Re: patch to DBD-Oracle to allow faster fetch from REF CURSORs

------=_Part_7292_11728057.1190122496613
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,

"fetchall_arrayref" is just as slow for REF CURSORs because internally, OCI
still does a DB roundtrip for each row.

Running a test will illustrate the issue. Here is a result of a test run:
++++++++++++++++++++++++++++++++++++++++++++++++++
Fetched 2500 rows (select): [0.06174] secs.
Fetched 2500 rows (ref cursor): [5.40367] secs.
Fetched 2500 rows (ref cursor array-fetch): [0.631323] secs.
++++++++++++++++++++++++++++++++++++++++++++++++++

Biswa





Alexander V Alekseev
Fri, 14 Sep 2007 12:42:39 -0700

                Hello!

On Fri, 14 Sep 2007, Biswadeep Chowdhury wrote:

    I am listing the patch (diff file).
    The REF CURSOR array-fetch can be turned on using something like
    "$dbh->{RefCursorCacheSize}=16;".
    Array fetch remains off by default (single row fetch) so that it will
not
    affect users who dont care about it.
    I have added tests to the end of "50cursor.t" that you may execute to
view
    performance improvements.
    I hope you will consider it for the next version.

        Shouldn't it be implemented as fetchall_arrayref( $slice,
$max_rows ) ? DBD::Oracle lacks it.

        Bye. Alex.

------=_Part_7292_11728057.1190122496613--
0
cbiswa
9/18/2007 1:34:56 PM
perl.dbi.dev 1960 articles. 0 followers. Follow

12 Replies
20812 Views

Similar Articles

[PageSpeed] 6
Get it on Google Play
Get it on Apple App Store

 		Hello!

On Tue, 18 Sep 2007, Biswadeep Chowdhury wrote:

> "fetchall_arrayref" is just as slow for REF CURSORs because internally, OCI
> still does a DB roundtrip for each row.
>
> Running a test will illustrate the issue. Here is a result of a test run:
> ++++++++++++++++++++++++++++++++++++++++++++++++++
> Fetched 2500 rows (select): [0.06174] secs.
> Fetched 2500 rows (ref cursor): [5.40367] secs.
> Fetched 2500 rows (ref cursor array-fetch): [0.631323] secs.
> ++++++++++++++++++++++++++++++++++++++++++++++++++
 	I know. I just mean that may be you should implement this
feature as fetchall_arrayref(), because there is no
DBD::Oracle's implementation of fetchall_arrayref().


 		Bye. Alex.
0
alex
9/18/2007 6:16:35 PM
------=_Part_8249_5867375.1190143056325
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,

This patch speeds up both the regular fetch, as well as the
fetchall_arrayref(). To me that seems better than ONLY speeding up
fetchall_arrayref.

Biswa



On 9/18/07, Alexander V Alekseev <alex@alemate.ru> wrote:
>
>                 Hello!
>
> On Tue, 18 Sep 2007, Biswadeep Chowdhury wrote:
>
> > "fetchall_arrayref" is just as slow for REF CURSORs because internally,
> OCI
> > still does a DB roundtrip for each row.
> >
> > Running a test will illustrate the issue. Here is a result of a test
> run:
> > ++++++++++++++++++++++++++++++++++++++++++++++++++
> > Fetched 2500 rows (select): [0.06174] secs.
> > Fetched 2500 rows (ref cursor): [5.40367] secs.
> > Fetched 2500 rows (ref cursor array-fetch): [0.631323] secs.
> > ++++++++++++++++++++++++++++++++++++++++++++++++++
>         I know. I just mean that may be you should implement this
> feature as fetchall_arrayref(), because there is no
> DBD::Oracle's implementation of fetchall_arrayref().
>
>
>                 Bye. Alex.
>

------=_Part_8249_5867375.1190143056325--
0
cbiswa
9/18/2007 7:17:36 PM
 		Hello!

On Wed, 19 Sep 2007, Biswadeep Chowdhury wrote:

> This patch speeds up both the regular fetch, as well as the
> fetchall_arrayref(). To me that seems better than ONLY speeding up
> fetchall_arrayref.
 	If application requires fast response (as typical user
interface does), you may slow down the process. If you fetch from
complex query, you may get the first row immediately, and the next
row may require a few seconds to calculate. So, if you fetch by 1
row, you receive immediate response. If you fetch by 3 rows, you have to
wait for database to calculate 3 rows, before you get the response.

 	As I understand DBD interface, there is fetchrow_... interface
for 1-row fetch, and fetchall_arrayref() for multi-rows fetch.
Am I right?
 	Data caching is a good idea, but it should be tuned
per-statement. Though I don't know the way to do it, as prepare()
doesn't accept any additional attributes.

 	Bye. Alex.

0
alex
9/19/2007 10:26:12 AM
------=_Part_10449_23633015.1190215646816
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,

There are several things that I would like to point out:

1. Look no further than what happens for SELECTs. Data caching is turned on
by default - and I have not heard much complaining about it. Perl is seldom
used for user interface development and the scenario you describe is an
exception rather than the norm. If turning on data caching by default was a
good decision for SELECTs, then there is little reason not to turn on
caching by default for REF CURSORs.

2. Nevertheless, in the patch I have kept data caching turned off by default
for REF CURSORs so that users who are currently happy fetching a single row
per DB roundtrip can remain in that state. Only if users set the caching
parameter, multiple rows will be fetched per DB roundtrip.

3. Per-statement cache tuning may be done by setting the RefCursorCacheSize
to an appropriate before executing/fetching from a REF CURSOR. This is just
as you would do for RowCacheSize - which controls rows fetched per DB
roundtrip for SELECTs (Set this to explicitly 1 and fetch a significant
number of rows and you will realize that data caching is turned on by
default for SELECT statement results).

The reason I recommend this patch is that there is currently no way (that I
am aware of) to improve performance of fetch from REF CURSORs in perl. I
dont mind it being turned off by default - but at least users who want the
efficiency will have an option to do so.

