Unlock a world of possibilities! Login now and discover the exclusive benefits awaiting you.
How can I speed up the AccessPoint load?
I have a clustered server environment, with 2 servers running QVS and a separate server running Accesspoint.
Version 11 SR8 with plenty of CPU and RAM, running 105 documents.
Accesspoint takes 6 - 10 seconds to load, doesn't matter if list or thumbnail view is selected, or if 1 favorite or all documents are displayed.
Migrated from version 10, which was displaying list all documents in less than 2 seconds.
Do the thumbnails load when list (details) view is selected?
Is there a way to prevent the thumbnails from loading?
If not, is there a way to cache the thumbnails?
While it is minor delay, my customers are disappointed we moved to a newer, faster environment and the response is slower.
Any suggestions would be appreciated,
B
Hi Brent,
I know this question has been sitting out there for a few months, and I was just wondering if you found anything that was helpful.
I'm dealing with the very same issue in our environment. It seems that regardless of the category selected or the number of documents/thumbnails being displayed, the loading content process takes 5-10 seconds. Sometimes longer.
Like you, our environment is clustered and AP is on a different system.
One thing that we have tried is to break NTFS security inheritance on subfolders that contain qvd generator documents, which are not accessed by our users; they're only used in document reloads. The users now only see the interface documents in AP. However, the impact of this change seemed to be pretty minimal.
Thanks!
Josh
Josh,
I've reverted to a single server, while I stabilize the environment. I'm pointing the DNS entry for the AccessPoint to primary server and speeds are significantly faster.
It might have something to do with the way we have configured the Management Services. They are running on each server, where it's recommended to centralize. The fact that we used a VM for the AccessPoint also might have caused slow downs.
We store our source and user documents in separate folder hierarchies, so the NTFS security item you mentioned wouldn't help our situation.
Regards,
B
HI,
Do you have a lot of documents with section access applied to them? Are you running IIS or QVWS.
Bill
Bill,
8 of the 113 documents have section access, and we are using QVWS for the web server.
Regards,
B
Did you find a solution to this problem?
I have a new install, which is connected to a AD.
AccessPoint takes about 20 secs to load!
What did I miss to configure?
The issue I had was centered around my configuration of the cluster. Each node was running it own services separately, and the load balancing Access Point server was running on a VM.
My assumption is, that the access point was having to query each of the nodes separately, doubling the processing time. On top of that, the VM may not had enough resources.
This issue does not seem to be caused by AD.
Is your architecture set-up similarly?
B
Hmm okay.
We have:
A) Physical server running QlikView Server + AccessPoint (IIS)
B) Virtual server running Publisher/Management Console / DSC
Our previous server ran QlikWebServer, everything on one server, no AD-connection...
So it's either the split of the servers, IIS or the AD that introduced the delay.
In which logfile should I search for clues?
My feeling is that there's some sort of function that does a timeout or similar...
EDIT:
Created a local test user, still slow. So I guess the AD isn't the problem
I'd start by looking in the QVS and IIS logs.
If you have a tool like Fiddler, it might help identify where the bulk of the lag is occurring.
You might want to try running the QV Web Server to see if it helps. It can exist with IIS.
Ok, thanks.
I'm happy to say that this issue resloved itself automagically.
We have been doing several reboots during the installation, but just to be extra sure I did another reboot of both machines, and now it works without the delay.
This seems really strange since both machines had been restarted several times after finishing the installations, so no idea why the extra reboot solved it