment made in SNA applications. 70 © 2006 IBM Corporation
Networking basics Chapter 8 SNA Advanced Peer-to-Peer Networking (APPN) © 2006 IBM Corporation
Networking basics Objectives § After completing this chapter, you will learn: –What are the APPN node types –How APPN routes SNA data –The role of an APPN control point –What is a boarder node –APPN support for dependent LUs –How to define VTAM APPN networks –How to define a combined VTAM and APPN network 72 © 2006 IBM Corporation
Networking basics Introduction § APPN Comprises a group or groups of connected type 2. 1 nodes § APPN allows for direct communication between any networkattached devices without the need for SSCP intervention. § Salient functions of APPN: –Distributed directory searches –Topology and route selection services –Adaptive pacing and transmission priority –Intermediate session routing 73 © 2006 IBM Corporation
Networking basics APPN node types - 1 § Types: –Low-entry networking (LEN) end node –APPN network node –Interchange node –Border node § Low entry networking node § APPN network node –Network node server –Provides the following: • LU-LU session services for its locally resident LUs • Intermediate session routing for LU-LU sessions of all types • Network server functions for attached APPN or LEN end nodes. • Management services focal point or relay services for network problem management 74 © 2006 IBM Corporation
Networking basics APPN node types - 2 § APPN end node § Interchange Node § Migration data host 75 © 2006 IBM Corporation
Networking basics Directory Services - 1 § Central directory server –APPN network node extensions to the directory, maintained in every network node –Central Directory Server acts as a focal point for origination broadcast searches in an APPN topology sub network. –Number can be from 0 to several on a network § Directed Search –Uses information stored in the directory services database of a network node server to direct the search to the location of the requested LU. –Sending Units: • To a network node server from an end node and to an end node from a network node server • From a network node to a central directory server when the network node does not have information on the location of the destination logical unit. 76 © 2006 IBM Corporation
Networking basics Directory Services - 2 – Who originates a search? • Network node server for the LUs it serves • Central Directory server – A directed search is needed when a network node server or a central directory server receives a request of an LU that stored in its database § Broadcast search – Each node receiving the search request sends the search to each of its network nodes. 77 © 2006 IBM Corporation
Networking basics Topology and route selection § Each node has a copy of the network topology § A network node uses the network topology database to computer routes for sessions that originate at LUs in its domain. § Topology database –Created and maintained by the topology database manager. –Consists of: • Local topology database, unique to a node • Network topology database (replicated across the network nodes) 78 © 2006 IBM Corporation
Networking basics Control point sessions - 1 § CP-CP Session –Established between type 2. 1 nodes. –Network node to other network nodes –Used to perform searches for resources –After a session is activated the topology database manager sends updates that contain the node’ topology database to the other note. § Route Selection –Network node control point services performed • Located the destination LU • Calculate the best route § Dependent LU support –Dependent LU Requester/Server 79 © 2006 IBM Corporation
Networking basics Control point sessions - 2 § Border Node –Two APPN networks can be interconnected by a border node, either extended or peripheral. –Peripheral subnetwork boundaries –Extended subnetwork boundaries § Connection Network –Representation of a shares access medium enabling dynamic direct connectivity between a pair of link stations attaching to the facility. 80 © 2006 IBM Corporation
Networking basics Summary APPN is the IBM strategic SNA protocol in the mainframe. It is required for sysplex, Enterprise Extender implementation, and many other technologies. APPN is dynamic in nature and reduces the amount of predefinition required in an SNA subarea network. In contrast to subarea networking, where special hardware and software required for intermediate session routing, every node that can act as network node can perform routing of SNA packets. 81 © 2006 IBM Corporation
Networking basics Chapter 9 SNA/IP Implementation © 2006 IBM Corporation
Networking basics Objectives § Be able to learn: –The background and motivation behind integrating SNA and TCP/IP networks. –Technologies and solutions applied within System z based organizations, for integrating SNA and TCP/IP networks. –The features and benefits of each of these technologies and why you might use one method over the other. –How a z/OS network administrator would implement these technologies. 83 © 2006 IBM Corporation