Biswa


On 9/19/07, Alexander V Alekseev <alex@alemate.ru> wrote:
>
>                 Hello!
>
> On Wed, 19 Sep 2007, Biswadeep Chowdhury wrote:
>
> > This patch speeds up both the regular fetch, as well as the
> > fetchall_arrayref(). To me that seems better than ONLY speeding up
> > fetchall_arrayref.
>         If application requires fast response (as typical user
> interface does), you may slow down the process. If you fetch from
> complex query, you may get the first row immediately, and the next
> row may require a few seconds to calculate. So, if you fetch by 1
> row, you receive immediate response. If you fetch by 3 rows, you have to
> wait for database to calculate 3 rows, before you get the response.
>
>         As I understand DBD interface, there is fetchrow_... interface
> for 1-row fetch, and fetchall_arrayref() for multi-rows fetch.
> Am I right?
>         Data caching is a good idea, but it should be tuned
> per-statement. Though I don't know the way to do it, as prepare()
> doesn't accept any additional attributes.
>
>         Bye. Alex.
>
>

------=_Part_10449_23633015.1190215646816--
0
cbiswa
9/19/2007 3:27:26 PM
On Wed, Sep 19, 2007 at 08:57:26PM +0530, Biswadeep Chowdhury wrote:
> Hello,
> 
> There are several things that I would like to point out:
> 
> 1. Look no further than what happens for SELECTs. Data caching is turned on
> by default - and I have not heard much complaining about it. Perl is seldom
> used for user interface development and the scenario you describe is an
> exception rather than the norm. If turning on data caching by default was a
> good decision for SELECTs, then there is little reason not to turn on
> caching by default for REF CURSORs.

Agreed. The fact it doesn't work now is purely a bug/limitation of the
Oracle OCI interface.

It would be good if someone with Oracle support would search for
(and/or submit) a bug report about this. Perhaps there is some magic
OCI API incantation that would make row caching for ref cursors work.

I believe there must be, otherwise a great many oracle applications
would be suffering the same slowdown.

Any volunteers to check Oracle support about this issue?

> 2. Nevertheless, in the patch I have kept data caching turned off by default
> for REF CURSORs so that users who are currently happy fetching a single row
> per DB roundtrip can remain in that state. Only if users set the caching
> parameter, multiple rows will be fetched per DB roundtrip.

It should be enabled by default. Lack of caching is a bug.

> 3. Per-statement cache tuning may be done by setting the RefCursorCacheSize
> to an appropriate before executing/fetching from a REF CURSOR. This is just
> as you would do for RowCacheSize - which controls rows fetched per DB
> roundtrip for SELECTs (Set this to explicitly 1 and fetch a significant
> number of rows and you will realize that data caching is turned on by
> default for SELECT statement results).
> 
> The reason I recommend this patch is that there is currently no way (that I
> am aware of) to improve performance of fetch from REF CURSORs in perl. I
> dont mind it being turned off by default - but at least users who want the
> efficiency will have an option to do so.

I'd want it on by default, with a suitable default cache size (ideally
calculated the same way the default row cache is sized), but I'd rather
not have to apply the patch at all.

Before applying a large workaround we should be quite sure there isn't
a simple fix via the OCI API.

Tim.

> Biswa
> 
> 
> On 9/19/07, Alexander V Alekseev <alex@alemate.ru> wrote:
> >
> >                 Hello!
> >
> > On Wed, 19 Sep 2007, Biswadeep Chowdhury wrote:
> >
> > > This patch speeds up both the regular fetch, as well as the
> > > fetchall_arrayref(). To me that seems better than ONLY speeding up
> > > fetchall_arrayref.
> >         If application requires fast response (as typical user
> > interface does), you may slow down the process. If you fetch from
> > complex query, you may get the first row immediately, and the next
> > row may require a few seconds to calculate. So, if you fetch by 1
> > row, you receive immediate response. If you fetch by 3 rows, you have to
> > wait for database to calculate 3 rows, before you get the response.
> >
> >         As I understand DBD interface, there is fetchrow_... interface
> > for 1-row fetch, and fetchall_arrayref() for multi-rows fetch.
> > Am I right?
> >         Data caching is a good idea, but it should be tuned
> > per-statement. Though I don't know the way to do it, as prepare()
> > doesn't accept any additional attributes.
> >
> >         Bye. Alex.
> >
> >
0
Tim
9/20/2007 8:22:00 AM
------=_Part_14966_9652108.1190404370982
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello Tim,

I agree the bug is really in OCI and ideally, that is where it should be
fixed. I really hope that there is some "switch" that can fix the issue -
but I have not been able to find any to date. I have tried this using Pro*C
too (using the PREFETCH=n commandline option) but the result is the same -
does not work for ref cursors.

People using Pro*C or OCI do however have the option to do array-fetch -
that is call OCIStmtFetch2 to fetch more than one row at a time - although
the pain of managing the data structures is now upon the user. It is not as
transparent as setting PREFETCH=n - but still, an option exists for them to
speed things up for REF CURSORs. This option is however not accessible to
perl users.

As you say, it would be better to turn some caching on by default - even if
a conservative amount of rows.

I have been looking for a fix for more than a year and have found none. I
see the same issue affecting nested cursors.  In the patch, I have addressed
REF CURSORs only, but I expect we could apply the same fix for nested
cursors. (You may run t/55nested.t independently to see degraded performance
of fetch from nested cursors).

