Problem Resource->rproccache_spin

Hi

Today in our Db we had strange problem.

1. 1.

Monitoring of spinlok on the basis of dbcc monitor works in my BD

Directly I write information on spinlock to sysmon.

As a result today during the whole day I receive next situation.

Resource Spinlock Activity Report

---------------------------------

Spinlock::Instance per sec per xact grabs waits

spins/wait contention

------------------------------------- ---------- ---------- ---------- ---------- -----

----- ----------

Resource->rproccache_spin::0              132322.1      186.9 4661282    4101986      344.0      88.0%

Resource->rdesmgr_spin::0                  1039.5        1.5 1001777      32226      120.8      3.2%

Resource->rlocksleeptask_spin::0            266.4        0.4      581837        8259        80.1      1.4%

Resource->rpssmgr_spin::0                    965.3        1.4 2126769      29923        69.9      1.4%

Resource->rpmctrl_spin::0                      1.0        0.0        4065 32    153493.7      0.8%

Resource->rlockobjpool_spin::0                0.9        0.0      11162          29      2087.1      0.3%

Resource->rdbts_spin::0                        3.4        0.0      47201        106        85.6      0.2%

Resource->rprocmgr_spin::0                    2.5        0.0      40565          77      129.2      0.2%

Resource->rlocksemaphore_spin::0              2.8        0.0      52591          87      234.9      0.2%

Resource->rdbt_spin::0                      222.8        0.3 4778877        6907      124.7      0.1%

Resource->rqueryplan_spin[i]::3                2.2        0.0      90442          67        97.7      0.1%

Resource->rqueryplan_spin[i]::2                0.8        0.0      50359          25        38.5      0.0%

Resource->rqueryplan_spin[i]::0                0.5        0.0      39838          16        24.2      0.0%

Resource->rqueryplan_spin[i]::1                0.2        0.0 26640          6        44.8      0.0%

And at the same time a utilization of engine extremely grows.

1. 2.

I know processes which appearance leads to a problem.( the process causing a problem)

This process launches multithreaded data handling.

At the same time working 3 similar process.(near 20 -25 spid )

The problem arises when it is launched 2 and more of them.

1 process can normally work.

1. 3.

In case of start of all 3 process, operation directly of this functionality begins to work extremely slowly.

Grows

Resource->rproccache_spin::0              132322.1      186.9 4661282    4101986      344.0      88.0%

At the same time all remained processes work stablly.

1. 4.

In case of switch-off of processes from the DBMS operation of process is normally work (in case of sharp reduction of connections).

1. 5.

Flags are at the moment switched on

Active traceflags: 753, 3604, 3973, 4086

Version

Adaptive Server Enterprise/15.7/EBF 25308 SMP SP135 HF1/P/RS6000/AIX 6.1/ase157sp133x/3936/64-bit/FBO/Thu Oct 8 05:16:23 2015

Sp_configure+sysmon in attach.(7z archiv)

What can be the cause of such behavior of operation of specific procedures?

What it is worth paying attention to?