Networking basics Key Terms § DLSW § EITF § DLSW+ § AIW § HPR § Token Ring § 802. 2 § SDLC § LLC 2 § Spoofing § Wan 84 © 2006 IBM Corporation
Networking basics Background § Organizations have a heavy investment in SNA based transaction programs, applications and networking infrastructure. § In many cases they provide the core to an organizations business. Based on figures in 1997, over twenty trillion have been invested in SNA applications in over 40, 000 organizations worldwide. § Surveys indicated SNA accounts for 61% or wide area network traffic, and 66% of WAN budgets § With the pervasiveness of the internet and IP networks being the de facto standard, organizations are standardizing on TCP/IP for intranet and external partner connectivity 85 © 2006 IBM Corporation
Networking basics Background § A transaction oriented program is dependent on the underlying protocol it uses. The API (application Programming Interface) is different if one uses SNA or TCP/IP as the transport in the network § Changing a transaction oriented program from one protocol (i. e. SNA) to another protocol (i. e. TCP/IP) requires a re-design of the communication part in the program, replacing the code that handles error recovery, exception processing, and many other tasks § Conversion of existing SNA applications to TCP/IP-enabled applications can be economically impractical § So, how can we enable IP applications and preserve SNAapplication and endpoint investment, while converging on a single network protocol? 86 © 2006 IBM Corporation
Networking basics Commonly used SNA over IP integration technologies § There are several different ways of running SNA and TCP/IP mixed protocol communication over single IP protocol transport networks. § The following System z Communications Server solutions are common place at organizations. – Data link switching (DLSw). SNA traffic is encapsulated in TCP packets. – Enterprise Extender. SNA (HPR) packets are carried as User Datagram Protocol (UDP) packets over an IP network –Telnet/3270 (TN 3270). 3270 data streams are carried over TCP connections to a server that replaces the TCP transport with SNA transport 87 © 2006 IBM Corporation
Networking basics Data Link Switching (DLSw) § DLSw was first developed by IBM and made available in 1993, to provide SNA support within multi-protocol routers § Cisco’s implementation of DLSw is known as DLSw+ , and contains additional enhancements to the original DLSw flavor § There are two types of DLSw: –Local: Local DLSw does not use TCP/IP, instead it enables communication between Lan-attached SNA devices and an SDLC device that is link-attached to the same DLSw route –Remote: Remote DLSw requires at least two routers (with DLSw feature loaded) connected to each other over an TCP/IP network. DLSw performs an encapsulation, decapsulation function, wrapping the SNA frames into TCP/IP packets for transportation across the IP network 88 © 2006 IBM Corporation
Networking basics Data Link Switching (DLSw) § The following points summarize some of the DLSw attributes: –DLSw is a forwarding mechanism, and supports SNA device types, PU 2. 1 and PU 4. –DLSw provides switching at the Data Link Layer (Layer 2), and encapsulates SNA data into TCP/IP packets for transport. –DLSw does not provide full routing capabilities, instead DLSw is implemented within routers, the SNA end points remain the same, no change is required. –DLSw, like any other TCP connection, utilizes the dynamic and rerouting capabilities of the IP network. 89 © 2006 IBM Corporation
Networking basics Enterprise Extender § What is an Enterprise Extender? EE is an extension of SNA HPR that provides encapsulation of SNA application traffic within UDP frames by EE-capable devices at the edges of an IP network. § Enterprise Extender capable devices would include: –z/OS EE configured Communications Server running on an z/Series Host. –Cisco Routers with SNASw defined –Microsoft Host Integration Server or Communications Server for Windows 90 © 2006 IBM Corporation
Networking basics Enterprise Extender - advantages § Ability to consolidate onto a single transport network. § Eliminates parallel networks, reduces equipment, lowers data circuit cost, simplifies network management. § No Changes required to SNA applications. § SNA can exploit the OSA Gigabit Ethernet interface cards. § Takes advantage of the re-routing capabilities of the IP router network. § Enterprise Extender solutions preserve session priority. 91 © 2006 IBM Corporation
Networking basics Summary Consolidating SNA onto IP is not a simple task. Each set of SNA components and applications should be evaluated on its own merits, and a solution should be found that best suits the requirements. In some cases a solution is to replace an application or device; in other cases it is to use DLSw, SNASw, BEX, Enterprise Extender, or a combination of these. 92 © 2006 IBM Corporation