Thanks.
Biswa




On 9/20/07, Tim Bunce <Tim.Bunce@pobox.com> wrote:
>
> On Wed, Sep 19, 2007 at 08:57:26PM +0530, Biswadeep Chowdhury wrote:
> > Hello,
> >
> > There are several things that I would like to point out:
> >
> > 1. Look no further than what happens for SELECTs. Data caching is turned
> on
> > by default - and I have not heard much complaining about it. Perl is
> seldom
> > used for user interface development and the scenario you describe is an
> > exception rather than the norm. If turning on data caching by default
> was a
> > good decision for SELECTs, then there is little reason not to turn on
> > caching by default for REF CURSORs.
>
> Agreed. The fact it doesn't work now is purely a bug/limitation of the
> Oracle OCI interface.
>
> It would be good if someone with Oracle support would search for
> (and/or submit) a bug report about this. Perhaps there is some magic
> OCI API incantation that would make row caching for ref cursors work.
>
> I believe there must be, otherwise a great many oracle applications
> would be suffering the same slowdown.
>
> Any volunteers to check Oracle support about this issue?
>
> > 2. Nevertheless, in the patch I have kept data caching turned off by
> default
> > for REF CURSORs so that users who are currently happy fetching a single
> row
> > per DB roundtrip can remain in that state. Only if users set the caching
> > parameter, multiple rows will be fetched per DB roundtrip.
>
> It should be enabled by default. Lack of caching is a bug.
>
> > 3. Per-statement cache tuning may be done by setting the
> RefCursorCacheSize
> > to an appropriate before executing/fetching from a REF CURSOR. This is
> just
> > as you would do for RowCacheSize - which controls rows fetched per DB
> > roundtrip for SELECTs (Set this to explicitly 1 and fetch a significant
> > number of rows and you will realize that data caching is turned on by
> > default for SELECT statement results).
> >
> > The reason I recommend this patch is that there is currently no way
> (that I
> > am aware of) to improve performance of fetch from REF CURSORs in perl. I
>
> > dont mind it being turned off by default - but at least users who want
> the
> > efficiency will have an option to do so.
>
> I'd want it on by default, with a suitable default cache size (ideally
> calculated the same way the default row cache is sized), but I'd rather
> not have to apply the patch at all.
>
> Before applying a large workaround we should be quite sure there isn't
> a simple fix via the OCI API.
>
> Tim.
>
> > Biswa
> >
> >
> > On 9/19/07, Alexander V Alekseev < alex@alemate.ru> wrote:
> > >
> > >                 Hello!
> > >
> > > On Wed, 19 Sep 2007, Biswadeep Chowdhury wrote:
> > >
> > > > This patch speeds up both the regular fetch, as well as the
> > > > fetchall_arrayref(). To me that seems better than ONLY speeding up
> > > > fetchall_arrayref.
> > >         If application requires fast response (as typical user
> > > interface does), you may slow down the process. If you fetch from
> > > complex query, you may get the first row immediately, and the next
> > > row may require a few seconds to calculate. So, if you fetch by 1
> > > row, you receive immediate response. If you fetch by 3 rows, you have
> to
> > > wait for database to calculate 3 rows, before you get the response.
> > >
> > >         As I understand DBD interface, there is fetchrow_... interface
> > > for 1-row fetch, and fetchall_arrayref() for multi-rows fetch.
> > > Am I right?
> > >         Data caching is a good idea, but it should be tuned
> > > per-statement. Though I don't know the way to do it, as prepare()
> > > doesn't accept any additional attributes.
> > >
> > >         Bye. Alex.
> > >
> > >
>

------=_Part_14966_9652108.1190404370982--
0
cbiswa
9/21/2007 7:52:50 PM
On Sat, Sep 22, 2007 at 01:22:50AM +0530, Biswadeep Chowdhury wrote:
> 
>    I have been looking for a fix for more than a year and have found none.

Have you asked Oracle?

Tim.
0
Tim
9/21/2007 8:03:50 PM
------=_Part_15187_5905353.1190457138992
Content-Type: multipart/alternative; 
	boundary="----=_Part_15188_24596908.1190457138992"

------=_Part_15188_24596908.1190457138992
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,

I had contacted Oracle through our DBA more than a year ago.
At that time, they had pointed me to bug that was fixed (attached). However,
that bug dealt with explicit-array-fetch and not the "prefetch" setting that
is being used in DBD.

Also if we take some of the Oracle documentation very literally:

The Pro*C documentation
athttp://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711
mentions "...The number of rows set by the PREFETCH option value is used for
all queries involving explicit cursors,..."

The OCI documentation at
http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254
mentions "...the OCI can prefetch result set rows when executing a query..."

Then it looks like the prefetch setting may not be intended to work for REF
CURSORs or nested cursors - although that is hardly reasonable.

I will see if I can contact Oracle once more in the near future.

Biswa


On 9/22/07, Tim Bunce <Tim.Bunce@pobox.com> wrote:
>
> On Sat, Sep 22, 2007 at 01:22:50AM +0530, Biswadeep Chowdhury wrote:
> >
> >    I have been looking for a fix for more than a year and have found
> none.
>
> Have you asked Oracle?
>
> Tim.
>

------=_Part_15188_24596908.1190457138992
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,<br>
<br>I had contacted Oracle through our DBA more than a year ago.<br>At that
time, they had pointed me to bug that was fixed (attached). However,
that bug dealt with explicit-array-fetch and not the &quot;prefetch&quot; setting
that is being used in DBD.<br><br>
Also if we take some of the Oracle documentation very literally:<br>
<br>The Pro*C documentation at<a href="http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)"> http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711
</a><br>mentions &quot;...The number of rows set by the PREFETCH option value is <span style="font-weight: bold;">used for all queries involving explicit cursors</span>,...&quot;<br>
<br>The OCI documentation at <a href="http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254
</a>&nbsp;
mentions &quot;...the OCI can prefetch result set rows <span style="font-weight: bold;">when executing a query</span>...&quot;<br><br>Then
it looks like the prefetch setting may not be intended to work for REF
CURSORs or nested cursors - although that is hardly reasonable. <br>
<br>
I will see if I can contact Oracle once more in the near future.<br>
<br>
Biswa<br>
<br><br><div><span class="gmail_quote">On 9/22/07, <b class="gmail_sendername">Tim Bunce</b> &lt;<a href="mailto:Tim.Bunce@pobox.com">Tim.Bunce@pobox.com</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On Sat, Sep 22, 2007 at 01:22:50AM +0530, Biswadeep Chowdhury wrote:<br>&gt;<br>&gt;&nbsp;&nbsp;&nbsp;&nbsp;I have been looking for a fix for more than a year and have found none.<br><br>Have you asked Oracle?<br><br>Tim.<br></blockquote></div>
<br>

------=_Part_15188_24596908.1190457138992--

------=_Part_15187_5905353.1190457138992
Content-Type: application/pdf; name="bug.pdf"
Content-Transfer-Encoding: base64
Content-Disposition: attachment; filename="bug.pdf"
X-Attachment-Id: f_f6vzb11q
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------=_Part_15187_5905353.1190457138992--
0
cbiswa
9/22/2007 10:32:18 AM
------=_Part_1504_13733409.1190997744527
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello Tim,

I have modified the patch taking into account your comments
-> Array fetch has been enabled for "nested cursors" too, in addition to
"ref cursors".
-> Array fetch is enabled by default (not for regular selects ofcourse -
which still uses OCI prefetch setting)
-> Array size is calculated based on fields (although I have restricted to a
max of 128 rows for now)
-> There is no separate parameter to manipulate array size
(RefCursorCacheSize), RowCacheSize itself is used.

I am listing the patch diff file below (shorter now).
Unless someone can point out that elusive OCI setting, I request you to
consider this.

