accelerate your business HA Considerations in a WAS

accelerate your business Introduction To High Availability § High Availability – System and Infrastructure Designed to Minimize or Eliminate Outages § Continuous Availability – “Nonstop Service” § Service Levels – Providing a Defined Availability Level IBM Israel Software Symposium 2003 3

accelerate your business The “ 5 Nines” Typical Non-clustered Unix Availability IBM Israel Software Symposium 2003 4

accelerate your business SPOF Elimination Technologies § Hardware/OS Clustering – Example Products • HACMP, Sun. Cluster , MCService. Guard, MS Clustering, Veritas Cluster Server § Software Clustering – Example Products • UDB/DB 2 ESE/DPF, Oracle RAC, Sybase Replication Server – Software Clustering Relies on Database Replication IBM Israel Software Symposium 2003 6

accelerate your business Active-Active / Active-Standby Client Server 1 Client Server 2 Active-Active IBM Israel Software Symposium 2003 Active Standby Active-Standby 7

accelerate your business Total System Availability 98. 5 % 99. 5 % 98. 1 % 99. 6 % 98. 2 % Network Switch /Router 1. 5 % + 0. 5 % + 1. 9 % + 0. 4 % + 1. 8 % + 2. 6 % + 1. 8 % + 0. 1 % + 0. 2 % + 0. 4 % + 0. 3 % Http Server Firewall Browser 97. 4 % ISP App Server 98. 2 % 99. 9 % Dist. DB Server Load balancer SNA WAN Gateway Host CICS/DB 99. 6 % 99. 7 % 99. 8 % Total system availability including all elements is : IBM Israel Software Symposium 2003 Firewall 87% 9

accelerate your business WAS Terminology § Application Server – Application Server Process § Node – Physical Machine hosting 1 or more App Servers § Node Agent – Manages App Servers on a Node (1 per node) § Cell – Administrative Domain (1 or more nodes) § Deployment Manager – Manages Cell via Node Agents § Cluster – One or more App Servers in a Cell IBM Israel Software Symposium 2003 11

accelerate your business Topology Components Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS WES Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 12

accelerate your business HTTP Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS WES Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 13

accelerate your business Web. Sphere Servlet Requests § HTTP Server Plug-in – Detects Failure – Marks Container as Unavailable Web Container – Tries Next Server in Server. Cluster App Server Servlet Requests HTTP Server Plug-in HTTP/S Protocol Traffic Web Container App Server IBM Israel Software Symposium 2003 14

accelerate your business Plug-in Configuration (plugin-cfg. xml) § Configuration Refresh – <Config Refresh. Interval=60> § Connect. Timeout – # of seconds plugin tries to contact an app server process – Default is zero (not specified prior to V 5. 02) Plugin will block many requests until OS times out attempting to connect to with dead machines – Server Clone. ID>=“xxx” Name=“clone 2” Connect. Timeout=“ 10”> § Load. Balance. Weight – Weight assigned to server for load distribution – Corresponds to Server Weight in SM § Retry. Interval – Depends on # of Servers TCP/IP Timeout for OS – Too Small & Performance Degrades – Recommended 10 Seconds + (# of clones * TCP Timeout) – Note: Every Rety. Interval every server/process will block one request for Connect. Timeout seconds IBM Israel Software Symposium 2003 15

accelerate your business HTTP Request HA HTTP Server HTTP Requests WES (Primary) HTTP Requests TCP/IP Traffic HTTP Server WES (Secondary) IBM Israel Software Symposium 2003 16

accelerate your business Web Container – Servlet/JSP Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS IP Sprayer Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 18

accelerate your business Session Manager – Failover Options § Database – Utilize Database HA § Memory-To-Memory Replication – Build on Small, Fast 'pub/sub' Engine – Runs in an Existing Server Process – Separate Thread per Queue – Can be set up Peer-to-Peer or Client-Server Default is Peer-to-Peer – Multiple 'channels' available for partitioning – Messages can be encrypted - DES or Triple DES – Does not provide a “persistent” store § Both Perform Essentially the Same!! – Object Serialization/Deserialization is 95% of Performance “Cost” IBM Israel Software Symposium 2003 19

accelerate your business Session Manager § Updates Occur – At End of the Servlet Service Method • The Default Prior to V 5. 0 – Manually • Requires Use of IBM Extension to Http. Session – At the End of a Specified Time Interval • V 5. 0/V 5. 01 Default Time Interval Defaults to 120 Seconds Recommended Tuning of 10 Seconds for Performance and Failover IBM Israel Software Symposium 2003 20

accelerate your business Memory To Memory Replication – Configurations Handling single point of failure (SPOF) Web Container WAS (store) Web Container WAS WAS (Store) Web Container WAS (store) Web Container WAS (Store) WAS Web Container WAS (store) IBM Israel Software Symposium 2003 Dedicated App Servers for HTTP Session Store 22

