Re: profiling connection overhead

On Wed, Nov 24, 2010 at 4:05 PM, Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> (You might be able to confirm or disprove this theory if you ask
> oprofile to count memory access stalls instead of CPU clock cycles...)

I don't see an event for that.

# opcontrol --list-events | grep STALL
INSTRUCTION_FETCH_STALL: (counter: all)
DISPATCH_STALLS: (counter: all)
DISPATCH_STALL_FOR_BRANCH_ABORT: (counter: all)
DISPATCH_STALL_FOR_SERIALIZATION: (counter: all)
DISPATCH_STALL_FOR_SEGMENT_LOAD: (counter: all)
DISPATCH_STALL_FOR_REORDER_BUFFER_FULL: (counter: all)
DISPATCH_STALL_FOR_RESERVATION_STATION_FULL: (counter: all)
DISPATCH_STALL_FOR_FPU_FULL: (counter: all)
DISPATCH_STALL_FOR_LS_FULL: (counter: all)
DISPATCH_STALL_WAITING_FOR_ALL_QUIET: (counter: all)
DISPATCH_STALL_FOR_FAR_TRANSFER_OR_RESYNC: (counter: all)

# opcontrol --list-events | grep MEMORY
MEMORY_REQUESTS: (counter: all)
MEMORY_CONTROLLER_PAGE_TABLE_OVERFLOWS: (counter: all)
MEMORY_CONTROLLER_SLOT_MISSED: (counter: all)
MEMORY_CONTROLLER_TURNAROUNDS: (counter: all)
MEMORY_CONTROLLER_BYPASS_COUNTER_SATURATION: (counter: all)
CPU_IO_REQUESTS_TO_MEMORY_IO: (counter: all)
MEMORY_CONTROLLER_REQUESTS: (counter: all)

Ideas?

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Company