Thank You.
Biswa

============================== patch file start
==============================
diff -r -u DBD-Oracle-1.19/dbdimp.c DBD-Oracle-1.19.refcursor/dbdimp.c
--- DBD-Oracle-1.19/dbdimp.c    2006-11-03 06:05:46.000000000 -0800
+++ DBD-Oracle-1.19.refcursor/dbdimp.c    2007-09-28 06:54:17.000000000-0700
@@ -1298,6 +1298,13 @@

     /* set ACTIVE so dbd_describe doesn't do explicit OCI describe */
     DBIc_ACTIVE_on(imp_sth_csr);
+/**********************************************************************
+*  rs_array: mark array fetch on for ref cursor - START
+**********************************************************************/
+        imp_sth_csr->rs_array_on=1;
+/**********************************************************************
+*  rs_array: mark array fetch on for ref cursor - END
+**********************************************************************/
     if (!dbd_describe(sth_csr, imp_sth_csr)) {
         return 0;
     }
diff -r -u DBD-Oracle-1.19/dbdimp.h DBD-Oracle-1.19.refcursor/dbdimp.h
--- DBD-Oracle-1.19/dbdimp.h    2006-10-27 08:50:52.000000000 -0700
+++ DBD-Oracle-1.19.refcursor/dbdimp.h    2007-09-28 06:49:33.000000000-0700
@@ -134,6 +134,17 @@

     /* (In/)Out Parameter Details */
     bool  has_inout_params;
+/************************************************************
+* rs_array: state variables - START
+************************************************************/
+    bool      rs_array_on;           /* if array to be used */
+    int       rs_array_size;         /* array size */
+    int       rs_array_num_rows;     /* num rows in last fetch */
+    int       rs_array_idx;          /* index of current row */
+    sword     rs_array_status;       /* status of last fetch */
+/************************************************************
+* rs_array: state variables - END
+************************************************************/
 };
 #define IMP_STH_EXECUTING    0x0001

Only in DBD-Oracle-1.19.refcursor: Makefile.old
diff -r -u DBD-Oracle-1.19/oci8.c DBD-Oracle-1.19.refcursor/oci8.c
--- DBD-Oracle-1.19/oci8.c    2006-10-13 05:12: 20.000000000 -0700
+++ DBD-Oracle-1.19.refcursor/oci8.c    2007-09-28 08:57:14.000000000 -0700
@@ -1294,6 +1294,23 @@
         oci_error(h, imp_sth->errhp, status, "OCIAttrSet
OCI_ATTR_PREFETCH_ROWS");
         ++num_errors;
     }
+/**********************************************************************
+* rs_array: initialize cache - START
+**********************************************************************/
+        if (imp_sth->nested_cursor)        /* array fetch for nested
cursors */
+          imp_sth->rs_array_on=1;
+        if (imp_sth->rs_array_on && cache_rows>0)
+          imp_sth->rs_array_size=cache_rows>128?128:cache_rows;    /*
restrict to 128 for now */
+        else
+          imp_sth->rs_array_size=1;
+        imp_sth->rs_array_num_rows=0;
+        imp_sth->rs_array_idx=0;
+        imp_sth->rs_array_status=OCI_SUCCESS;
+        if (DBIS->debug >= 3)
+          PerlIO_printf(DBILOGFP, "    sth_set_row_cache: rs_array_size set
to %d\n",imp_sth->rs_array_size);
+/**********************************************************************
+* rs_array: initialize cache - END
+**********************************************************************/
     }
     else {                /* set cache size by memory    */
     cache_mem  = -imp_sth->cache_rows; /* cache_mem always +ve here */
@@ -1579,7 +1596,13 @@
     sb4 define_len = (ftype==94||ftype==95) ? fbh->disize+4 : fbh->disize;
     fb_ary_t  *fb_ary;

-    fbh->fb_ary = fb_ary_alloc(define_len, 1);
+/**********************************************************************
+* rs_array: allocate field buffer array - START
+**********************************************************************/
+    fbh->fb_ary = fb_ary_alloc(define_len, imp_sth->rs_array_size);
+/**********************************************************************
+* rs_array: allocate field buffer array - END
+**********************************************************************/
     fb_ary = fbh->fb_ary;

     if (fbh->ftype == 116) { /* RSET */
@@ -1663,8 +1686,26 @@
     else {
     if (DBIS->debug >= 3)
         PerlIO_printf(DBILOGFP, "    dbd_st_fetch %d fields...\n",
DBIc_NUM_FIELDS(imp_sth));
-    OCIStmtFetch_log_stat(imp_sth->stmhp, imp_sth->errhp,
-        1, (ub2)OCI_FETCH_NEXT, OCI_DEFAULT, status);
+/**********************************************************************
+*  rs_array: iteration - START
+**********************************************************************/
+    imp_sth->rs_array_idx++;
+    if (imp_sth->rs_array_status==OCI_SUCCESS &&
imp_sth->rs_array_num_rows<=imp_sth->rs_array_idx)
+    {
+       OCIStmtFetch_log_stat(imp_sth->stmhp, imp_sth->errhp,
+        imp_sth->rs_array_size, (ub2)OCI_FETCH_NEXT, OCI_DEFAULT, status);
+       imp_sth->rs_array_status=status;
+       OCIAttrGet_stmhp_stat(imp_sth, &imp_sth->rs_array_num_rows,0,
+                OCI_ATTR_ROWS_FETCHED, status);
+       imp_sth->rs_array_idx=0;
+    }
+    if (imp_sth->rs_array_num_rows>imp_sth->rs_array_idx)
+      status=OCI_SUCCESS;
+    else
+      status=imp_sth->rs_array_status;
+/**********************************************************************
+*  rs_array: iteration - END
+**********************************************************************/
     }

     if (status != OCI_SUCCESS) {
@@ -1698,7 +1739,15 @@
     for(i=0; i < num_fields; ++i) {
     imp_fbh_t *fbh = &imp_sth->fbh[i];
     fb_ary_t *fb_ary = fbh->fb_ary;
-    int rc = fb_ary->arcode[0];
+/**********************************************************************
+*  rs_array: offset values - START
+**********************************************************************/
+        int row_idx=imp_sth->rs_array_idx;
+    int rc = fb_ary->arcode[row_idx];
+        ub1* row_data=&fb_ary->abuf[0]+(fb_ary->bufl*row_idx);
+/**********************************************************************
+*  rs_array: offset values - END
+**********************************************************************/
     SV *sv = AvARRAY(av)[i]; /* Note: we (re)use the SV in the AV    */

