There is a known bug in the 7.0 MP’s for SQL that cause WMI to show out of memory errors when monitoring randomly. I have a case open with MS on this and they have confirmed the issue and it has been escalated to the product team. I would watch for an update to the MP coming out in the next few weeks.
In the most recent release, SQL MP team presented a change to reduce the number of WMI request established by the MP, but it didn’t solve the “out of memory” problem in all environments where it had existed. It looks like the problem can’t be solved from the MP side.
Which version of the SQL 2014 Always On MP do you have? Sounds like there could be a memory leak in the script it mentions. Or, if you correlate the time of the script failing vs SCOM perf data for memory on these servers, is there any sign that the server is running with high memory usage?
Microsoft SQL Server 2014 Always On (Monitoring) version 7.0.0.0 .there is no sign that the server is running with high memory usage during the time of alert