View Drop Folder

[TB]

We are having problems clearing the view drops folder of various xml files,
and understanding exactly is stored in this folder. We know it impacts the
accuracy of the CUBe build process and the publishing queue...but

1)what exactly is the view drops folders and what data does it contain?

2) I open a xml file and it stores Project Name and then it displays
EntProjectID>nnnn<
then later it displays ProjectID<nnnn< but it is different than
EntProjectID, what is the difference?

Finally,
3) what does the timestamp represent?

 

[Marc Soester]

Hi TB,

I cannot answer all of your questions, but some

The view drop folder is the folder where Project server puts xml files when
saving/publishing a project and / or trigger Server requests ( like the cube).

I wouldnt necessarily know what the date stamp and inforamtion within the
file mean, but bottom line you dont need to know, as long as Project Server
does.

Your problem is different. The view drop folder should be empty. It looks
like that you have a permission issue on the folder. I would give your
service accounts modify access on the folder. This should rectify your issue
that the xml files will remain in the folder ( which they shouldnt)
Hope this helps

 

[Paul Conroy]

All project users need read/write access to the view drop folder !

Make for the views processing service is running on the applicatio server.

 

[Marc Soester]

Hi Paul and TB,

I am sure that you know what you are talking about, but I believe that the
service account should be enough, otherwise this would be a potenital
security issue.
I know that at the installations I done, the service account was sufficient.
TB, you may want to try both options, first the service account and if that
doesnt do the job, you may want to include "everyone".
Lets hope one of the suggestions will solve your issue

 

[Paul Conroy]

It's under the users security context which the spooler writes to the view
drop folder. The service account running the Views Processing Service also
needs access to retrieve the files and publish. This is documented in the
installation guide.

Paul

 

[TB]

Thank you all for the info...although the jobs clear out eventually, it can
take up to 24 hours before the publishing of a project can clear the view
drops folder. Which makes it difficult to know if you have all the
information you are looking for in the OLAP cube. Seems that the queuing in
the view drops folders is a problem due to our number of projects...over
7000, and publishing is occurring regularly. Does throwing hardware at the
problem, clear the queue faster, or should we expect this folder always to be
empty, or cleared out faster?

Any thoughts on what the two project IDs represent in one xml file content?
I have found that when a person publishes a plan, we may find 3 files with
the same project name in the views drop folder, not sure why 3 files would
exist? I can understand the one job, when they hit Collaborate, Publish, all
information, but that is all the user is doing. The did not publish the plan
3 times. The three xml files will be relatively close in time of each other
(within minutes).

Thanks for your knowledge on this, it is very helpful.
Regards,
TB

 

[Gary L. Chefetz [MVP]]

Tommy:

Two XML packages get generated when you publish and close the project. One
is the publish, the other tells the system to checkin the project. With a
portfolio as large as yours, you should have a second views processing
service running on another machine. You should also make sure that none of
your users have auto save turned on.