     if (rc == 1406                /* field was truncated    */
@@ -1724,8 +1773,14 @@
             ++err;    /* fetch_func already called oci_error */
         }
         else {
-        int datalen = fb_ary->arlen[0];
-        char *p = (char*)&fb_ary->abuf[0];
+/**********************************************************************
+*  rs_array: use offset data - START
+**********************************************************************/
+        int datalen = fb_ary->arlen[row_idx];
+        char *p = (char*)row_data;
+/**********************************************************************
+*  rs_array: use offset data - END
+**********************************************************************/
         /* if ChopBlanks check for Oracle CHAR type (blank padded)    */
         if (ChopBlanks && fbh->dbtype == 96) {
             while(datalen && p[datalen - 1]==' ')
@@ -1747,8 +1802,13 @@
         if (!fbh->fetch_func) {
             /* Copy the truncated value anyway, it may be of use,    */
             /* but it'll only be accessible via prior bind_column()    */
-            sv_setpvn(sv, (char*)&fb_ary->abuf[0],
-              fb_ary->arlen[0]);
+/**********************************************************************
+*  rs_array: use offset data/len - START
+**********************************************************************/
+            sv_setpvn(sv, (char*)row_data,fb_ary->arlen[row_idx]);
+/**********************************************************************
+*  rs_array: use offset data/len - END
+**********************************************************************/
             if (CSFORM_IMPLIES_UTF8(fbh->csform))
             SvUTF8_on(sv);
         }
diff -r -u DBD-Oracle-1.19/t/50cursor.t DBD-Oracle-1.19.refcursor
/t/50cursor.t
--- DBD-Oracle-1.19/t/50cursor.t    2006-07-07 05:17: 29.000000000 -0700
+++ DBD-Oracle-1.19.refcursor/t/50cursor.t    2007-09-28 08:41:17.000000000-0700
@@ -96,6 +96,49 @@
     ok(0, $close_cursor->execute);
 }

+##################################################
+# Row fetch timing: START
+##################################################
+use Time::HiRes qw(gettimeofday tv_interval);
+
+sub timed_fetch {
+  my($rs,$caption) = @_;
+  my $tm_start=[gettimeofday];
+  my $rows = $rs->fetchall_arrayref();
+  my $elapsed=tv_interval($tm_start);
+  print "Fetched ".(@{$rows})." rows ($caption): [$elapsed] secs.\n";
+}
+
+##################################################
+# execute (select)
+##################################################
+my $sth = $dbh->prepare(
+ 'SELECT object_name FROM (SELECT object_name '
+  .'FROM all_objects WHERE ROWNUM<=70),(SELECT 1 FROM all_objects '
+  .'WHERE ROWNUM<=70)'
+);
+$sth->execute();
+timed_fetch($sth,'select');
+
+##################################################
+# execute (refcursor default-cache)
+##################################################
+$sth = $dbh->prepare(
+ 'BEGIN OPEN ? FOR '
+  .'SELECT object_name FROM (SELECT object_name '
+  .'FROM all_objects WHERE ROWNUM<=70),(SELECT 1 FROM all_objects '
+  .'WHERE ROWNUM<=70); '
+  .'END;'
+);
+my $rs;
+$sth->bind_param_inout(1,\$rs,0,{ ora_type => ORA_RSET });
+$sth->execute();
+timed_fetch($rs,'refcursor default-cache');
+
+##################################################
+# Row fetch timing: END
+##################################################
+
 $dbh->disconnect;

 exit 0;
============================== patch file end ==============================





On 9/20/07, Tim Bunce <Tim.Bunce@pobox.com> wrote:
>
> On Wed, Sep 19, 2007 at 08:57:26PM +0530, Biswadeep Chowdhury wrote:
> > Hello,
> >
> > There are several things that I would like to point out:
> >
> > 1. Look no further than what happens for SELECTs. Data caching is turned
> on
> > by default - and I have not heard much complaining about it. Perl is
> seldom
> > used for user interface development and the scenario you describe is an
> > exception rather than the norm. If turning on data caching by default
> was a
> > good decision for SELECTs, then there is little reason not to turn on
> > caching by default for REF CURSORs.
>
> Agreed. The fact it doesn't work now is purely a bug/limitation of the
> Oracle OCI interface.
>
> It would be good if someone with Oracle support would search for
> (and/or submit) a bug report about this. Perhaps there is some magic
> OCI API incantation that would make row caching for ref cursors work.
>
> I believe there must be, otherwise a great many oracle applications
> would be suffering the same slowdown.
>
> Any volunteers to check Oracle support about this issue?
>
> > 2. Nevertheless, in the patch I have kept data caching turned off by
> default
> > for REF CURSORs so that users who are currently happy fetching a single
> row
> > per DB roundtrip can remain in that state. Only if users set the caching
> > parameter, multiple rows will be fetched per DB roundtrip.
>
> It should be enabled by default. Lack of caching is a bug.
>
> > 3. Per-statement cache tuning may be done by setting the
> RefCursorCacheSize
> > to an appropriate before executing/fetching from a REF CURSOR. This is
> just
> > as you would do for RowCacheSize - which controls rows fetched per DB
> > roundtrip for SELECTs (Set this to explicitly 1 and fetch a significant
> > number of rows and you will realize that data caching is turned on by
> > default for SELECT statement results).
> >
> > The reason I recommend this patch is that there is currently no way
> (that I
> > am aware of) to improve performance of fetch from REF CURSORs in perl. I
> > dont mind it being turned off by default - but at least users who want
> the
> > efficiency will have an option to do so.
>
> I'd want it on by default, with a suitable default cache size (ideally
> calculated the same way the default row cache is sized), but I'd rather
> not have to apply the patch at all.
>
> Before applying a large workaround we should be quite sure there isn't
> a simple fix via the OCI API.
>
> Tim.
>
> > Biswa
>

------=_Part_1504_13733409.1190997744527--
0
cbiswa
9/28/2007 4:42:24 PM
------=_Part_17896_16081344.1191338768521
Content-Type: multipart/alternative; 
	boundary="----=_Part_17897_14626695.1191338768521"

------=_Part_17897_14626695.1191338768521
Content-Type: text/plain; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,

I found the following information on metalink.oracle.com that mentions that
there was no solution as of March 2006.

