Our current production server has the following specs: Single Quad processor, Windows 2003 x64, 32 GB RAM, IIS 6, and recently got a new test server with following specs: Dual Quad processor, Windows 2008 R2 x64, 30 GB RAM, IIS 7. Both servers have the same version of SSRS 2005 and service packs. On the test server, any report with charts when rendering in PDF format takes 3 three times longer compared to the current production server. HTML, TIFF or EXCEL formats renders really fast on the test server. Also, other reports without charts, the test server renders PDF very fast compared to the current production server. When rendering reports with chart in PDF format, the CPU utilization is very high on the test server. We are trying to figure out why our old production server outperforms the new test server for PDF rendering. We have reported this issue to Microsoft Support (Case# 110072976750051) two weeks ago, and they have not still figure out the issue. Is the issue with IIS 7? or Dual Quad processor? or ASP 2.0? I would appreciate if somone can help me figure out this issue.

Hi, This does sound confusing. Just to investigate the problem in more detail please do the following (assuming the data source is a SQL Server database); - 1) Create an SQL Server Profile based on the relevant Database Engine instance for your production server, execute your report on the production server and record this trace. 2) Create an SQL Server Profile based on the relevant Database Engine instance for your test server, execute your report on the production server and record this trace. 3) Compare and contrast 1 and 2. 4) Double check the indexes that these reports use (sorry if you have already done this) 5) When running the T-SQL in SSMS which these reports use, do you have the same difference in speed? I hope following the above steps helps you understand the problem a bit better. I'm sorry I can't give you a direct answer. Kind Regards, Kieran.If you have found any of my posts helpful then please vote them as helpful. Kieran Wood PGD SoftDev (Open), MBCS http://uk.linkedin.com/in/bileeds

Good day, I think this issue is related to Acrobat Reader version and its compatibility with the OS. Can you provide the Acrobat Reader version on both servers? Regards,My Profile Jordan Group Raed Taha

Hi Raed, I think SSRS uses it's own rendering process for PDFs separate to Acrobat Reader; - http://www.eggheadcafe.com/software/aspnet/33836304/how-reporting-services-generates-pdfs.aspx I am still at a loss as to what is causing the different between production and test servers though. Kind Regards, Kieran.If you have found any of my posts helpful then please vote them as helpful. Kieran Wood PGD SoftDev (Open), MBCS http://uk.linkedin.com/in/bileeds

Hi Kieran, You are correct; the Acrobat Reader impact should be excluded. My Profile Jordan Group Raed Taha

I have been working with Microsoft close to last four weeks. They are really good support and very helpful, but they seems to not get our production performance, for example one report that we gave them takes 22 secs to render in PDF on our production server (on the test server it takes 60+ secs) but on Microsoft support machine with similar specs it takes 46 secs. I found out this week that all the testing Microsoft did using Enterprise edition and AMD processors, and I suggested to them that this maybe why for Microsoft single or dual quad processor giving the same or similar results because our is Standard edition and both our servers are Intel Xeon. Also, I told them that whenever the report renders in PDF on the test server, only one processor is busy and other is idle. Here is what I got from Microsoft support last update: "I also wanted to provide an update on this case. I installed SQL 2005 x64 Standard with Service pack 3 Verified Microsoft SQL Server 2005 - 9.00.4035.00 (X64) Nov 24 2008 16:17:31 Copyright (c) 1988-2005 Microsoft Corporation Standard Edition (64-bit) on Windows NT 6.1 (Build 7600: ) Tested with Standard and it uses 4 CPU's out of the 8 the report takes about 60+ seconds Tested again with Enterprise and it used all 8 processors and takes about 49 seconds. So it looks like moving the Enterprise would help a little with the performance since it takes advantage of all the processors but it’s not a 2x increase in speed." I'm really confused. The Standard edition should use all processors according to the licensing model, i.e. i have two CPUs on the test server. Can someone please shed some light on this issue? Thanks.

As a note to the previous reply, Microsoft is confused with fact that 4 cores on each Quad as 4 CPUs totalling 8 CPUs for the Dual Quad machine. I know licensing does not treat cores as separate CPUs. Check http://www.microsoft.com/sqlserver/2005/en/us/compare-features.aspx Does the Standard edition have issues with Dual Quads, especially with PDF rendering?

To what extent have you followed my previous guidance given on Friday, August 20, 2010 8:35 PM.? What were the results when you performed the following; - "1) Create an SQL Server Profile based on the relevant Database Engine instance for your production server, execute your report on the production server and record this trace. 2) Create an SQL Server Profile based on the relevant Database Engine instance for your test server, execute your report on the production server and record this trace. 3) Compare and contrast 1 and 2". I'm always open to learn, but I really don't think the core issue in the difference in performance relates to the PDF renderer, it's probably just a symptom of the core issue you have noticed rather than the route cause. If you have found any of my posts helpful then please vote them as helpful. Kieran Wood PGD SoftDev (Open), MBCS http://uk.linkedin.com/in/bileeds

I have the same PDF rendering issue on our new server with SQL 2008 Standard 64 bit version, Windows 2008 Enterprise 64bit. Was this solved?