After successful installation of a two node Oracle RAC 11.2 on Windows 2008 R2 we investigate a daily growing number of cmd.exe all running the following statement:
*C:\Windows\system32\cmd.exe /K D:\oracle\cluster\11.2.0\grid\bin\cluvfy comp health -_format* Does anyone know the originator/source of this "cluvfy" command?
Maybe the Grid Control Agent?
We think, that the "/K" option is the problem, i.e. the reason while the cmd.exe will not terminate.
When we run the "cluvfy" manually everythings results successfully.
Thanks for your reply.
the cluvfy is either executed by the grid control agent as you expected, but with 22.214.171.124. the grid infrastructure is also executing cluvfy commands.
(See crsctl stat res -t and you will find the cluster resource ora.cvu, which is the clufvy tool called by the grid infrastructure).
Since you are using 126.96.36.199 I suspect you hit: Bug 12529945: CVU HEALTH CHECKS EXHAUST WINDOWS HEAP MEMORY
Best is to open an SR and request/ask for a bugfix... Maybe it is already fixed in the latest Patch # for 188.8.131.52 on Windows.
thanks for your quick reply. You were probably right with BUG 12529945.
Seems that the patch 12714463: ORACLE 11G 184.108.40.206 PATCH 8 BUG FOR WINDOWS (64-BIT AMD64 AND INTEL EM64)
has the fix included.
I've opened a SR ro request a patch for current 220.127.116.11.. There is no patch for this release ("Bug 1252994 is fixed in 18.104.22.168 Patch 8 and 22.214.171.124.")
Edited by: user597986 on 06.12.2012 04:28
I have done the same and get the run around from the support tech when I ask for clarification. Even asked duty manager for assistance and was sent right back to the support tech that I could not get conclusive answers to my question from the first time. Seems he is telling me that he opened a new bug up for my answer but the description of the problem is not the same so how can I believe him. He wants me to apply patch 12 but I thought the later patches set are supposed to contain previous patches, so why not apply patch 14 or newest. Seems none of the patches show that they solve the Windows Heap Issue for 126.96.36.199. Even the bug note says it will be fixed in 188.8.131.52. Just one more reason we are planning (although against my better judgement) to move away from oracle.