Thanks.
Biswa


  Subject:

*OCCI Prefetch Does Not Work With REF CURSOR*



Doc ID <https://metalink.oracle.com/help/usaeng/Search/search.html#file>:

*Note:293030.1*

Type:

*PROBLEM*



Last Revision Date:

*28-MAR-2006*

Status:

*PUBLISHED*

*In this Document*
  Symptoms<https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#SYMPTOM>
  Cause<https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#CAUSE>
  Solution<https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#FIX>
  References<https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#REF>
 ------------------------------
 Applies to:

Oracle Server - Enterprise Edition - Version: 9.2.0 to 10.2.0
PL/SQL - Version: 9.2.0 to 10.2.0
This problem can occur on any platform.
Symptoms

The OCCI prefetch doesn't work invoking stored procedure with REF CURSOR.

The OCCI programs that invoke stored procedures with REF CURSOR are very
slow.
Cause

This is currently expected behavior, identified in
Bug 3067376<https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&id=3067376>PERFORMANCE
ISSUE WITH OCCI CODE
in which is reported that OCCI PREFETCH doesn't work with REF cursor

However there have been raised an enhancement request to be implemented in
the future releases for this in
Bug:3140718 OCCI PREFETCH DOES NOT WORK WITH REF CURSOR
Solution

There is no solution available at the moment when writing this article
(March 2006) to use PREFETCH with REF CURSORS.
References

Bug 3067376<https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&id=3067376>-
Performance Issue With Occi Code
Bug 3140718<https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&id=3140718>-
Occi Prefetch Does Not Work With Ref Cursor


On 9/22/07, Biswadeep Chowdhury <cbiswa@gmail.com> wrote:
>
> Hello,
>
> I had contacted Oracle through our DBA more than a year ago.
> At that time, they had pointed me to bug that was fixed (attached).
> However, that bug dealt with explicit-array-fetch and not the "prefetch"
> setting that is being used in DBD.
>
> Also if we take some of the Oracle documentation very literally:
>
> The Pro*C documentation athttp://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711
>
> mentions "...The number of rows set by the PREFETCH option value is used
> for all queries involving explicit cursors,..."
>
> The OCI documentation at http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254
>   mentions "...the OCI can prefetch result set rows when executing a query
> ..."
>
> Then it looks like the prefetch setting may not be intended to work for
> REF CURSORs or nested cursors - although that is hardly reasonable.
>
> I will see if I can contact Oracle once more in the near future.
>
> Biswa
>
>
> On 9/22/07, Tim Bunce <Tim.Bunce@pobox.com> wrote:
> >
> > On Sat, Sep 22, 2007 at 01:22:50AM +0530, Biswadeep Chowdhury wrote:
> > >
> > >    I have been looking for a fix for more than a year and have found
> > none.
> >
> > Have you asked Oracle?
> >
> > Tim.
> >
>
>
>

------=_Part_17897_14626695.1191338768521
Content-Type: text/html; charset=ISO-8859-1
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

Hello,<br><br>I found the following information on <a href="http://metalink.oracle.com">metalink.oracle.com</a> that mentions that there was no solution as of March 2006.<br><br>Thanks.<br>Biswa<br><br><br>

<table class="MsoNormalTable" style="width: 80%;" border="0" cellpadding="0" cellspacing="6" width="80%">
 <tbody><tr style="">
  <td style="padding: 0in;" valign="top">
  <p class="MsoNormal" style="text-align: right;" align="right">Subject: </p>
  </td>
  <td colspan="4" style="padding: 0in;">
  <p class="MsoNormal"><strong>OCCI Prefetch Does Not Work With REF CURSOR</strong></p>
  </td>
 </tr>
 <tr style="">
  <td style="padding: 0in;">
  <p class="MsoNormal">&nbsp;</p>
  </td>
  <td style="padding: 0in;" valign="top">
  <p class="MsoNormal" style="text-align: right;" align="right"><a href="https://metalink.oracle.com/help/usaeng/Search/search.html#file">Doc ID</a>:
  </p>
  </td>
  <td style="padding: 0in;">
  <p class="MsoNormal"><strong>Note:293030.1</strong></p>
  </td>
  <td style="padding: 0in;" valign="top">
  <p class="MsoNormal" style="text-align: right;" align="right">Type: </p>
  </td>
  <td style="padding: 0in;">
  <p class="MsoNormal"><strong>PROBLEM</strong></p>
  </td>
 </tr>
 <tr style="">
  <td style="padding: 0in;">
  <p class="MsoNormal">&nbsp;</p>
  </td>
  <td style="padding: 0in;" valign="top">
  <p class="MsoNormal" style="text-align: right;" align="right">Last Revision Date: </p>
  </td>
  <td style="padding: 0in;">
  <p class="MsoNormal"><strong>28-MAR-2006</strong></p>
  </td>
  <td style="padding: 0in;" valign="top">
  <p class="MsoNormal" style="text-align: right;" align="right">Status: </p>
  </td>
  <td style="padding: 0in;">
  <p class="MsoNormal"><strong>PUBLISHED</strong></p>
  </td>
 </tr>
</tbody></table>

<p><span style="font-family: Helvetica;"></span><b>In this Document</b><br>
<span style="font-size: 10pt;">&nbsp;&nbsp;<a href="https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#SYMPTOM">
Symptoms</a><br>
&nbsp;&nbsp;<a href="https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#CAUSE">Cause</a><br>
&nbsp;&nbsp;<a href="https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#FIX">Solution</a><br>
&nbsp;&nbsp;<a href="https://metalink.oracle.com/metalink/plsql/f?p=130:14:9270921068823151778::::p14_database_id,p14_docid,p14_show_header,p14_show_help,p14_black_frame,p14_font:NOT,293030.1,1,1,1,helvetica#REF">References</a></span>
</p>

<div class="MsoNormal" style="text-align: center;" align="center">

<hr align="center" size="1" width="100%">

</div>

<h2>Applies to: </h2>

<p class="MsoNormal">Oracle Server - Enterprise Edition - Version: 9.2.0 to
10.2.0<br>
PL/SQL - Version: 9.2.0 to 10.2.0<br>
This problem can occur on any platform.</p>

<h2><a name="SYMPTOM"></a>Symptoms</h2>

<p class="MsoNormal">The OCCI prefetch doesn&#39;t work invoking stored procedure with
REF CURSOR.<br>
<br>
The OCCI programs that invoke stored procedures with REF CURSOR are very slow.</p>

