Hi
I have an issue when extracting data via ODBC.
Progress 10.1C on Win2003 server////4 quad core processors 4Gb RAM
DataDirect OpenEdge 10.1C Driver 5.30.00.74 configured as READ_UNCOMMITTED and FETCH_ARRAY_SIZE = 50 NO TRACE
The issue is that when extracting data via ODBC the perfomance degrades over the period of the extract.
an example of this is as follows:
SELECT * FROM TABLE.....2.8M rows in table
Elapsed time ..........30sec----60secs----90secs----120secs----150secs
Row count returned 670000 1327000 1445000 1519000 1560000
Rows/sec ........22333 21900 3933 2466 1366
To investigate further I am running SQLDBX on this server to return results and eliminate any network bottlenecks.
What I find is that if I run the following
SELECT PK_COLUMN FROM PUB.TABLE
I can return all 2.8m rows in under 7 secs
If I run a single column select on any indexed column I get similar sub 7 sec performance
However if I run a single column select on a non indexed column the extract performance degrades rapidly after about 1.3M rows.
....and if i run a two column select (both indexed) then it still slows down.
The data is being extracted from a 3rd pty Progress Solution and we are not in control of the start up parameters nor do we have a developer licence.
Has anyone any suggestions on why we experoence this slow down.
many thanks
I have an issue when extracting data via ODBC.
Progress 10.1C on Win2003 server////4 quad core processors 4Gb RAM
DataDirect OpenEdge 10.1C Driver 5.30.00.74 configured as READ_UNCOMMITTED and FETCH_ARRAY_SIZE = 50 NO TRACE
The issue is that when extracting data via ODBC the perfomance degrades over the period of the extract.
an example of this is as follows:
SELECT * FROM TABLE.....2.8M rows in table
Elapsed time ..........30sec----60secs----90secs----120secs----150secs
Row count returned 670000 1327000 1445000 1519000 1560000
Rows/sec ........22333 21900 3933 2466 1366
To investigate further I am running SQLDBX on this server to return results and eliminate any network bottlenecks.
What I find is that if I run the following
SELECT PK_COLUMN FROM PUB.TABLE
I can return all 2.8m rows in under 7 secs
If I run a single column select on any indexed column I get similar sub 7 sec performance
However if I run a single column select on a non indexed column the extract performance degrades rapidly after about 1.3M rows.
....and if i run a two column select (both indexed) then it still slows down.
The data is being extracted from a 3rd pty Progress Solution and we are not in control of the start up parameters nor do we have a developer licence.
Has anyone any suggestions on why we experoence this slow down.
many thanks