IBM Support

PH28725: BIG SQL QUERY MAY FAIL WITH SQL1229N WHEN ONE OR MORE WORKER NODES IS DOWN

Subscribe

You can track all active APARs for this component.

 

APAR status

  • Closed as fixed if next.

Error description

  • A Big SQL query may fail with SQL1229N if one or more nodes are
    down. In Big SQL, multi-node
    subsections should be runtime partitioned. In a rare scenario,
    the query's access plan may
    include a subsection that is multi-node, but not flagged as
    runtime partitioned.  That subsection
    will run on a fixed set of nodes, instead of a set of nodes
    dynamically determined by the
    scheduler at runtime. As a result, the query cannot complete if
    one node fails, returning sqlcode
    -1229.
    
    This issue may be seen if the query is accessing complicated
    (i.e. nested) views.  If you run
    EXPLAIN and db2exfmt against the query, in the optimized
    statement you might see an
    outer join with a VALUES expression; for example:
    
    RIGHT OUTER JOIN
    (SELECT
       NULL,
       NULL,
       NULL
     FROM
       (VALUES
       ) AS Q2
    
    and in the access plan, you will see the above outer join that
    the partitioning specifies
    "Replicated Data on all nodes"
    
    i.e
     Input Streams:
     -------------
             1) From Object SYSIBM.GENROW
                     Estimated number of rows:       0
                     Partition Map ID:                       1
                     Partitioning:                   (REPL )
                                                     Replicated Data
    on all nodes
                     Number of columns:              0
                     Subquery predicate ID:          Not Applicable
                     Partition Column Names:
                     ----------------------
                     +NONE
    

Local fix

  • 1) Ensure all nodes are up and running.
    2) If a node will not stay up, remove the problematic node from
    the cluster.
    3) Try to run the query with a different optimization level, to
    generate a
    different access plan.
    i.e SET CURRENT QUERY OPTIMIZATION 3
    

Problem summary

  • Please see problem description.
    

Problem conclusion

Temporary fix

Comments

APAR Information

  • APAR number

    PH28725

  • Reported component name

    IBM BIG SQL

  • Reported component ID

    5737E7400

  • Reported release

    504

  • Status

    CLOSED FIN

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Special Attention

    NoSpecatt / Xsystem

  • Submitted date

    2020-08-21

  • Closed date

    2020-09-09

  • Last modified date

    2020-09-09

  • APAR is sysrouted FROM one or more of the following:

  • APAR is sysrouted TO one or more of the following:

Fix information

Applicable component levels

[{"Line of Business":{"code":"LOB10","label":"Data and AI"},"Business Unit":{"code":"BU059","label":"IBM Software w\/o TPS"},"Product":{"code":"SSCRJT","label":"IBM Db2 Big SQL"},"Platform":[{"code":"PF025","label":"Platform Independent"}],"Version":"504"}]

Document Information

Modified date:
10 September 2020