1. TCP/IP for VSE The Last Word in Performance Presented by John Rankin CSI International Phone: (800) 795-4914 Web: HTTP://WWW.e-VSE.COM Copyright (C) 2006 - CSI International

2. Agenda 4 General Discussion –Performance Tuning –Data Flow 4 Performance –Network –Stack –Applications

3. Performance Tuning Steps 4 Isolate your performance problem –Collect Symptoms –Collect Data 4 Create performance theory –Matching symptoms to the theory 4 Adjust performance controls –Testing the theory –re-adjusting as necessary

4. TCP/IP Collection Facilities 4 SET DIAGNOSE=PERFORM –Performance of specific connections 4 QUERY STATS –Overall System Statistics 4 QUERY IPSTATS –Cumulative statistics by IP address 4 SET RECORD=FULL 4 QUERY FRAGMENT

5. Life of a Datagram 4 To understand performance, you must understand how data flows through the network and subsequently through TCP/IP –An application such as FTP –The TCP/IP engine –A TCP/IP link driver –The network –And back again

6. Performance is Flow of Data 4 Performance is divided into three areas: –Network Data Flow Ethernet, token-ring, etc. –Stack Data Flow TCP, UDP, IP, etc. –Application Data Flow FTP, Telnet, LPR, etc.

7. Network Data Flow 4 Network Connections –Type of Connection –Type of Conduit 4 Flow of the actual information –Large to smaller pipe considerations –Fragmentation

8. Control Unit Considerations 4 Integrated adapters –General performance –Sharing Token-ring with SNA 4 The CLAW interface –Cisco Router with CIP card –Channel attached RS/6000 4 The Open Systems Adapter

9. The LAN Station Controller 4 Devices –IBM 3172/OSA –OEM such as Netshuttle for VSE 10mbs –IBM 2216 100mbs 4 Throughput –Theory –Control 4 Multiple control units

10. OSA Express/2 Adapter 4 Specialized Interfacing 4 Full IP packet transfers 4 Performance –Maximizes Packet Transmissions –Reduces fragmentation –Improves operation overlapping

11. Hyper Sockets 4 Hybrid operating interface 4 Low Level Hardware interface 4 Performance –Extremely Fast and high volume –Maximize transmission size –64k IP datagrams. –Maximum TCP transmission size

12. Retransmission 4 What is happening in TCP/IP to cause this to happen –The theory –The practice 4 Performance –How can it be measured –How can it be effected

13. Global Verses Route Statements 4 Retransmission effected for entire environment 4 Specific Parameteres –Available on specific routes –Definable by class of net or subnet

14. Retransmission Parameters 4 MAXSEGMENT=576 - 65535 4 CRETRAN=10 - 1000 4 DRETRAN= 10 - 5000 4 FIXRETRAN= YES | NO 4 MINRET=10 - 1000 4 MAXRET=10 - 5000 4 RETRY=5 - 1000 4 RPAUSE=10 - 5000 4 WINDOW=1500 - 65535 4 PULSE= 0 - 9999999

15. Ethernet 4 Datagram size 1500 4 Considerations –Data collision problems –Types of physical wiring –Size of local network

16. Token-Ring 4 Datagram size 2000-8000 4 Considerations –Multiple Ring Considerations –Different ring segment flow restrictions –Size of local network –Data collision problems

17. Direct connection of Router 4 CISCO or RS/6000 –Although the MTU size is large the overall destination may require a smaller MTU size 4 CTCA for VSE or VM –Tremendous control can be managed over the MTU size in the specific direction of flow

18. Hyper Sockets 4 64K mtu size 4 Pass the maximum data across to Linux 4 Properly identify the final destination –Multiple Link definitions depending

19. Fragmentation 4 IP Datagrams can be larger the can fit across a connection –Fragmentation creates Overhead at outbound time Overhead at inbound time –Can be eliminated if attention is paid MTU size in relation to the direction of flow Always pass a larger MTU size to a smaller one

20. Stack Data Flow 4 Data Units –The breaking of data into transmission units –The grouping of these units into transmission chunks 4 Flow Control –How is the flow managed –How is the flow impeded –How is the flow maximized

21. Transmission 4 MTU (Maximum transmission unit) –What does it mean –How does it effect the network 4 TCP segments –What is a segment size –How does it effect the network

22. TCP Windows 4 Theory –Exactly what is a window –How does it control data transfers 4 Performance –SET WINDOW –SET WINDOW_DEPTH –SET WINDOW_RESTART –SET CLOSE_WINDOW

23. Silly Window Syndrome 4 General TCP/IP buffered implementation –Fixed buffer is the size of WINDOW –Window decreases as data is read from the buffer –Once a close window has occurred. 4 Performance –SET ADDITIONAL_WINDOW

24. Application Flow 4 Far more tradition in tuning 4 Network data arriving and departing –Controlled by size and number of buffers 4 Directly effected by performance of I/O devices

25. FTP Performance of Files 4 Sequential disk –Variable –Fixed 4 VSAM –ESDS –KSDS 4 Power 4 Custom Built

26. FTP Performance Considerations 4 CPU utilization –Objective –Controls 4 Using set maximum_buffers –Theory –Practice –Control over CPU

27. FTP Performance Controls 4 List of things to check when CPU is high –TCP/IP partition utilization –Network utilization –Hesitate= 4 Running more then one TCP/IP partition

28. FTPBATCH for Multi-processors 4 Allows greater used of multiple processors –Concurrent I/O activity –Concurrent Buffer creation and processing 4 SET BUFFSIZE nnn 4 SET BUFFMAX nnn

29. TN3270 Performance Considerations 4 Telnet buffers –Pooled telnet daemon –Non-pooled telnet daemon 4 CPU utilization –Telnet daemon flow –Effect of larger number of daemons 4 CAF performance enhancements –VTAM elimination –CPU reduction

30. TN3270 Performance Controls 4 Number of defined daemons 4 Sequence ordering –Unordered definitions –Specific definitions –Forced connections 4 The use of FULL_SCAN 4 Control with POOL=YES and POOL=NO

31. API Considerations 4 Language –Assembler –C using the BSD interface –PL/1 and Cobol 4 Protocol –UDP –TCP 4 Buffering