November 18, 1998
If you have a network with more than 15 nodes running software release 9.1.x, an event that causes the BNI to be reprogrammed, this network may cause database corruption, which would cause a BPX node on this network node to abort logging a 514 software error or 1M3 abort. This problem could occur in a network consisting of more than 15 BPX nodes, however, it is more likely to occur in a network larger than one with 30 or 40 BPX nodes. A BPX node with this problem will continue to abort until it is removed from the network. Any one of the following events can cause the BNI to reprogram:
If you switch from the active to the standby processor card.
Software upgrade to software release 9.1.x.
If you do a dncd or upcd command for a BNI.
If you do a resetcd command on the BNI card.
In software release 9.1, a feature was added where several channels are programmed with only one CommBus message, which is sent from the processor card, in order to improve the performance. There is a problem in the code such that the loop will not terminate properly if there are more than 15 nodes on the network. This causes memory corruption when the CommBus message is being built beyond 15 nodes.
The exact combination of the number of BNI trunks and the number of nodes that may cause this problem is unknown. This problem was reproduced in the lab with 30 BPX nodes. In a smaller network, the problem has not been reported.
This problem does not occur if the network is a pure IGX/IPX network. The problem also doesn't occur in a mixed network with IPX/IGX/BPX without any BNI trunks.
The node will continue to experience software error 514, or software abort 1,000,003.
Remove the node that is reporting the 514 aborts from the network. Issue a clrcnf command on the node. Upgrade the network and the isolated node to software release 9.1.06. Read the node to the network. Software release 9.1.06 is scheduled for release on December 4th 1998.
To follow the bug ID link below and see detailed bug information, you must be a registered user and you must be logged in.
Memory corruption in BNI Network channal programming may leads to abort.
For More Information
If you require further assistance, or if you have any further questions regarding this field notice, please contact the Cisco Systems Technical Assistance Center (TAC) by one of the following methods:
Receive Email Notification For New Field Notices
Product Alert Tool - Set up a profile to receive email updates about reliability, safety, network security, and end-of-sale issues for the Cisco products you specify.