<h2><a name="CAUSE"></a>Cause</h2>

<p class="MsoNormal" style="margin-bottom: 12pt;">This is currently expected
behavior, identified in <br>
<a href="https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&amp;id=3067376">Bug
3067376</a> PERFORMANCE ISSUE WITH OCCI CODE<br>
in which is reported that OCCI PREFETCH doesn&#39;t work with REF cursor<br>
<br>
However there have been raised an enhancement request to be implemented in the
future releases for this in<br>
Bug:3140718 OCCI PREFETCH DOES NOT WORK WITH REF CURSOR</p>

<h2><a name="FIX"></a>Solution</h2>

<p class="MsoNormal">There is no solution available at the moment when writing
this article (March 2006) to use PREFETCH with REF CURSORS.</p>

<h2><a name="REF"></a>References</h2>

<p class="MsoNormal"><a href="https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&amp;id=3067376">Bug
3067376</a> - Performance Issue With Occi Code<br>
<a href="https://metalink.oracle.com/metalink/plsql/showdoc?db=Bug&amp;id=3140718">Bug
3140718</a> - Occi Prefetch Does Not Work With Ref Cursor</p>

<br><br><div><span class="gmail_quote">On 9/22/07, <b class="gmail_sendername">Biswadeep Chowdhury</b> &lt;<a href="mailto:cbiswa@gmail.com">cbiswa@gmail.com</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
Hello,<br>
<br>I had contacted Oracle through our DBA more than a year ago.<br>At that
time, they had pointed me to bug that was fixed (attached). However,
that bug dealt with explicit-array-fetch and not the &quot;prefetch&quot; setting
that is being used in DBD.<br><br>
Also if we take some of the Oracle documentation very literally:<br>
<br>The Pro*C documentation at<a href="http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)"> http://download.oracle.com/docs/cd/B19306_01/appdev.102/b14407/pc_06sql.htm#i4711
</a><br>mentions &quot;...The number of rows set by the PREFETCH option value is <span style="font-weight: bold;">used for all queries involving explicit cursors</span>,...&quot;<br>
<br>The OCI documentation at <a href="http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">http://download-uk.oracle.com/docs/cd/B14117_01/appdev.101/b10779/oci04sql.htm#428254
</a>&nbsp;
mentions &quot;...the OCI can prefetch result set rows <span style="font-weight: bold;">when executing a query</span>...&quot;<br><br>Then
it looks like the prefetch setting may not be intended to work for REF
CURSORs or nested cursors - although that is hardly reasonable. <br>
<br>
I will see if I can contact Oracle once more in the near future.<br>
<br>
Biswa<div><span class="e" id="q_1152cc65f94785e0_1"><br>
<br><br><div><span class="gmail_quote">On 9/22/07, <b class="gmail_sendername">Tim Bunce</b> &lt;<a href="mailto:Tim.Bunce@pobox.com" target="_blank" onclick="return top.js.OpenExtLink(window,event,this)">Tim.Bunce@pobox.com
</a>&gt; wrote:</span><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
On Sat, Sep 22, 2007 at 01:22:50AM +0530, Biswadeep Chowdhury wrote:<br>&gt;<br>&gt;&nbsp;&nbsp;&nbsp;&nbsp;I have been looking for a fix for more than a year and have found none.<br><br>Have you asked Oracle?<br><br>Tim.<br></blockquote></div>

<br>
</span></div><br clear="all"></blockquote></div><br>

------=_Part_17897_14626695.1191338768521--

------=_Part_17896_16081344.1191338768521--
0
cbiswa
10/2/2007 3:26:08 PM
On Fri, Sep 28, 2007 at 12:42:24PM -0400, Biswadeep Chowdhury wrote:
> Hello Tim,
> 
> I have modified the patch taking into account your comments
> -> Array fetch has been enabled for "nested cursors" too, in addition to
> "ref cursors".
> -> Array fetch is enabled by default (not for regular selects ofcourse -
> which still uses OCI prefetch setting)
> -> Array size is calculated based on fields (although I have restricted to a
> max of 128 rows for now)
> -> There is no separate parameter to manipulate array size
> (RefCursorCacheSize), RowCacheSize itself is used.
> 
> I am listing the patch diff file below (shorter now).
> Unless someone can point out that elusive OCI setting, I request you to
> consider this.

John Scoles at pythian.com takes the lead for DBD::Oracle development now
(for which I am eternally grateful). I'm sure he'll be considering it carefully.

Thank you for your help - this is an important issue to resolve.

Tim.
0
Tim
10/2/2007 11:12:25 PM
Hello,

I hope a solution to the "prefetch issue for ref cursors and nested
cursors" will be considered in the next DBD-Oracle version - be it my
patch or an alternative fix.

To make it (seem) less daunting, I am attaching here a shorter version
of the patch diff file.
This diff file is about 100 lines shorter than the previous one I
sent. I have removed some (extra) comments and all changes made to
50cursor.t. Since this patch works for nested cursors, the performance
benefit can be tested can be tested using the existing test script
55nested.t

Thank You.
Biswadeep

============================== patch diff file start
==============================
diff -r -u DBD-Oracle-1.19/dbdimp.c DBD-Oracle-1.19.refcursor.min/dbdimp.c
--- DBD-Oracle-1.19/dbdimp.c	2006-11-03 06:05:46.000000000 -0800
+++ DBD-Oracle-1.19.refcursor.min/dbdimp.c	2007-11-06 13:18:50.000000000 -0800
@@ -1298,6 +1298,7 @@

 	/* set ACTIVE so dbd_describe doesn't do explicit OCI describe */
 	DBIc_ACTIVE_on(imp_sth_csr);
+        imp_sth_csr->rs_array_on=1;
 	if (!dbd_describe(sth_csr, imp_sth_csr)) {
 	    return 0;
 	}
diff -r -u DBD-Oracle-1.19/dbdimp.h DBD-Oracle-1.19.refcursor.min/dbdimp.h
--- DBD-Oracle-1.19/dbdimp.h	2006-10-27 08:50:52.000000000 -0700
+++ DBD-Oracle-1.19.refcursor.min/dbdimp.h	2007-11-06 13:30:21.000000000 -0800
@@ -134,6 +134,11 @@

     /* (In/)Out Parameter Details */
     bool  has_inout_params;
+    bool      rs_array_on;           /* if array to be used */
+    int       rs_array_size;         /* array size */
+    int       rs_array_num_rows;     /* num rows in last fetch */
+    int       rs_array_idx;          /* index of current row */
+    sword     rs_array_status;       /* status of last fetch */
 };
 #define IMP_STH_EXECUTING	0x0001

