on ‎2019 Aug 22 12:18 PM
Hello,
I have an issue with a client where performance for the Launchpad, opening Web Intelligence reports and opening queries in WebI is slow. It can take 30 seconds or more from the authentication screen to the documents view at logon, over 3 minutes to open a large report for editing, and almost a minute to open the query panel and view WebI queries.
The problem system is 4.2 SP6 running on a Windows Server 2016 virtual server, with dedicated resource. It has 8 cores of AMD Opteron Processor 6172 at 2.1 GHz and 24GB RAM. I am using Chrome v76 on the desktop of the server itself.
At first logon of a day the logon process takes up to a minute, with this output in the Chrome console:

This 'Waiting (TTFB)' (Time to first byte) is consistent for the first logon on a day - subsequent logins are quicker.
Then when I come to open a large report, albeit one where the data has been purged from I get these timing from the Chrome console:

52 seconds to download the file

27 seconds wait time again - after which the report first page has come to the screen and the structure/formatting is visible.

then the left hand pane to display the universe available objects takes another 3.7 minutes to become visible.
I accept that this report is very large - 19 queries, 21 report tabs, dozens of variables. However I loaded the same report into another system and had these comparative timings:
For the test machine timings I am using Chrome v76 not on the server desktop but on my own laptop. The server is Windows 2012 R2 running Intel Core i7-4790 @ 3.6GHz (8 cores) with 32 GB RAM on a cloud hosted VM.
These timings show some obvious findings - that a faster server returns data more quickly even over a wider network. However the logon wait time far exceeds anything I've seen at other clients, and the general slow timings for things like opening of query panels, moving between queries, opening smaller reports etc etc suggests something is not right.
Can anyone help with suggestions of what might be wrong, where the extreme latency is coming from?
thanks
Keith
Request clarification before answering.
I'd recommend checking the CMS database. BO generates hundreds of queries, and high latency there can cause significant delays. If that's not a problem, you can do an end-to-end trace and submit the results to tech support for analysis.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.
| User | Count |
|---|---|
| 15 | |
| 9 | |
| 6 | |
| 5 | |
| 4 | |
| 4 | |
| 3 | |
| 2 | |
| 2 | |
| 2 |
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.