Piyush
05-05-2005, 10:56 AM
Hi everyone,
I hope someone out there can provide some tips or insight into this
issue.
I am executing performance testing on certain reports for my work.
This report is 4 pages with a total of 7 line graphs. Now I realize
there are multiple issues here, underlying SQL, report server config,
hardware, # CPUs, RAM, etc...
my test server is a windows 2000 server, 3 GB Ram, 1 2.4GHz CPU. all
allocated to a VMWare image running the app server/report server. The
reports server has the default config except I changed the initEngine
to 4, maxEngine to 6, and minEngine to 1. There are 4 users
concurrently running reports by invoking a URL via a JSP page.
Most rdf reports that are run come back between 10 to 20 seconds. but
this report takes 20 to 70 seconds. I followed the reports tuning
section in the documentation. The only thing I can't do at this point
is replace my rdf with rep files (long story). The trace file seems to
indicate a the vast majority of time (>65%) dedicated to processing
images. There doesn't seem to be anything I can do about that or is
there?? btw, the SQL takes about 100 milliseconds when I run it in
TOAD.
Their white paper on scalability tests a windows box with 1 GB RAM and
1 866MHz CPU and it shows it can scale up to hundreds of users with
response time less than 10 seconds! Base on my config above, I should
be spitting these reports out in no time! They did not show the report
that was being executed. If I created a text only report and kept
retrieving from cache then it would be blazing fast also. but that's
not the real world scenario for me.
I guess long story short...is there anything I can do to quicken the
response time via the reports or reports config or is it just a matter
of getting more hardware resources?
I hope someone out there can provide some tips or insight into this
issue.
I am executing performance testing on certain reports for my work.
This report is 4 pages with a total of 7 line graphs. Now I realize
there are multiple issues here, underlying SQL, report server config,
hardware, # CPUs, RAM, etc...
my test server is a windows 2000 server, 3 GB Ram, 1 2.4GHz CPU. all
allocated to a VMWare image running the app server/report server. The
reports server has the default config except I changed the initEngine
to 4, maxEngine to 6, and minEngine to 1. There are 4 users
concurrently running reports by invoking a URL via a JSP page.
Most rdf reports that are run come back between 10 to 20 seconds. but
this report takes 20 to 70 seconds. I followed the reports tuning
section in the documentation. The only thing I can't do at this point
is replace my rdf with rep files (long story). The trace file seems to
indicate a the vast majority of time (>65%) dedicated to processing
images. There doesn't seem to be anything I can do about that or is
there?? btw, the SQL takes about 100 milliseconds when I run it in
TOAD.
Their white paper on scalability tests a windows box with 1 GB RAM and
1 866MHz CPU and it shows it can scale up to hundreds of users with
response time less than 10 seconds! Base on my config above, I should
be spitting these reports out in no time! They did not show the report
that was being executed. If I created a text only report and kept
retrieving from cache then it would be blazing fast also. but that's
not the real world scenario for me.
I guess long story short...is there anything I can do to quicken the
response time via the reports or reports config or is it just a matter
of getting more hardware resources?