Script that would gather 24hrs data for folder

[rickoic]

Don't know if it's been suggested before or not, but my idea would allow a folder to run a script from Stanford that would:

1. Gather a list of each work unit his machines had submitted for the previous 24 hours.
2. Along with the list provide the points for each work unit submitted and accepted.

This would allow each individual to see what work his/her machine(s) had contributed each day and to see if they have a problem with the incorrect settings one a machine.

If it is determined that it would create too much load on the servers then a limit to the number of times it is run by each individual could be introduced.

Tks
Rick

[bollix47]

https://apps.foldingathome.org/cpu

[Akaanc]

This link shows only the last wu. I think it would be a good idea to show a list for last 100wus as in boinc.

[bruce]

A script could be run on your local machine to capture the information that is posted to fah's log when a WU is successfully uploaded. Such a script would be very easy to write for Linux. (I'm not particularly familiar with Window's scripting so I can't guess how difficult it would be for Windows.) If there is no problem, fine. If you suspect a problem, you can look up each of those WUs since you know the PRCGs that were uploaded over the last "day." If there's a WU that failed to upload, you could look at the filtered log through FAHControl -- or you could extend the script to capture that information too.

You might want to start by running hfm, since it already captures most of that information.

[rickoic]

Bruce
That would work for our side of it, but wouldn't give a check on which wu's were credited to us.

Tks
Rick

[bruce]

I predict that the number of credited WU would match the number of uploaded WUs at least 99% of the time as long as the applicable time periods matched and the points would be reasonably close.

[rickoic]

I suspect so also, however there is still that 1% of the time when something like this would help.

[bruce]

In that 1% of the time, all you need is the list of PRCGs -- you can look up each one here: https://apps.foldingathome.org/wu and add them up.

I still recommend hfm.net

[rickoic]

Tks Bruce, wasn't aware of that link.

[MeeLee]

Isn't that information written in the log files?
All you'd need is a script that can find certain search criteria in the log file, and parse it inside a new file.
This script would be interesting to help find Failed WUs, as well as any Overclocking errors (since the Console log is too difficult to find for multi GPU setups that run days on end).

[toTOW]

I still recommend hfm.net

I agree that HFM built in Work Unit History Viewer will to the job without adding extra load on the stats servers ...