Try this. In taskmgr process list, look for cmd.exe instances where command line is "%GRID_HOME%\bin\cluvfy comp health -_format". On the node asm conn failed, there should be lots of processes the same as this. First, kill all of this processes once a week or two. This is current available workaround and it seems there is no fix for this bug.
If it doesn't help, there should be long running SQLs like "SELECT * FROM gv$archive_dest_status". Hundreds of these SQLs cause Network wait event. Kill all of these sessions.
Do I need to end process or End process tree?. Is there is any path available to fix this ?
Hi , Again I can see lot of conhost.exe also in the task manager. Does this also related to the above cmd.exe?
Can I Kill that also?
Its bloody WINDOWS !!!!!!!!!!!!!!!!!! What do you expect ?
Switch to Linux !!!
Kill process tree. conhost executables will be ended when you kill cmd executables. I found no fixes. I'm also planning to switch to oracle linux.
If your db have lots of network waits caused by something like select apply_gap from db_archivelog_status, kill them too. And all will be clean
What is the Desktop heap size?
check it on following location,
Edit the Windows value of Type REG_EXPAND_SZ.
It will look similar to:
%SystemRoot%\system32\csrss.exe ObjectDirectory=\Windows SharedSection=1024,3072,512 Windows=On SubSystemType=Windows ServerDll=basesrv,1 ServerDll=winsrv:UserServerDllInitialization,3 ServerDll=winsrv:ConServerDllInitialization,2 ProfileControl=Off MaxRequestThreads=16
At this location the following set of values can be seen for "SharedSection",
1024,3072,512. The third value (512) is the size of the desktop heap for each desktop that is associated with a "noninteractive" window station.
And monitor the system for few days.