Networking basics Chapter 10 Telnet/3270 © 2006 IBM Corporation
Networking basics Introduction § During the last several decades, before the Internet became popular, hundreds of companies established their own SNA networks. As a result, thousands of mainframe applications were written to communicate with 3270 -type terminals § During the last decade, corporate networks started implementing IP as the transport protocol on their backbone. Because so many SNA and 3270 applications existed, they looked at integrating the SNA protocol into their IP backbone. § The technology used to move from SNA 3270 applications to TCP/IP is called TN 3270. 94 © 2006 IBM Corporation
Networking basics 3270 data stream § The 3270 data stream operations are designed primarily for transmitting data between an application program and a 3270 display with keyboard or 3270 printer § The 3270 data stream is implemented using a mapped character buffer in the device. § Data received from the application program and data to be transmitted to the application program are stored in a device buffer and displayed on the screen in the form of alphanumeric characters and symbols § The mode of operation used by 3270 terminals is called full screen. Full screen means that the buffer or parts of the buffer are transferred from the application program to the 3270 device and from the 3270 device to the application program. 95 © 2006 IBM Corporation
Networking basics 3270 data stream § Field Attributes § The field attribute defines the start of a field and the characteristics of the field. Field attribute defines the following field characteristics: –Protected or unprotected –Autoskip –Nondisplay or display –Intensified display –Alphanumeric or numeric. 96 © 2006 IBM Corporation
Networking basics TN 3270 description § Telnet/3270 is an enhancement of the old Telnet protocol where Telnet clients can use this protocol to access the resources on a Telnet server as if the client was directly attached to the server § The Telnet protocol is based on three principles: –The Network Virtual Terminal (NVT) concept –A symmetric view of terminals and processes –Negotiation of terminal options 97 © 2006 IBM Corporation
Networking basics TN 3270 description § TN 3270 client and server must negotiate certain additional options if they are to succeed in communication. The most important of these are: –Binary Transmission –End of Record –Terminal Type 98 © 2006 IBM Corporation
Networking basics Positioning the TN 3270 E server § One of the dilemmas for many mainframe installations was where to place the TN 3270 E server. Should it be implemented outside the mainframe on one of the following boxes? –AIX with IBM's Communications server –WINDOWS with IBM's communications server –WINDOWS with Microsoft's host integration server –One of the other many TN 3270 E software packages that are available § If all your 3270 applications are in the same place (same host or same Parallel Sysplex) then it probably makes sense to implement TN 3270 E in CS for z/OS § If your 3270 applications are in the same location but not quite as closely coupled, you can still implement TN 3270 in the mainframe and use the channel to channel connection between the LPARs to connect the TN 3270 server to the target application 99 © 2006 IBM Corporation
Networking basics TN 3270 functions supported in z/OS communications server § TN 3270 functions that the Communications Server supports: –Secure Sockets Support –Multiple Ports Support –IP Address to LU Name Mapping (IP Filtering) 100 © 2006 IBM Corporation
Networking basics Implementing TN 3270 server in z/OS communications server § You can implement TN 3270 in z/OS communications server either within the TCP/IP stack (the TN 3270 definition are imbedded in the TCP/IP profile) or in a dedicated address space 101 © 2006 IBM Corporation
Networking basics Summary The TN 3270 E protocol represents the evolution of SNA as it converged into the world of TCP/IP. It is the primary method of connecting end users to mainframe computers. It consists of a character-based data stream. TN 3270 E connectivity is handled on the mainframe by the TN 3270 E server. The TN 3270 E server converts a TN 3270 E TCP/IP connection to an SNA session. The terminal LU of the SNA session is emulated so that the SNA application functions as though it were connected to a nonprogrammable 3270 terminal. 102 © 2006 IBM Corporation
Networking basics Chapter 11 Operations Procedures © 2006 IBM Corporation
