AW: parallel_max_servers and the number of sessions involved in a SQL

Hi,  

only for information (in case someone has the same problem): Setting the parameter "parallel_force_local" to true, solved the problem.  

Best regards
Ahmed          

-----Original-Nachricht-----
Betreff: Re: parallel_max_servers and the number of sessions involved in a SQL
Datum: 2020-11-29T13:55:16+0100
Von: "Mikhail Velikikh" <mvelikikh_at_gmail.com> An: "Jonathan Lewis" <jlewisoracle_at_gmail.com>      

40 processes per CPU sounds like the default value of parallel_max_servers:
<https://docs.oracle.com/database/121/REFRN/GUID-B6CE9EA3-B846-4AA7-B668-FD3BEE433A24.htm#REFRN10158>
PARALLEL_THREADS_PER_CPU * CPU_COUNT * concurrent_parallel_users * 5   The number of concurrent parallel users running at default degree of   parallelism on an instance depends on the memory initialization parameter   settings for the instance. For example, if the MEMORY_TARGET or   SGA_TARGET initialization parameter is set, then the number of   concurrent_parallel_users = 4. If neither MEMORY_TARGET or SGA_TARGET is   set, then PGA_AGGREGATE_TARGET is examined. If a value is set for   PGA_AGGREGATE_TARGET, then concurrent_parallel_users = 2. If a value is   not set for PGA_AGGREGATE_TARGET, then concurrent_parallel_users = 1.  

PARALLEL_THREADS_PER_CPU = 2
concurrent_parallel_users = 4
It results in exactly 40 (=2 * 4 * 5).  

However, I don't see that this limit is honored in my 19.9 database, so that I am able to allocate more parallel processes despite the fact that documentation says about some adjustments:
<https://docs.oracle.com/en/database/oracle/oracle-database/19/refrn/PARALLEL_MAX_SERVERS.html#GUID-B6CE9EA3-B846-4AA7-B668-FD3BEE433A24>
  The lower of the two values is used as the default value of   PARALLEL_MAX_SERVERS, and if you attempt to explicitly set   PARALLEL_MAX_SERVERS to a value that is higher than either of the values,   then the setting is adjusted to the lower of the two values. The adjustment can be about reserved processes and entries such as these in the alert log (Second Alert_<Sid>.Log In ORACLE_HOME/dbs Directory Next To The Standard Alert Log Under ORACLE_BASE/admin (Doc ID 1322075.1)):   Sun Feb 20 11:33:46 PST 2011
  Adjusting the default value of parameter parallel_max_servers  

There are really two issues here:
1. the wrong results issue that Jonathan Lewis mentioned. It might be something like this: (DB36) Bug 20509482 - ORA-600 [3020], ORA-752 Wrong Results or RMAN ORA-600 [krcrfr_nohist] after Parallel Direct Load in RAC (caused by fix for bug 9962369) (Doc ID 2139374.1) 2. the DOP downgrade issue. That is a separate thing. I usually analyze the tracefile: alter session set events 'trace[px_scheduler]'; but it can be approached differently: How to View Why the Degree Of Parallelism (DOP) was Downgraded for an SQL in 12c (Doc ID 2011375.1)  

On Sun, 29 Nov 2020 at 12:30, Jonathan Lewis <jlewisoracle_at_gmail.com
<mailto:jlewisoracle_at_gmail.com> > wrote:
   

  This is why I asked you about all your parallel parameters and what   parameters you were leaving to default.   In 19.3 (for example) if you don't set "processes" then the default   number of processes is "80 * CPU_count + 40" - which almost looks like   Oracle deciding that it has to have 40 processes for the critical   background processes and a maximum of 80 processes per CPU is a sensible   limit (for an OLTP system).    

  In the same vein it's perfectly reasonable for someone in Oracle to   decide that if 80 processes per CPU is sensible for "normal" processing   then 40 per CPU is equally sensible for the "batch-like" processes of   parallel execution. In fact they might be thinking in terms of the impact   of 20 batch-like processes per CPU on the assumption that DOP 20 usually   gets 40 processes but only 20 of them are likely to be very busy at any   one instant.    

  Regards
  Jonathan Lewis    

  On Sat, 28 Nov 2020 at 17:15, ahmed.fikri_at_t-online.de   <mailto:ahmed.fikri_at_t-online.de> <ahmed.fikri_at_t-online.de   <mailto:ahmed.fikri_at_t-online.de> > wrote:     sorry my bad. Indeed, there is a correlation with the process number     and the pga (which also logical is):     After setting the pga_aggregate_target to 10M (an extreme value) and     processes to 1500, I got this correlation:      

    cpu (host) n_max

    1              40
    2              80
    3             120
    4             160
    5              200
    6              240
    7              280
    8              320
     

    The question now is where the value 40 comes from. (I'll also try this     test on 11.2)      



--
http://www.freelists.org/webpage/oracle-l