accelerate your business EJB – RMI/IIOP Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS IP Sprayer Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 23

accelerate your business EJB – Two Major Steps § Locating the EJB Home in the Name Server … Initial. Context ic = new Initial. Context () ; Object o = ic. lookup (“java: comp/env/ejbs/Customer. Home”) ; … § Invoking methods on remote objects … Customer cust = customer. Home. find. By. Primary. Key (id) ; cust. get. Name () ; . . . IBM Israel Software Symposium 2003 24

accelerate your business IIOP WLM v 5. 0. . . Hashtable env = new Hashtable(); env. put( Context. INITIAL_CONTEXT_FACTORY, 7: Additional EJB Lookup 1: "com. ibm. websphere. naming. Wsn. Initial. Context. Factory"); EJB Container App. Server 1 5: env. put( Request EJB Method Requests Server 1 Context. PROVIDER_URL, "corbaloc: : myhost 1: 9810, : myhost 1: 9811, : myhost 2: 9810"); Context initial. Context = new Initial. Context(env); WLM Info +. . . IOR 6: EJB Method Result & EJB Container App. Server 2 Server Routing List Server 2 1: Initial Request 2: Indirect IOR Name Service 4: IOR for EJB Server & WLM Context 3: Direct IOR EJB Container Server 3 LSD App. Server 3 Client JVM Process IBM Israel Software Symposium 2003 Server Cluster Node. Agent 25

accelerate your business Web. Sphere IIOP Requests § EJB Requests EJB Container Web Container App Server Java Client App Server IIOP Traffic EJB Container App Server IBM Israel Software Symposium 2003 26

accelerate your business IIOP Failover § Handled by SM Runtime and ORB Plug-in – Application Servers • • Deployment Manager/Node Agent is “parent” to All Application Servers State Changes Are “pushed” to Application Servers – Clients • • Updated with Response from Application Server Epoch Number Change is Used as Indicator Marks Server as Unavailable After Failures com. ibm. ejs. wlm. Max. Comm. Failures on command line com. ibm. ejs. wlm. Unusable. Interval on command line – Time out for network requests (when machines fail) • • OS TCP/IP timeout (prior to connection being established) com. ibm. CORBA. request. Timeout on command line (after connection is established) – Avoid Dynamic Port Assignment after App Server Restart and for Firewalls • Configure End Points in SM (Admin Browser) ORB_LISTENER_ADDRESS com. ibm. CORBA. Listener. Port in V 4 and before SAS_*_LISTENER_ADDRESS and CSIv 2_*_LISTENER_ADDRESS IBM Israel Software Symposium 2003 27

accelerate your business IIOP Failover Extensions in V 5. 02 Enterprise § Dynamic WLM – Weights Adjusted Based On PMI Feedback – Adjusts Every 7 Seconds (same as Edge Server Default) – V 5. 02 Adjustment Based on • • CPU Utilization - 88% Response Time - 2% EJB + 2% Web Conc. Requests - 2% EJB + 2% Web Total Req. Rate - 2% EJB + 2% Web – Not Configurable/Customizable in V 5. 02 § Cross Domain/Cell Failover – Requires Identical Cluster in Two Cells (Mirroring) • Same Application and Cluster Name • Number of Cluster Members Can Differ IBM Israel Software Symposium 2003 28

accelerate your business EJB § Stateless Session Bean – WLM by infrastructure – No HA issues § Stateful Session Bean – WLM on lookup only – Doesn’t support auto-faliover (according to J 2 EE Spec. ) § Entity Bean – State is in persistent store (usually DB) – WLM with session and TX affinity – TX log for XA Transactions IBM Israel Software Symposium 2003 29

accelerate your business JMS Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS IP Sprayer Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 30

accelerate your business Embedded JMS Server Deployment Manager Node Agent Web. Sphere JMS Server Web. Sphere App Server MDB MDB IBM Israel Software Symposium 2003 31

accelerate your business HA Embedded JMS Server Heartbeat HA Cluster SW Web. Sphere JMS Server (Primary) HA Cluster SW Mirrored Web. Sphere JMS Server (Secondary) Disks Node Agent Web. Sphere App Server MDB MDB IBM Israel Software Symposium 2003 32

accelerate your business HA Web. Sphere MQ Heartbeat HA Cluster SW Web. Sphere MQ (Primary) HA Cluster SW Mirrored Web. Sphere MQ (Secondary) Disks Node Agent Web. Sphere App Server MDB MDB IBM Israel Software Symposium 2003 33

accelerate your business Systems Management Deployment Mgr. HTTP Client Database(s) HTTP Server WAS WAS App App Node Agent WAS 5 ND JMS IP Sprayer Internet HTTP Server WAS WAS App App Node Agent LDAP WAS 5 ND Firewall IBM Israel Software Symposium 2003 42