Networking basics Objectives § Provide an overview of the types of operational tasks a z/OS network administrator would perform or assign to another group. § List network control products that are used for controlling and displaying z/OS network component status. § List z/OS network component commands and displays that are commonly used to monitor and control the network. 104 © 2006 IBM Corporation
Networking basics Operational tasks § Within z/OS networking components on the mainframe, some common tasks you will be expected to fulfill include: –Fix z/OS network related faults –Change and configure the network components –Monitor and control the network components. –Provide performance and usage network statistics. Work with other groups on projects, tasks and faults. 105 © 2006 IBM Corporation
Networking basics Description § Where to start? 106 © 2006 IBM Corporation
Networking basics Network startup, shutdown § The network subsystems such as VTAM and TCP/IP, and related components are normally started as part of the IPL startup sequence under z/OS § There is normally an automation software product that controls the startup, and this will have dependency checks or parent child relationships built into it, where VTAM would have to start prior to TCP/IP, and that FTP (an application daemon under TCP/IP) cannot start prior to TCP/IP is active § VTAM startup § TCP/IP startup 107 © 2006 IBM Corporation
Networking basics Network startup, shutdown § Example dependencies –The network cannot start before the z/OS operating system and JES 2 is up and going –The network devices also need to be varied online by the operating system –The network will startup before any applications, there is no point in starting these before the underlying network is available –Some of the VTAM subsystem and major nodes must startup prior to TCP/IP starting –EE cannot be enabled until both VTAM and TCP/IP are up –Omproute is started after TCP/IP is available 108 © 2006 IBM Corporation
Networking basics Network startup, shutdown § Stopping the network § Taking VTAM down, effectively means that communication can no longer occur between end users, devices and VTAM applications, including EE. There are features and processes that organizations deploy to manage, a network takeover or movement of control when an outage occurs § The order in which the network is shutdown, is in reverse to that of a startup. Generally it would be: –TCP/IP and VTAM Applications are stopped Each application will have its own shutdown command. –TCP/IP is stopped via: P TCPIP –VTAM is then stopped via: Z NET, QUICK command 109 © 2006 IBM Corporation
Networking basics z/OS administrator tasks § The network administrator tasks are usually derived from: –The Customer or organizations problem management or request system. Problems or requests might get called into a central helpdesk, then be assigned to your group for action. They will normally have a severity classification –Project work. You may be assigned to a specific project that introduces a new function or upgrades a network software component 110 © 2006 IBM Corporation
Networking basics VTAM § The following commands provide examples of some of the VTAM commands that you might use in a z/OS network administration role § Displaying VTAM resources § Controlling VTAM resources § Changing VTAM resources –Vary command –Modify command 111 © 2006 IBM Corporation
Networking basics TCP/IP § TCP/IP, like VTAM, has a number of commands available to monitor, change and verify the environment. § TCP/IP provides two methods of issuing commands, either via z/OS console or SDSF log commands or TSO commands § Monitoring TCP/IP on z/OS § Changing the TCP/IP configuration –You can alter the TCP/IP configuration with the Obeyfile command § Verifying the TCP/IP configuration 112 © 2006 IBM Corporation
Networking basics TCP/IP § This command displays the status of all known TCP/IP stack(s) within the z/OS LPAR 113 © 2006 IBM Corporation
Networking basics Documenting the network environment § Good documentation is worth its weight in gold. It is a great tool for learning about your network environment and helps reduce the time it takes for you to resolve a problem or find out information § The type of information you might expect to find in this type of documentation includes: –Network component overview diagram –Network component description –External connections –Network naming conventions –Network processes –Network related products, tools, exits, and automation –Change log –Problem log –Contact details 114 © 2006 IBM Corporation
Networking basics Chapter 12 Problem Determination © 2006 IBM Corporation
Networking basics Objectives § Be able to learn: • The different system dumps • The different VTAM traces • The z/OS commands to display or modify data and settings • The messages associated with each component 117 © 2006 IBM Corporation
Networking basics Key Terms 118 § Dump § Trace § Command § Symptoms § Component § Tools § CSM § Storage § GF § VIT § NETSTAT § Buffer © 2006 IBM Corporation
