1 Reply Latest reply: Nov 27, 2012 8:28 AM by stefanik RSS

    LDM0 trace explanation

    stefanik
      Hy,
      in a recent "surf" on the ldm0 trace of our production DB (oracle rac 11.2.0.3 with 6 nodes) i found this peace of file:

      Global blockers dump end:-----------------------------------
      Global Wait-For-Graph(WFG) at ddTS[0.736b7] :
      BLOCKED 0x1a4d70ee80 3 wq 2 cvtops x1001 TX 0x1764003e.0x757e(ext 0x6,0x0)[80000-0005-0000E9EB] inst 5
      BLOCKER 0x1d38cbfcd8 3 wq 1 cvtops x28 TX 0x1764003e.0x757e(ext 0x6,0x0)[9C000-0004-0000C3EC] inst 4
      BLOCKED 0x1b339ca1d0 3 wq 2 cvtops x1001 TX 0xde4003c .0xd72a(ext 0x6,0x0)[9C000-0004-0000C3EC] inst 4
      BLOCKER 0x1d49020e20 3 wq 1 cvtops x28 TX 0xde4003c .0xd72a(ext 0x6,0x0)[A0000-0004-00010980] inst 4
      BLOCKED 0x19dbe41ee8 3 wq 2 cvtops x1 TX 0x160c003e.0x1812(ext 0x7,0x0)[A0000-0004-00010980] inst 4
      BLOCKER 0x1d587d7a08 3 wq 1 cvtops x28 TX 0x160c003e.0x1812(ext 0x7,0x0)[7C000-0005-0000C518] inst 5
      BLOCKED 0x1b9ae8e528 3 wq 2 cvtops x1001 TX 0x1647000b.0x1175(ext 0x7,0x0)[7C000-0005-0000C518] inst 5
      BLOCKER 0x1d17115110 3 wq 1 cvtops x28 TX 0x1647000b.0x1175(ext 0x7,0x0)[80000-0005-0000E9EB] inst 5
      * Cancel deadlock victim lockp 0x1a4d70ee80

      It seems that i'm experimenting a deadlock in TX mode 4 but im not able to identify the objects involved.
      In particular, the ex number after TX will be the resource number, but how can i identify the object_id?
      I convert it with select to_number('1647000b','xxxxxxxx') from dual but there is anything!
      Also for select to_number('1647000b1175','xxxxxxxxxxxx') from dual.

      Can enyone explain me the meaning of the columns of the trace?

      Tanks
        • 1. Re: LDM0 trace explanation
          stefanik
          i've found on metalink this article: ID 1443482.1 in witch they say:


          The causes for TX deadlock in S mode wait can be:

          a. ITL contention, eg: INITRANS setting for the object is too small, it can not handle the number of concurrent transactions.

          The solution is to increase INITRANS setting for the object involved in the deadlock using "alter table" or "alter index" command

          b. If the object invovled is an unique key index, the wait could be cauesd by uniqueness validation. Application need to be checked to aovid unique key violation.


          tanks.