Are the units configured for active/active with multiple context and failove groups or you are saying that the units
keep failing over when configured as active/standby?
Either way, we need to know quite a few things before we can say anything on this matter.
1. asa code?
2. active/active or active/standby configuration?
3. sh fail history - output
4. what do the syslogs say during the time of the problem?
-KS