Networking basics Houston, we have a problem § Keeping the network problem free and responsive is a priority, so determining where a problem lies and fixing it quickly is imperative § Your first sign that there is a network problem likely will come from users or operators. Users might complain about: –An unusual message received –The system behavior being different than the past –Slow response time on the network –No response from the system –System operators might report: • Poor performance • A TCP/IP or VTAM abend • A loop or wait on the sub-system (TCP/IP, VTAM) 119 © 2006 IBM Corporation
Networking basics Houston, we have a problem § To begin, determine the general cause of the problem: § Read messages in the system log. Each z/OS component has a unique message prefix that identifies the component: –VTAM messages are prefixed with “IST” –TCP/IP messages are prefixed with “EZZ” –CSM messages are prefixed with “IVT” –TSO messages are prefixed by “IKT” § Look for suffix “E, ” which stands for “Error. ” Check the messages manual for an explanation of the error. § Check to see if the system created a dump at error time. § Find out if the system has changed; for example, whether the procedures for any components have changed: –VTAM start –TCP/IP profile –BUFFERS 120 © 2006 IBM Corporation
Networking basics Tools and diagnostic aids § VTAM tools and diagnostic aids –Abnormal end (Abend) Dump –Stand-Alone Dump –SVC Dump –First Failure Support Technology (FFST) Dump –Network Traces –Sense Codes –VTAM Internal Trace –Other VTAM Traces –VTAM DISPLAY Commands 121 © 2006 IBM Corporation
Networking basics Tools and diagnostic aids § From SDSF, you can issue the commands to display aspects of VTAM. For output of the command view the “LOG” (System Log) panel. Examples include: –D NET, APPLS Status of application program major and minor nodes –D NET, BFRUSE VTAM buffer usage –D NET, CSM Communications storage manager buffer pools –D NET, MAJNODES Status of major nodes 122 © 2006 IBM Corporation
Networking basics Tools and diagnostic aids § Examples of the DISPLAY command 123 © 2006 IBM Corporation
Networking basics TCP/IP tools and diagnostic aids § Abend Dump § Stand-Alone Dump § SVC Dump § TCP/IP component trace § TCP/IP DISPLAY Commands for Problem Determination § PING command § TRACERTE command 124 © 2006 IBM Corporation
Networking basics TCP/IP tools and diagnostic aids § Example of a DISPLAY command 125 © 2006 IBM Corporation
Networking basics VTAM problems § Abend message § Activating network nodes takes too long § Application program reports an unexpected return or sense code § Deactivating network nodes takes too long § Error message § Hung session, LU, or terminal § Hung system § IST error message § LOGON takes too long to complete § Performance is degraded after a network outage § Response time is slow § Storage message IST 154 I or IST 561 I-IST 566 I 126 © 2006 IBM Corporation
Networking basics VTAM Problems § Determining VTAM problems on the Network § PTF Number § Device Type § Terminal Action § VTAM Application Program § Hardware Error Condition § Coding Change 127 © 2006 IBM Corporation
Networking basics TCP/IP Problem Determination § Various messages appearing in the console log or in the SYSPRINT, SYSERROR, and SYSDEBUG data sets, together with alerts and diagnostic aids provide information that helps you to find the source of a problem § To find the SYSPRINT, SYSERROR, and SYSDEBUG data sets, check the DDNAMEs specified in the startup procedure for TCPIP. 128 © 2006 IBM Corporation
Networking basics TCP/IP Problem Determination § Recommended Documentation for TCP/IP Problems § PTF Number –If the problem appears after you apply a PTF (program temporary fix). § MCL Number –If the problem appears after IBM has applied microcode level. § Profile TCP/IP –If the problem appears after a modification or an addition of a device. § Network Topology –If the problem is associated at a NETWORK modification (equipment change, configuration change). § Hardware Error –Sometimes it is immediately apparent that a problem is related to a specific hardware error condition. –If a hardware error occurred, note the failure condition that accompanied it, such as UNIT CHECK or CHANNEL-END/DEVICE-END. 129 © 2006 IBM Corporation
Networking basics CSM § CSM is a component of VTAM that enables host applications to share data with VTAM and other CSM users without having to physically copy the data. CSM reduces CPU utilization and optimizes system performance during the transfer of bulk data by enabling applications to share buffers § The code for CSM comes with VTAM, but once CSM is started it runs independently of VTAM and can be used by other MVS tasks and subsystems § CSM is started automatically when it is first invoked, and continues to run even if VTAM terminates. While CSM is not in use it retains a minimum amount of storage. It terminates only when MVS itself terminates 130 © 2006 IBM Corporation
