Hello,

I've completed 2 tasks and received another one since the last "breakdown". I still don't get any credit. Is it normal os is there a problem with my account? I hope the data I've sent are not forever lost.

Thank you.

Hi Marie,

You are not alone -- we are all missing credit. As has been discussed in several other threads, there are certain server functions that are not yet working properly. Based on what other knowledgeable volunteers have posted, our work is not lost and credit will catch up eventually. I know it can be frustrating, but we all need to be patient.

Happy crunching,

MarkR

---

Marie

Credits for this project are based on the trickle_up files returned to the server.
There is a list of trickles received, on the page of each model.
From there, "normally" several scripts run overnight (UK time), to create credit for each of these trickles. (For models, user totals, teams.)

However, since the recent server replacement, there has been a problem with these scripts, which the project people are trying to find and correct. Which wasn't helped by the failure of the power supply unit in the server soon after. :(

Data that's been returned for each model is totally separate to the trickles/credit.

Thank you very much for your quick and nice answers.

I thought it is all on our "Pending Credit" page in our accounts, that's not the case? I have a tremendous amount of credit there and I know I won't get credit for all of it but I assumed most of it.

Pending Credit is whatever BOINC puts into that variable.
It's not used by this project, nor reset by it.

Every night, (or at least, 'back in the good old days'), some scripts, ('cron jobs'), run, which re-calculate from scratch, all granted credits. (Based on a predetermined value per trickle for that model type.) This newly calculated value is put into the appropriate variables, and then shows up in various places. Including, hours or days later, the external stats sites.

And this re-calc is why, when someone leaves a team, their credits disappear from the team totals after the next re-calc. Which, in turn, is why the team owner has to be nice to people here for ever and ever. :)

I'm guessing this won't be helpful, but maybe someone should look at /var/spool/cron/crontabs/* from the old server?

I'm guessing this also won't be helpful, but here's a cleanup line in my crontab which deletes old lines from my "sessions" and "challenges" tables in my database called "http_auth":

50 * * * * PATH=$PATH:/sbin mysql -h localhost -u username http_auth -e 'DELETE FROM sessions WHERE expire<NOW();' ; mysql -h localhost -u username http_auth -e 'DELETE FROM challenges WHERE timeout<NOW();'

Does the calculation script access a Mysql server with ssh? Maybe you need the servers's private key file?

(Sorry guys. You're in the penalty, and so are now subject to spurious suggestions from those of us who know precious little about Mysql.)

My guess is that the solution to the credits is straightforward but is waiting till other more pressing work is done.

Hopefully you're right, but it could have more to do with the admin who setup the scripts being long gone from the project. That shouldn't be taken as criticism of the current staff: one of the most difficult things in IT is to take over a complex, production database with which you had nothing to do with setting up. It honestly makes me shudder thinking of the enormity and the complex math underlying the CPDN model databases.

I'm pretty sure my above suggestions are worthless. It's just fun to posit solutions from the safety and isolation from one's living room.

And back to suggestions:

Script user and/or ip address granted access (update database name set Host='ip address' ; update username set Host='ip address')? Remote access to 3306/tcp (or explicity configured) port open on Mysql server firewall? And the classic head-thumper from every phone representative's manual: is the ethernet cable plugged in?

As Leslie Nielson would say: "I just want to tell you both good luck. We're all counting on you." x5 times.

Video. It's really funny--if you watch the whole movie.

Okay, from here on out I'm spelling it �thernet.

Well, why not just phone Carl and ask how it works..
I recall from earlier days that these Oxfordians debated if credits are necessary.

Who's Carl? The only system admins I know of right now are Andy and Jonathan:



(Based on Jon's photo, he could be a vampire.)

I see here he's listed under "Former Staff".

And another suggestion:

Maybe the script sourced credentials in a .bashrc file, like sections 1.2.1.1 to 1.2.1.2 here.

The above maybe my best suggestion, because it's one of those things that may not get copied to a new system, it's obscure enough to be totally left out of any documentation, and will have vacated the memory cells of former admins after nine years from when it was written.

My talking to Carl probably wouldn't achieve much. What might bear fruit is for anyone who's even looked at a database to post ten suggestions to this thread. I'd like to see 100 more in the next hour. Come on people!

If I remember well: Carl wrote the origional stats scripts, then he left, stats got overloaded by additional crunchers and he came back on a short visit and rewroted it to the slimmer way that works today.

Miss you Carl :-)

Ed: Oh, Belfry, silly me better leave this thread then, heh

[crickets,...crickets]

Hmm, probably with so many people typing at once it's overloaded the forum.

Ed: but doesn't explain how I got through.

Another suggestion:

Could it be the older system used a different default shell, e.g. Bash? The new one might use Dash, in which case the scripts would need "#!/bin/bash" instead of "#!/bin/sh" at the top.

Probably not it, but if we get a few thousand of these ideas up here, the statistics will break in our favor.