Possible single points of failure in the WebSphere system
Table 8-3 lists potential single points of failure in the WebSphere system and possible solutions.
Table 8-3
Failure point Possible solutions Note Entity EJBs, Application DB HA DBs, parallel DBs Catch StaleConnectionException and retry Log files HA DBs, parallel DBs, clustering, HA files system, disk mirroring, RAID-5 LDAP HA LDAP, master-replica, clustering, sprayer Manual recovery WebSphere Deployment Manager Multiple WebSphere domains (cells), OS-service, hardware-based clustering WebSphere Master Repository Data HA shared file system, Networked file system, hardware based clustering WebSphere Node Agent Multiple Node Agents with WebSphere built-in HA LSD, OS-service, hardware-based clustering WAS EJB WLM, servlet clustering, hardware-based clustering Use different appservers/nodes for EJBs and servlets Web server Multiple Web servers with network sprayer, hardware-based clustering Load balancer HA network sprayers Firewall Firewall clustering, HA firewall, firewall sprayer Internal network Dual internal networks Hubs Multiple interconnected network paths NIC failures Multiple NICs in a host Disk failures Disk mirroring, RAID-5 Disk bus failure Multiple buses Disk controller failure Multiple disk controllers Network service failures (DNS, ARP, DHCP, etc.) Multiple DNS, etc. network services OS or other software crashes Clustering, automatically switch to a healthy node Host dies Clustering, automatically switch to a healthy node Power outages Two-power systems Room disaster (fire, flood, etc.) Put the system in a different room Floor disasters (fire, flood, etc.) Put the system on a different floor Building disasters (fire, flood, tornado, etc.) Put the system in a different building City disasters (earthquake, flood, etc.) Remote mirror, replication, geographical clustering Region disasters Put two data centers far away with geographical clustering or remote mirroring People error Train people, simplify system management, use clustering and redundant hardware/software, disable DDL operations Software upgrades Rolling upgrades with clustering and/or WLM for 7x24x365, planned maintenance for others For 7x24x365, suggest using two-domain WAS; single domain with WLM works, but is not suggested. Hardware upgrades Rolling upgrades with clustering and/or WLM for 7x24x365, planned maintenance for others Possible SPOF in the WebSphere system
Prev | Home | Next WebSphere is a trademark of the IBM Corporation in the United States, other countries, or both.
IBM is a trademark of the IBM Corporation in the United States, other countries, or both.