Networking basics SNIFFER § When the problem seems to be outside the mainframe, we need to take a SNIFFER trace on different LAN segments to determine what is the problem root cause (station, fiber, protocol, router) § There a lot of tools utilizing packet sniffers, network monitors, and protocol analyzers designed for LAN administrators, security professionals and network programmers. § The most used SNIFFER Tool is from NAI or ETHEREAL which is an Open Source Software released under the GNU General Public License. § An example of a Sniffer is the Distinct Network Monitor, which is a packet capture and network protocol analyzer software that translates complex protocol negotiation into natural language, pinpointing where errors occurred. 131 © 2006 IBM Corporation
Networking basics SNIFFER § This example shows traffic distribution by MAC address 132 © 2006 IBM Corporation
Networking basics Chapter 13 Performance tuning and capacity planning © 2006 IBM Corporation
Networking basics Objectives § Be able to: –Understand network tuning and capacity planning concepts –Monitor the network for performance problems –Determine network performance bottlenecks –Use tools to fix performance problems 135 © 2006 IBM Corporation
Networking basics Key Terms 136 § Performance § Tuning § Bottleneck § Monitor § Capacity Planning § RMF © 2006 IBM Corporation
Networking basics Tuning and capacity planning overview § In a computer system, performance tuning is the skill of allocating existing resources to meet business objectives; these are often specified by the business in a service level agreement or SLA. In contrast, capacity planning is the means by which an IT professional predicts the resources that will be needed to meet future service objectives. § Similar approaches can be used for both performance tuning and capacity planning: –Comparison with other systems –Creating a transaction profile –Assessing the costs of implementing new functions –Using a performance modeling system –Testing the system with industry standard benchmarks, such as the Teleprocessing Network Simulator (TPNS) 137 © 2006 IBM Corporation
Networking basics Tuning and Capacity Planning Overview 138 © 2006 IBM Corporation
Networking basics Defining performance management § Performance management means monitoring and allocating data processing resources to an application, according to a service level agreement (SLA) or informal objectives § The SLA is a contract that objectively describes such measurables as: –Average transaction response time for network, I/O, CPU, or total –The distribution of these response times (for example, 90% TSO trivial at less than 0. 2 of a second) –Transaction volumes –System availability 139 © 2006 IBM Corporation
Networking basics Defining capacity planning § Capacity planning is a process of planning for sufficient computer capacity in a cost-effective manner to meet the service needs for all users. § Capacity planning involves asking the following questions: –How much of your computer resources are being used? –CPU, Processor Storage, I/O, Network –Which workloads are consuming the resources (workload distribution)? –What are the expected growth rates? –When will the demands on current resources impact service levels? 140 © 2006 IBM Corporation
Networking basics What is a performance problem? § Most performance problems revolve around unacceptably high response times or resource usage, which we can collectively refer to as “pain”. § Some network bottlenecks can be caused by the following: –Client-network interface –Network bandwidth –Network topology –Server network interface –Server CPU load –Server memory usage –Server bandwidth –Inefficient configuration 141 © 2006 IBM Corporation
Networking basics Monitoring using RMF § RMF issues reports about performance problems as they occur, so that the mainframe IT staff can take action before the problems become critical. § Your installation can use RMF to: –Determine that your system is running smoothly –Detect system bottlenecks caused by contention for resources –Evaluate the service your installation provides to different groups of users –Identify the workload delayed and the reason for the delay –Monitor system failures, system stalls, and failures of selected applications 142 © 2006 IBM Corporation
Networking basics Monitoring using RMF § Processor Problems § I/O Problems § Storage Problems § Sysplex Problems 143 © 2006 IBM Corporation
Networking basics Network performance and capacity planning § Performance and capacity management –Resolving performance problems –Extending the life of hardware –Maintaining service levels in the face of application growth –Exploiting technology –Supporting software and hardware changes –Ensuring new applications perform well –Managing with scarcer skills § Examine the performance data in depth and focus on the essence of the questions you want to solve 144 © 2006 IBM Corporation
