RE: [PERFORM] Dbt2 with postgres issues on CentOS-5.3‏ - Mailing list pgsql-performance

From MUHAMMAD ASIF
Subject RE: [PERFORM] Dbt2 with postgres issues on CentOS-5.3‏
Date
Msg-id BAY154-w577ED84EDB41A1D3FB92C2FFDC0@phx.gbl
Whole thread Raw
In response to Re: [PERFORM] Dbt2 with postgres issues on CentOS-5.3‏  (Mark Wong <markwkm@gmail.com>)
List pgsql-performance
I am facing sar related issues on Redhat Enterprise Linux64 5.4 too (60G Ram, No Swap space, Xeon Processor).

sar -o /var/dbt2_data/PG/Output/driver/dbt2-sys1/sar_raw.out 60 204
    |___ sadc 60 205 -z /var/dbt2_data/PG/Output/driver/dbt2-sys1/sar_raw.out

It generates following sar data i.e.
….
….
03:52:43 AM      2.31
03:53:43 AM      2.31
03:54:43 AM      2.28
03:55:43 AM      2.31
03:56:43 AM      1.67
03:57:43 AM      0.29
03:58:43 AM      0.29
04:00:43 AM      0.30
04:04:00 AM      3.52
04:07:07 AM      0.30
04:09:36 AM      0.23
04:12:04 AM      0.36
04:14:25 AM      0.23
04:16:45 AM      0.26
04:19:10 AM      0.24
04:21:30 AM      0.38
04:23:55 AM      0.24
04:26:25 AM      0.35
04:28:48 AM      0.24
04:31:10 AM      0.27
04:33:40 AM      0.33
04:36:45 AM      0.41
04:39:12 AM      0.27
04:41:41 AM      0.26
04:44:11 AM      0.33
04:46:35 AM      0.25
04:49:06 AM      0.33
04:51:27 AM      0.27
04:53:56 AM      0.23
04:56:19 AM      0.36
04:58:43 AM      0.24
05:01:10 AM      0.35
05:03:43 AM    &nbs p; 0.33
05:06:53 AM      0.29
05:09:25 AM      0.23
….
….

To fix this issue I have modified the sysstat-9.1.2/sadc.c and replaced signal based pause (That is not real time) with "select" based pause. That fixed the issue. Thanks.

-------------------------------------------------------------------------------------------------------------------------------------------------------
sadc.c.patch
-------------------------------------------------------------------------------------------------------------------------------------------------------
--- sadc.c.org    2010-06-14 21:44:18.000000000 +0500
+++ sadc.c    2010-06-14 22:52:51.693211184 +0500
@@ -33,6 +33,10 @@
 #include <sys/stat.h>
 #include <sys/utsname.h>
 
+#include <sys/types.h>
+#include <sys/time.h>
+#include <time.h>
+
 #include "version.h"
 #include "sa.h"
 #include "rd_stats.h"
@@ -792,6 +796,15 @@
     }
 }
 
+void pause_new( void )
+{
+    struct timeval tvsel;
+    tvsel.tv_sec = interval;
+    tvsel.tv_usec = 0;
+
+    select( 0, NULL, NULL, NULL, &tvsel );
+}
+
 /*
  ************************************ ***************************************
  * Main loop: Read stats from the relevant sources and display them.
@@ -899,7 +912,7 @@
         }
 
         if (count) {
-            pause();
+            pause_new();
         }
 
         /* Rotate activity file if necessary */
-------------------------------------------------------------------------------------------------------------------------------------------------------

Best Regards,
Asif Naeem

> Date: Wed, 21 Apr 2010 18:10:35 -0700
> Subject: Re: [PERFORM] Dbt2 with postgres issues on CentOS-5.3‏
> From: markwkm@gmail.com
> To: anaeem.it@hotmail.com
> CC: pgsql-performance@postgresql.org
>
> 2010/4/20 MUHAMMAD ASIF <anaeem.it@hotmail.com>:
> > Hi,
> >
> > I am using dbt2 on Linux 64 (CentOS release 5.3 (Final)) . I have compiled
> > latest postgresql-8.4.3 code on the machine and run dbt2 against it. I am
> > little confused about the results. I ran dbt2 with the following
> > configuration i.e.
> >
> > DBT2 Options :
> >     WAREHOUSES=75
> >     DB_CONNECTIONS=20
> >     REGRESS_DURATION=1 #HOURS
> >     REGRESS_DURATION_SEC=$((60*60*$REGRESS_DURATION))
> >
> > DBT2 Command :
> >         ./dbt2-pgsql-create-db
> >         ./dbt2-pgsql-build-db -d $DBDATA -g -r -w $WAREHOUSES
> >         ./dbt2-run-workload -a pgsql -c $DB_CONNECTIONS -d
& gt; > $REGRESS_DURATION_SEC -w $WAREHOUSES -o $OUTPUT_DIR
> >         ./dbt2-pgsql-stop-db
> >
> > I am not able to understand the sar related graphs. Iostat,mpstat and vmstat
> > results are similar but
> > sar results are strange. I tried to explore the dbt2 source code to find
> > out the how graphs are drawn and why sar results differ.DBT2.pm : 189 reads
> > sar.out and parse it and consider 1 minute elapsed time between each record
> > i.e.
>
> That is certainly a weakness in the logic of the perl modules in
> plotting the charts accurately. I wouldn't be surprised if the other
> stat tools suffer the same problem.
>
> Regards,
> Mark


Your E-mail and More On-the-Go. Get Windows Live Hotmail Free. Sign up now.

pgsql-performance by date:

Previous
From: Dave Crooke
Date:
Subject: Re: Fwd: Dead lock
Next
From: Scott Marlowe
Date:
Subject: Re: query hangs