accelerate your business V 5 SM Failover – Deployment Manager § Deployment Manager failover not addressed in Network Deployment – Can be "Nannied" (Unix initab or Windows Service) § Consequence of Deployment Manager failure: – Unable to broadcast configuration changes to Node Agents – Admin Console unavailable – wsadmin unavailable (unless manually directed to specific server) – In short, you cannot make any changes to the central configuration § Deployment Manager handles IIOP WLM routing table 'masters' – If the Deployment Manager is down, failed or stopped servers will still have routed requests • Until plug-in marks as down • Performance degradation is possible if Deployment Manager stays down • IIOP WLM for New ORB Clients will “pin” to One Application Server In V 5. 01 New ORB Clients Utilize “stale” Server Cluster List IBM Israel Software Symposium 2003 43

accelerate your business V 5 SM Failover – Node Agent § Node Agents have copies of all configuration information – Can be "Nannied" (Unix initab or Windows Service) § Consequence of Node Agent failure – Local Configuration May Not Reflect Global Configuration Until Synchronization (restart) – Application Servers Cannot Be Started Until Node Agent is Restarted • Can Be Stopped Via command line. IBM Israel Software Symposium 2003 44

accelerate your business Simple HA Topology Cell Deployment Mgr. WAS App HTTP Server Node Agent WAS 5 ND HA Software DB HTTP Server WAS App Node Agent WAS 5 ND IBM Israel Software Symposium 2003 50

accelerate your business HA “Gold Standard” Cell 1 HTTP Server WAS WAS App App Node Agent WAS 5 ND Deployment Mgr. WES IP Sprayer HTTP Server WAS WAS App App Deployment Mgr. Node Agent WAS 5 ND MQ Clustering Cell 2 HTTP Server WAS WAS App App Node Agent Application Session & LDAP WAS 5 ND Deployment Mgr. WES IP Sprayer HTTP Server WAS WAS App App Deployment Mgr. Node Agent WAS 5 ND IBM Israel Software Symposium 2003 51

accelerate your business HA “Gold Standard” § Two (or More) Cells – Hardware Isolation – Software Isolation – Planned Maintenance – Insurance Against Catastrophic Outage – May Require More Administrative Effort § Don’t Forget “Rule of 3” – With 2 of “Everything”, an Outage (Planned or Unplanned) Reduces Capacity by 50% • Is No Longer Fault Tolerant IBM Israel Software Symposium 2003 52

accelerate your business HA Equipment Requirements § Minimum “Price of Admission” for HA – Firewall Servers =2 – IP/HTTP Sprayers =2 – HTTP Servers =2 – Application Servers =2 – Deployment Manager = 2 – JMS Servers =2 – Database Servers =2 – LDAP Servers =2 – Total = 16 § Some “Layer Compression” is Possible Through Collocation of Components – Security May Be Compromised – Administration More Difficult IBM Israel Software Symposium 2003 53

accelerate your business Web. Sphere HA Take-aways § Use HA Topology that suites your needs § Consider entire system § WAS has built in capabilities for HA – use them § Speak to IBM – we can help IBM Israel Software Symposium 2003 56

accelerate your business Additional Resources § Whitepapers – Establishing database failover support with High Availability Cluster Multi-Processing (HACMP) • http: //www-4. ibm. com/software/webservers/appserv/hascenario. html – A Highly Available & Scalable LDAP Cluster in an IBM AIX Environment • http: //www-1. ibm. com/servers/esdd/articles/ldap/index. html – Web. Sphere Connection Pooling • http: //www-4. ibm. com/software/webservers/appserv/whitepapers/connection_pool. pdf – Server Clusters For High Availability in Web. Sphere Application Server Network Deployment Edition 5. 0 • http: //www-1. ibm. com/support/entdocview. wss? rs=180&context=SSEQTP&q=&uid=swg 27002473&loc=en_US&cs=utf 8&lang=en – Implementing a Highly Available Infrastructure for Web. Sphere Application Server Network Deployment, Version 5. 0 without Clustering • http: //www 7 b. software. ibm. com/webapp/dd/transform. wss? URL=/wsdd/library/techarticles/0 304_alcott/alcott. xml&xsl. URL=/wsdd/xsl/document. xsl&format=one-column – Redbooks (http: //www. redbooks. ibm. com) • IBM Web. Sphere V 5. 0 Applications: Ensuring High Performance and Scalability SG 24 -619800 IBM Israel Software Symposium 2003 57

accelerate your business ? שאלות IBM Israel Software Symposium 2003 58

accelerate your business Don’t forget to give us feedback Presentation Code: A 10 IBM Israel Software Symposium 2003 59