Hi! Currently I’m(my server) struggling from FileStore\Input folder’s size growing. Currently it is more than 70GB, with 15GB in output folder. As I know input folder is the location for report definition file, that actualy must not be in size of 800Mb or more. But now I’ve got to many files with that big sizes.
In old version (BOXI R2) I’ve solved this problem in the following way:
found one big file in Input folder;
using query builder searched si_name (SELECT * FROM CI_INFOOBJECTS WHERE SI_ID = ‘18928069’ – files folder name);
found si_name (reports full name) in InfoView, opened it and purged the data;
then saved it back to repository;
in current version above listed steps meets with failure, e.g. SELECT * FROM CI_INFOOBJECTS WHERE SI_ID = ‘18928069’ gives no result, or there is no data to purge in report.
Anyone knows how to handle the situation in BOXI31? Comments and suggestions are welcome.
I’d question those sizes - take some actual cases to see if they do actually equate to their KB equivalent etc.
I have scripts that I run across directories and you can also use explorer search - but yo then only get the WID name and not the report name - and that’s where Query Builder comes in handy …
Do you see the number of files growing significantly over time?
I can’t verify that this is correct and I don’t know if it’s the same for 3.x, but I was told by a BO engineer a couple of years ago that files saved in your users’ favorites folder are stored and managed by the Input FRS not the Output FRS as you might expect.
Running “select si_name, si_files, si_size from ci_infoobjects where SI_RUNNABLE_OBJECT = 1 and SI_INSTANCE_OBJECT = 0” in Query builder gives me the list of about 120 rows. Each row looks like:
The SI_VALUE1 represents the size of the .WID file in bytes
(1MB = 1048576).
Short-term solution could be to export the Query Builder output into an Excel spreadsheet and manipulate the data to go from columns to a single row for each entry (it’s ugly but can be done in a matter of 10-15 minutes). Then perform a data sort.
We use a home-grown process to extract this info to tables for used in a universe. You may want to look into doing that as well or purchasing a third party tool.