credits!

Our first validator snafu was more minor in scope and something that did not seem to justify the time commitment to work on correcting the credits. However, the issues over the last 24-48 hours have been so severe in scope and impact that there was nothing fair about just say tough. Also, based on data that continues to remain favorable it seems that somewhere between: heat, load, or power supply lays the culprit. We continue to run in as trimmed down a fashion as possible to allow WU hand out, validation, and credit. This will continue until our host gets some high cfm fans in place or we need to stop work hand out due to disk space. I spent this evening writing code to grant credit to results that failed validation but match the result marked as valid for the workunit. The updates have been applied for the result, host, user, and team total_credit. I did not try to adjust all the expavg values *just* total_credit. Also, please note that applies to WUs that already have a valid result turned in *and* up to and including the past 10 days. There are many results waiting for a result to mark the workunit valid and will allow me to run this code again. Please note this code is very specific to the two issues encountered between the validator failures and the upload handler failures that resulted in validation failures. Also, please note this level of credit adjustment marks the most extreme needed and applied in the history of the project under BOINC in over 3 years. We certainly cannot promise that every single cobblestone has been granted or will be but I think everyone should be fairly accepting of their credits versus a few days ago. Please use the forums to discuss this and other issues as much as possible. Our admin staff of 2 people, who both have full time day jobs, is not able to investigate every single user. Hopefully though this adjustment and ones to come as soon as a valid result+workunit are available to DC against will overall be fair in compensating everyone.

More...