SharePoint 2010 Fast Disk Space Part 2 (Solution)

by Balaji Kandasamy

Plan

The solution for the issue would be to stop the service and clear the logs. Start the service back try to process the old click analysis. It should typically process those files and improve the link rankings.

Note: This will have an impact on the search results proceed with caution.

Backup

Backup the following locations to a different server or a drive.

:\FASTSearch\var\log\sprel\

:\FASTSearch\data\sprel\ (all folder below this)

:\FASTSearch\components\resourcestore\generic\clickthrough

Execution

Stop the sprel process deletes the logs, start the service back and run the SPRel process cmd and ensure everything works as expected.Test all the process in a Test environment before trying it in a live environment.
  1. Run the following command to stop the sprel process:
nctrl stop sprel walinkstorerreceiver
  1. Clean the data/sprel directory (Except config folder)
  2. Run the following command to start the sprel process again:
nctrl start sprel walinkstorerreceiver
  1. Confirm everything is ok by running the following command
spreladmin ShowStatus
  1. Start the processing by executing the following command
spreladmin StartProcessing
  1. Regularly check the status by executing the following command
spreladmin ShowStatus
  1. Once completed, status shows ready doing nothing: ensure the process is completed

Update

we faced issue while incremental crawl. The following error popped up and due to this, some search results were affected. 
The Content Plugin received a "Processing Error" response from the backend server
for the item. ( ERROR: PostProcessBatch failure in processor SPCIDStorer. 
Traceback (most recent call last): 
\\ File "SPCIDStorer.py", line 63, in PostProcessBatch \\ File "TableSender.py", 
line 111, in send \\ WindowsError: [Error 3] The system cannot find the path 
specified: 'E:\\FASTSE~1\\data\\sprel\\cidstorer_backup/*.*' \\ [17]
Fix: Create an empty folder :\FASTSearch\data\sprel\cidstorer_backup

Results

The expectation is the logs shouldn't grow at a rapid pace, the clickthrough archive should be processed and important the search service shouldn't see any outage.drive>:\FASTSearch\components\resourcestore\generic\clickthrough (data pushed by the SP extraction job) should be empty if the process worked as expected. Monitor the process for a couple of days and ensure there are no errors.

Check the following location should be empty if the process worked as expected.

:\FASTSearch\components\resourcestore\generic\clickthrough

(data pushed by the SP extraction job)

Start an incremental crawl to ensure there is no impact on the results. Monitor the process for a week and ensure there are no errors in the sprel logs or crawl logs.

Ref Links:

https://blogs.msdn.microsoft.com/microsoft_search_bloggers/2015/02/10/sprel-consuming-more-disk-space-and-spreladmin-status-not-showing-as-ready/

https://social.technet.microsoft.com/Forums/exchange/en-US/a76fad93-35a4-4de9-8662-c4c7f8b24225/sprelexe-high-utilization?forum=fastsharepoint

http://blog.comperiosearch.com/blog/2011/07/12/learning-about-nctrl-disabling-fast-search-web-crawler/

https://technet.microsoft.com/en-us/library/ee943516(v=office.14).aspx