After uploading previous task qcnb_023507_0 (in between, a second CPDN task grabbed one more gigabyte), BOINC refused to get a replacement, because "Quake-Catcher Network | Message from server: QCN Sensor needs 631.57MB more disk space. You currently have 322.10 MB available and it needs 953.67 MB.". But after upping BOINC's allowed disk usage and obtaining another task, QCN's disk usage still remained at mere 6.32 MB.
Is it because of the 20 similar blocks in the sched_reply_qcn.stanford.edu_sensor.xml file?
<file_info>
<name>qcnb_030014_0_1.zip</name> ... <name>qcnb_030014_0_20.zip</name>
<generated_locally/>
<upload_when_present/>
<max_nbytes>50000000</max_nbytes>
<url>http://qcn-upl.stanford.edu/qcn_cgi/file_upload_handler</url>
</file_info>
My trigger files are just a couple of kilobytes large. Or is the reason somewhere else?<name>qcnb_030014_0_1.zip</name> ... <name>qcnb_030014_0_20.zip</name>
<generated_locally/>
<upload_when_present/>
<max_nbytes>50000000</max_nbytes>
<url>http://qcn-upl.stanford.edu/qcn_cgi/file_upload_handler</url>
</file_info>
Could the "theoretical maximum" QCN disk requirements be a bit reduced?
Peter
PS: I do hate akismet! It is a dumb obstacle, triggering just false positives!
/ 
Reply
Quote
yeah I probably should drop that down from a gigabyte, that was back when we started and our output files were much bigger.
/ Reply
Quote