Re: "control file sequential read" wait during RAC Benchmark ?

Vivek

Over what time period was the snapshot taken?

Q1: someone else can answer that, I certainly can't. Do you really care about 7 
seconds? Although this wait is a significant fraction of the total wait time, 
is the load meaningful? 

Q2: yes if possible. But IMHO, the worse the application, the more it helps to 
partition in this way (conversely, the better the application, the less there 
is need for or benefit from partitioning). You choose whether to tune the 
application code (eg by reducing redundant data reads etc) or workaround by 
partitioning. The first could  give you orders of magnitude better results - 
but will cost more... and for other vendors' COTS applications you may have no 
choice but workarounds.


Regards Nigel

----- Original Message ----
From: VIVEK_SHARMA <VIVEK_SHARMA@xxxxxxxxxxx>


Folks
 
During internal 2 Node RAC Benchmark OLTP Nature Transaction Runs, following 
are the TOP Events from Statspack.
 
Qs 1) How can the ?control file sequential read? wait be reduced after 
identifying the respective ?busy? Objects? 
Qs 2) Should Transactions be grouped Logically so that Transactions Hit 
different Data Blocks on the 2 Nodes?
 
Top 5 Timed Events                                                    Avg %Total
~~~~~~~~~~~~~~~~~~                                                   wait   Call
Event                                            Waits    Time (s)   (ms)   Time
----------------------------------------- ------------ ----------- ------ ------
CPU time                                                        12          43.6
control file sequential read                    21,969           7      0   25.4
...

Other related posts: