Expand my Community achievements bar.

SOLVED

WatchedFolderEndpoint, Stage folder filling up

Avatar

Former Community Member

I have a service in Livecycle ES that processes PDF files, via a WatchedFolderEndpoint. When the process is complete the original files are no longer needed. I have the watched folder endpoint configured to only preserve on failure, and that is working correctly. The issue is that the files that are in the stage folder never get deleted, even after they have been processed. I know that the files have been processed because I can see them in the database, which is the last step of the service. The folder is on a remote (to the livecycle server) path, and I don't think its a time issue because there are folders in there from the previous day.

My question is this, Why are the PDFs staying in the stage folder after being processed? The watched folder endpoint is used to pickup documents that have been placed there by scanners and we get around potentially 800~1000 a day, so this issue could eat up a lot large amount of disk space very quickly. Right now its manually being deleted periodically.

Here are the settings for the watched folder endpoint:

Path = \\xxx.xxx.xxx.xxx\SomeFolder\AnotherFolder\

Asynchronous = True

Cron Expression: = 0/20 * * * * ?  (Every 20 Secs)

Repeat Count = -1

Throttle = false

Username = SuperAdmin

Domain = DefaultDom

Batch Size = 15

Wait Time = 1000

Include File Pattern = *.pdf

Result Folder = empty

Preserve Folder = empty

Failure Folder = failure/input

Preserve On Failure = true

Overwrite Duplicate Filenames = false

Purge Duration = -1

1 Accepted Solution

Avatar

Correct answer by
Level 7

We found this same issue where the stage folder filled up to about 29,000+ folders and files and the CPU utilisation went to 100% and never dropped. The server system process was constantly trying to parse the stage directory and as it filled up was thrashing the CPU & disk.

You HAVE to set the result directory parameter in the WatchedFolder Startpoint Endpoint configuration even if you arent using it and set it to purge after 1 day (or anything other than -1 which is never delete). This way the stage folder will self-delete.

View solution in original post

2 Replies

Avatar

Correct answer by
Level 7

We found this same issue where the stage folder filled up to about 29,000+ folders and files and the CPU utilisation went to 100% and never dropped. The server system process was constantly trying to parse the stage directory and as it filled up was thrashing the CPU & disk.

You HAVE to set the result directory parameter in the WatchedFolder Startpoint Endpoint configuration even if you arent using it and set it to purge after 1 day (or anything other than -1 which is never delete). This way the stage folder will self-delete.

Avatar

Former Community Member

Yea we figured that out after getting in contact with our support provider. Its a bug with how the watchedfolder endpoints proceses documents. If the result folder is empty it never gets rid of the stage files. Thanks for you help though