Networking basics Performance tools § Net. View(R) Performance Monitor § Tivoli Monitoring for Network Performance § ASG-TMON for TCP/IP § OMEGAMON § z. PCR 145 © 2006 IBM Corporation
Networking basics Performance Tools § TCP/IP performance checklist 146 © 2006 IBM Corporation
Networking basics Summary § The network administrator should first determine the general cause of the problem by reading error messages, checking for system memory dumps, checking to see if software or hardware has changed, and reading the system log. After determining the general cause of the problem, the network administrator should use the tools and diagnostic aids at hand to determine the specific cause of the problem. Lastly, tuning tasks should be carried out to ensure good network performance. § z/OS has diagnostic aids that the network administrator can use: abend dumps, standalone dumps, and supervisor call (SVC) dumps, which the Interactive Problem Control System can format for easier reading. Additionally, VTAM has specific aids, such as First Failure Support Technology, CSDUMPs, network traces, sense codes, VTAM traces, and commands that display the state of VTAM components and resources. TCP/IP has component traces and diagnostic commands (such as the NETSTAT command) that help determine problems in the IP network. Communications Storage Manager (CSM) problems generally manifest themselves as central storage problems. The network administrator can display CSM’s use of storage, activate CSM VTAM traces, and dump CSM storage for analysis. 147 © 2006 IBM Corporation
Networking basics Chapter 14 Change Control Procedures © 2006 IBM Corporation
Networking basics Objectives § Be able to learn: –Why Change Control is important from a z/OS networking perspective –IT related Groups and areas that the z/OS networking administrator interacts with you will need to interact with during Changes –Examples of types of changes and process that should be followed –Some good practice and habits to use 149 © 2006 IBM Corporation
Networking basics Importance of change control § Change Control has become increasingly important as organizations focus on availability to their clients and revenue. § The z/Series platform should be treated with respect, the z/OS platform normally provides the core to an organizations business applications. If any of the components are unavailable on the z/OS environment, it can have a severe impact on revenue and client satisfaction § Implementing change is probably the greatest cause of component failure on z/OS, whether it be a coding error, incorrect command issued, unknown relationship with another component, the list goes on § The Change Control process is all about reducing risk and implementing change successfully. 150 © 2006 IBM Corporation
Networking basics Remember, this is not a single user system § Do not forget, that z/OS systems are used by many users and applications. Some systems may have in access of 30, 000 interactive users. Any network change should be treated with respect, be well planned, and have processes in place for implementation and back out. 151 © 2006 IBM Corporation
Networking basics Change control process and relationships § All IT organizations will have a change control application or process, that you will be required to use and follow. There a lot of different applications available, and are sometimes not that intuitive § In general the process might look like: –Research component change, problem or request. –Establish what needs to be done. –How will you test the change ? –Build and submit your change plan –Test your change –Change gets approved and scheduled –Are there any other groups that need to know about your change ? –Implement your change –Complete and close change control form 152 © 2006 IBM Corporation
Networking basics Groups you will interact with § Change Control § Security § Operations § System Programmers § business users § Testers § Team leader § WAN group 153 © 2006 IBM Corporation
Networking basics Good habits § Log your changes, the How Why and When § Take a backup, take a backup § Document your naming convention § Add comments to your configuration members § Take regular snapshots 154 © 2006 IBM Corporation
Networking basics Change Interaction § Change control § WAN group § Server group § z/OS group § Customer 155 © 2006 IBM Corporation
Networking basics Example changes and process § Telnet server update, dynamic vs. static changes 156 © 2006 IBM Corporation
Networking basics Good practice and habits § Telnet server Test=yes § Backup members , libraries § Always have back out, contingency plan in place § Create repeatable change templates, or refine process § Document, document § Understand impact of change 157 © 2006 IBM Corporation
- Slides: 155