Download for Force 10 S50 v2.3.1.5

Resolving The Problem

You are now leaving the IBM Systems support Web site

IBM makes no representations or warranties about any other Web site which you may access through this one. When you access non-IBM Web sites, even though they contain the IBM logo and content regarding IBM's products and services, such Web sites are independent of IBM and IBM has no control over the operation of non-IBM Web sites. In addition, a link to a non-IBM Web site does not mean that IBM endorses that Web site or has any responsibility for the use of such Web site.

Force10 Networks

All S-Series customers should upgrade to version 2.3.1.9, including Routing Package (Layer 3) users and Switching Package (Layer 2) users.
As described below, Force10 Networks has identified a limited set of conditions under which the S50 might experience either a hang of management functions (console, telnet and SSH) and/or a failure to forward Layer 2 and/or Layer 3 traffic. A series of link flaps is most likely to precede any traffic disruption.

Symptom 1: Only Console/Telnet access is affected

When symptom 1 is evident, the S50 cannot be accessed via the console or Telnet session (“management hang”). Network traffic might or might not continue switching.
This condition occurs when SFTOS detects an improper release of an internal buffer; the CLI task is suspended, preventing access to the system via the console, Telnet, or SSH. The failure of this task should not affect other running tasks. This symptom does not affect SFTOS version 2.2.1.9 or version 2.3.1.5 or higher.

Symptom 2: All Layer 2 and Layer 3 traffic is affected

When symptom 2 pertains, the S50 experiences the management hang problem described for symptom 1. In addition, the switch ceases to perform Layer 2 forwarding, Layer 3 routing, and/or protocol operations (“system hang”). Specific manifestations of this symptom include:

• The switch experiences a complete system lockup; no traffic is passed through the system. A series of link flaps might precede this condition. To verify that link flaps occurred, review the syslog for a series of link up/ down messages.
• The switch experiences a partial system lockup; only Layer 2 traffic passes, but routing adjacencies are lost, and Layer 3 forwarding fails. A series of link flaps might precede this condition. To verify that link flaps occurred, review the syslog for a series of link up/down messages.
• If the primary unit fails in a stacking configuration, as described in the previous scenarios, a stack member will take over the management function as the primary unit of the stack in what appears to be a successful failover. However, the failed unit might or might not continue to forward traffic. Both the failed primary and the new primary switches will illuminate the primary LED (PRI) located on the front panel.