Not certain this is your problem, but atleast on other BOINC-projects if the sched_request-file gets too large the server can't handle it since this will use more memory or something than a single scheduling-thread is allowed to use. While this normally have happened with users trying to report thousands of results at once, it is possible your WCG-results has maxed-out the logfile and/or the WCG-server is allowed to use less memory.

If this is your problem, you can try reporting fewer results in a single request. For this to work, you'll need to run BOINC v6.12.xx or later. In your BOINC data-directory, edit a file (or make the file if not present) called cc_config.xml

This file should minimum include:

<cc_config>
    <log_flags>
    </log_flags>
    <options>
        <max_tasks_reported>50</max_tasks_reported>
    </options>
</cc_config>

If you've already got a cc_config.xml, it's possible you've already got the <max_tasks_reported> in it, if so just edit this line to a fairly low limit. I've used 50 in my example.

After saving the file, in BOINC Manager's Advanced view, choose the menu-option "Advanced" - "Read config files".

Afterwards, try to report your results. If this now works, you can try increasing #tasks reported at once. If it doesn't work, try with a very low setting of 5 or something.

If this doesn't work, someone more familiar with Linux would need to help you...

Brilliant, that's fixed it! Many thanks again! super!

Thought I was going to loose all those wu's for a while!