Processes consuming 100% CPU but SPID not in v$process
On my 12.1.0.2 dataguard standby (running on RHEL 7.2), I regularly see high CPU usage, caused by processes which don't show up in v$process.
e.g. right now I'm looking at the output of top, and the top lines are:
8112 oracle 20 0 10.320g 41276 30092 R 99.7 0.3 2:08.38 oracle_8112_****
7329 oracle 20 0 10.332g 44104 26368 R 99.3 0.3 40:56.26 oracle_7329_****
5030 oracle 20 0 10.327g 6380 3644 S 0.7 0.0 295:09.60 ora_pr00_****
Running the query:
select * from v$process where spid in (8112,7329);
Returns no rows.
This is very similar to 285026.1, but I've got EXPIRE_TIME commented out in sqlnet.ora, and this is a _much_ more recent version than 9i!
I've tried running strace on the processes involved, and it just shows a _lot_ of clock requests, presumably for ASH/monitoring purposes.