Only in DBD-Oracle-1.19.refcursor.min: Makefile.old
diff -r -u DBD-Oracle-1.19/oci8.c DBD-Oracle-1.19.refcursor.min/oci8.c
--- DBD-Oracle-1.19/oci8.c	2006-10-13 05:12:20.000000000 -0700
+++ DBD-Oracle-1.19.refcursor.min/oci8.c	2007-11-06 13:19:57.000000000 -0800
@@ -1294,6 +1294,17 @@
 	    oci_error(h, imp_sth->errhp, status, "OCIAttrSet OCI_ATTR_PREFETCH_ROWS");
 	    ++num_errors;
 	}
+        if (imp_sth->nested_cursor)		/* array fetch for nested cursors */
+          imp_sth->rs_array_on=1;
+        if (imp_sth->rs_array_on && cache_rows>0)
+          imp_sth->rs_array_size=cache_rows>128?128:cache_rows;	/*
restrict to 128 for now */
+        else
+          imp_sth->rs_array_size=1;
+        imp_sth->rs_array_num_rows=0;
+        imp_sth->rs_array_idx=0;
+        imp_sth->rs_array_status=OCI_SUCCESS;
+        if (DBIS->debug >= 3)
+          PerlIO_printf(DBILOGFP, "    sth_set_row_cache:
rs_array_size set to %d\n",imp_sth->rs_array_size);
     }
     else {				/* set cache size by memory	*/
 	cache_mem  = -imp_sth->cache_rows; /* cache_mem always +ve here */
@@ -1579,7 +1590,7 @@
 	sb4 define_len = (ftype==94||ftype==95) ? fbh->disize+4 : fbh->disize;
 	fb_ary_t  *fb_ary;

-	fbh->fb_ary = fb_ary_alloc(define_len, 1);
+	fbh->fb_ary = fb_ary_alloc(define_len, imp_sth->rs_array_size);
 	fb_ary = fbh->fb_ary;

 	if (fbh->ftype == 116) { /* RSET */
@@ -1663,8 +1674,20 @@
     else {
 	if (DBIS->debug >= 3)
 	    PerlIO_printf(DBILOGFP, "    dbd_st_fetch %d fields...\n",
DBIc_NUM_FIELDS(imp_sth));
-	OCIStmtFetch_log_stat(imp_sth->stmhp, imp_sth->errhp,
-		1, (ub2)OCI_FETCH_NEXT, OCI_DEFAULT, status);
+    imp_sth->rs_array_idx++;
+    if (imp_sth->rs_array_status==OCI_SUCCESS &&
imp_sth->rs_array_num_rows<=imp_sth->rs_array_idx)
+    {
+       OCIStmtFetch_log_stat(imp_sth->stmhp, imp_sth->errhp,
+		imp_sth->rs_array_size, (ub2)OCI_FETCH_NEXT, OCI_DEFAULT, status);
+       imp_sth->rs_array_status=status;
+       OCIAttrGet_stmhp_stat(imp_sth, &imp_sth->rs_array_num_rows,0,
+                OCI_ATTR_ROWS_FETCHED, status);
+       imp_sth->rs_array_idx=0;
+    }
+    if (imp_sth->rs_array_num_rows>imp_sth->rs_array_idx)
+      status=OCI_SUCCESS;
+    else
+      status=imp_sth->rs_array_status;
     }

     if (status != OCI_SUCCESS) {
@@ -1698,7 +1721,9 @@
     for(i=0; i < num_fields; ++i) {
 	imp_fbh_t *fbh = &imp_sth->fbh[i];
 	fb_ary_t *fb_ary = fbh->fb_ary;
-	int rc = fb_ary->arcode[0];
+        int row_idx=imp_sth->rs_array_idx;
+	int rc = fb_ary->arcode[row_idx];
+        ub1* row_data=&fb_ary->abuf[0]+(fb_ary->bufl*row_idx);
 	SV *sv = AvARRAY(av)[i]; /* Note: we (re)use the SV in the AV	*/

 	if (rc == 1406				/* field was truncated	*/
@@ -1724,8 +1749,8 @@
 		    ++err;	/* fetch_func already called oci_error */
 	    }
 	    else {
-		int datalen = fb_ary->arlen[0];
-		char *p = (char*)&fb_ary->abuf[0];
+		int datalen = fb_ary->arlen[row_idx];
+		char *p = (char*)row_data;
 		/* if ChopBlanks check for Oracle CHAR type (blank padded)	*/
 		if (ChopBlanks && fbh->dbtype == 96) {
 		    while(datalen && p[datalen - 1]==' ')
@@ -1747,8 +1772,7 @@
 		if (!fbh->fetch_func) {
 		    /* Copy the truncated value anyway, it may be of use,	*/
 		    /* but it'll only be accessible via prior bind_column()	*/
-		    sv_setpvn(sv, (char*)&fb_ary->abuf[0],
-			  fb_ary->arlen[0]);
+		    sv_setpvn(sv, (char*)row_data,fb_ary->arlen[row_idx]);
 		    if (CSFORM_IMPLIES_UTF8(fbh->csform))
 			SvUTF8_on(sv);
 		}
============================== patch diff file end
==============================


On Oct 2, 2007 6:12 PM, Tim Bunce <Tim.Bunce@pobox.com> wrote:
> On Fri, Sep 28, 2007 at 12:42:24PM -0400, Biswadeep Chowdhury wrote:
> > Hello Tim,
> >
> > I have modified the patch taking into account your comments
> > -> Array fetch has been enabled for "nested cursors" too, in addition to
> > "ref cursors".
> > -> Array fetch is enabled by default (not for regular selects ofcourse -
> > which still uses OCI prefetch setting)
> > -> Array size is calculated based on fields (although I have restricted to a
> > max of 128 rows for now)
> > -> There is no separate parameter to manipulate array size
> > (RefCursorCacheSize), RowCacheSize itself is used.
> >
> > I am listing the patch diff file below (shorter now).
> > Unless someone can point out that elusive OCI setting, I request you to
> > consider this.
>
> John Scoles at pythian.com takes the lead for DBD::Oracle development now
> (for which I am eternally grateful). I'm sure he'll be considering it carefully.
>
> Thank you for your help - this is an important issue to resolve.
>
> Tim.
>
0
cbiswa
11/6/2007 9:56:23 PM
Reply: