It's probably going to take another 24 hours or more to sort everything out. I'm restoring the last backup overnight so that we can identify what's gone missing.

Let's see...the project as a whole is (- 3,074,783,684) credits and (- 8,367) registered users since yesterday, according to BOINCStats. I can't wait to see what the graphs look like when it all comes back ;D

hi,

within the last 24hours i lost 30% of my credits!

http://de.boincstats.com/stats/user_graph.php?pr=cpdn&id=417573

whats going on?

ralf

I got hit also. According to the Team USA stats I lost more than 200,000 credits. Milo, please put them back.

---

Hello mo.v,

The tasks were missing from my computers MICHAEL1 (Host ID 982 801) and MICHAEL2 (Host ID 975 929). From a total of 44 tasks are 5 tasks disappeared.

I have the task numbers and work unit numbers in my Excel list. The task names are also there. Here they are:

Task ID, work ID unit, computer ID, Sent (UTC), Time reported or deadline (UTC), Task Name
9091392, 6390814, 975 929, 23 Jun 2009 03:52:57, 01 Aug 2009 01:42:10, hadam3p_mhor_1959_2_006124757_2
9082262, 6394539, 982 801, 17 Jun 2009 18:19:00, 17 Jun 2009 18:21:46, hadam3p_mkk8_1989_2_006128482_2
9082261, 6395540, 982 801, 17 Jun 2009 18:21:46, 20 Jun 2009 19:51:39, hadam3p_mlc1_1976_2_006129483_4
8960588, 6405836, 975 929, 10 Jun 2009 19:27:52, 16 Jun 2009 18:24:57, hadam3p_mta1_1989_2_006139779_0
8948236, 6399667, 982 801, 14 Jun 2009 20:07:21, 17 Jun 2009 18:19:00, hadam3p_moio_1989_2_006133610_1

I think the rest of the data are first unimportant. If they are still needed, just ask, or here is the table as an HTML document.

The credits are missing from Task 8960588 with 415.80 credits (Error while computing after 5th Trickle Message) and Task 9091392 with 2,079.00 credits (Completed). For the others there was no credit, since this ended before the first trickle message with an error while computing.

---

My Boinc Stats

Thanks for your effort on this Milo. Once the dust settles, it may be worth dumping a backup copy on a test server and running thru the archive process step by step. Something isn't right somewhere. Since this happened last year during the same process as this year, that could possibly save a lot of folks a lot of grief, especially you, unless you particularly enjoy dealing with this sort of thing every year (NOT!!!). ;)

Boinc Manager 24,675.66 Account 16,774.17 loss of 7901.49

Milo thanks for being on top of this.

---

What you do today you will have to live with tonight

I dropped from 1040000 credits to 720000 on 9/1/2010.
User ID 137222

---

I've had a credit problem as well..

I didn't have much, but I was in the area of 37,000 or so.

Now, when I log in to climateprediction.net, it registers me at 18,754.26 credits.

BOINC Stats also reflected this change. I actually had an overall NEGATIVE score for September 1, 2010 as reflected on my BOINC Stats page.

User ID 574780

I have lost credit of around 10,000 from around 73,000 to below 63,000. about 15% of total

Nearly 25 k down. Hope everything will back 2 normal in some days.

Thank you very much for the information from your Ecxel file, Michael.

I've checked that the tasks you list are in fact missing from your two affected computers.

I think that block(s) of workunits have probably disappeared, not blocks of tasks (though of course the tasks are in the WUs).

Here is one workunit that has disappeared:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=6390814
If I edit the workunit number in that address and look for the next WU I find:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=6390815
also disappeared because the archiving program has missed a whole block. It would be possible to search forwards and back by editing the numbers to find the size of the 'black hole' but I don't think this is necessary at the moment.

There could f course be more than one black hole of missing WUs.

So in WU number order, missing WUs we know about are:

6390814
6394539
6395540
6399667
6405836

We will be able to check that when Milo restores his backup these tasks and their workunits reappear. When the tasks reappear their credit will also return to our accounts.

I don't think (at the moment) that this problem can affect tasks before April 2009. Carl archived those old models last year so I think Milo probably didn't archive them again. But I'm not a programmer and I'm not an expert in these matters.

Milo has decided that when he archives the database again he will not do this as one big job. He will archive smaller sections separately. He thinks this will make errors less likely, and if errors occur they should be quicker to locate and correct.

Thank you again, Michael. I have no list of all my models!

---

Cpdn news

I have lost my credit too. More than 20k were lost (recently: more than 169k, now: 148,880)

---

Credit has dropped from 697,000 to 537,059 also three places in team.

---

My Credit has dropped from 800190 to 506226.

---

My credit dropped from 368,947 to 129,048 (BOINCstats) !
When this gets corrected, the "best five days" is sure going to get updated ;-)

-- T.

My credit dropped from 368,947 to 129,048 (BOINCstats) !
When this gets corrected, the "best five days" is sure going to get updated ;-)

-- T.

Yes, be ready for the best best day in history ;)

---

Grüße vom Sänger

Silly question, but what will happen when I try to trickle-up models that have dissappeared? I'm assuming that I shouldn't try this until the restore has happened...

My current model is:
Task ID: 7934758
Work Unit ID: 6212331
Computer: 865981

Silly question, but what will happen when I try to trickle-up models that have dissappeared? I'm assuming that I shouldn't try this until the restore has happened...

My current model is:
Task ID: 7934758
Work Unit ID: 6212331
Computer: 865981

You shouldn't be able to, as I turned the BOINC daemons off once it was clear there was a serious problem. I'll turn them back on again when I've dealt with it.

Also, the vanished results should be old ones which have already finished.

Whoa! I've lost at least 210k credits (about 29%) since the month of Sept started. Hope Milo is on this.

MacDitch, every single task for your computer has disappeared! I don't really know the answer to your question but I would make sure the computer has plenty of work to keep it busy then suspend Boinc network activity.

The info you gave about that model is interesting. When I try to go to the web page for that task I get 'Unable to handle request: no such task' ie it's disappeared. But when I look for the workunit page:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=6212331
it exists but contains no tasks. It looks as if this could be a separate(?) missing block consisting of a run of tasks, not WUs.

I've looked further back and forwards from that task page looking for every 10th task page. They've all gone:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7934708
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7934798
Here is 1000 tasks back:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7933758
and much further back:
http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=3720758
This seems to be a very large missing block (if it's all one block).

---

Cpdn news