Sudden reboots and partition failovers due to insufficient memory in VDX systems with large MAC tables
In a VCS with a large MAC table, occasional reboots due to insufficient memory. Sudden partition failovers due to insufficient memory or memory allocation failures followed by interface counters being cleared.
Large MAC table
Frequently changing MAC table
An increase in the rate of MAC age-outs due to too short MAC aging-time can in fact lead to memory depletion, suddent reboots, partition failovers, interface counter resets, and, counterintuitively, dynamic MAC tables that balloon in size due to failure to age out MAC addresses.
Increase the dynamic and conversational mac aging-times. This will in turn decrease the rate of MAC age-outs and decrease the workload on all VDXs in the VCS. Even if the size of the dynamic and conversational MAC tables appear to be near the maximum, increasing aging times will in fact allow all VDXs in the VCS to age out MAC addresses more smoothly and lead to smaller MAC tables.
For example: mac-address-table aging-time 600 mac-address-table aging-time conversational 600
Furthermore, decreasing MAC aging-times to lower than the default 300 can in fact make problems worse. If problems occur after decreasing MAC aging-times lower than the default, first try returmning them to their default values of 300 seconds.