HP Neoview Release 2.4 Software - Manual (Page 86)

These three views provide access to R2.4 query and runtime statistics information. This information
is maintained on the Neoview platform for your convenience, but it will not change, nor are the
tables maintained. These three views use the pre-R2.4 SP1 "two rows per query" semantics of
query statistics.
NOTE: The old data displayed in those views will not be transferred to the new Repository
table and cleaned up. Also, the old Repository tables will no longer collect new data after Release
2.4 SP1 is installed. To remove old Repository data (and the tables that contain it), contact your
HP service provider.
NOTE: Although data collected in Release 2.4 is preserved by the installation script, no tools
are provided for migrating existing data to Release 2.4 SP1 tables.
Also, in certain cases HP Support might explicitly uninstall an older Repository version before
upgrading to Release 2.4 SP1. Conditions that might necessitate an uninstall include
• System malfunctions affecting the database
• Database malfunctions that cause tables to be corrupted
• Missing prerequisite software
• Mismatches between the database and Repository processes (for example, if a prior
installation of Repository software succeeded but the corresponding database upgrade
failed)
In such cases, accumulated data will not be retained. Please consult your HP representative prior
to a platform upgrade if you require preservation of old Repository data.
VIEW NEO.HP_METRICS.ODBC_QUERY_STATS_V1_2400
This view preserves separate rows for data collected at the start and end of a query. If information
is not available for a field (because the record represents the start of a session and the field is not
populated until the end), the value in that field is NULL. To distinguish easily between start and
end records, you can filter results according to the value of STATEMENT_STATUS.
Character columns use ISO88591 or UCS2 encoding, as appropriate to the Neoview platform
configuration. If the platform is UCS2 or SJIS, all character columns are created with CHARACTER
SET UCS2. If the platform is ISO88591, all CHAR columns are created with CHARACTER SET
ISO88591.
The primary key for the ODBC_QUERY_STATS_V1_2400 field definition is SEGMENT_ID,
NODE_ID, SEQUENCE_NUM and the date/time fields (QUERY_EVENT_DATETIME_UTC and
other date fields).
Table C-2 ODBC_QUERY_STATS_V1_2400 Field Definitions
Description Data Type Field Name
Segment number of the NDCS server
that reported the statistics. INTEGER UNSIGNED SEGMENT_ID
Processing node number of the NDCS
server which reported the statistics.
This is the node where the master
executor process which executed the
query was running. INTEGER UNSIGNED NODE_ID
ANSI SQL timestamp in Local Civil
Time when the statistics were
recorded in the Repository. TIMESTAMP(6) QUERY_EVENT_DATETIME
ANSI SQL date in Local Civil Time
when the statistics were recorded in
the Repository. DATE QUERY_EVENT_DATE
ANSI SQL time in Local Civil Time
when the statistics were recorded in
the Repository. TIME(6) QUERY_EVENT_TIME
86 Pre-R2.4 SP2 Query Statistics Views