For the past two days, I\'ve seen a different problem, while there are daily updates, the size of the credit is significantly less than \'normal\' -- that is, I\'ve been reliably running about 3500 a day, and for the past two days it has been 1/3 of that. It seems this is not a \'just me\' event, looking at the daily credits for others via BOINC stats, this seems to be a general thing.

I\'ve seen the \'no trickles\' problem in the past, but *partial trickles* is a new one I believe. Any comment?

---

There\'s no such think as a partial trickle as far as credit is concerned.
It may be that the credits program is running at different times now and then, as programs are stopped to move data around on the servers.
Also, the programs that move the trickles from the upload servers to the trickle server seems to be slow at present.

This would mean that sometimes people don\'t get all of thge trickles for the day included in that days credit run, and the export file run.
But these WILL get included in the next days runs.

Hmm, for three days now there has been a substantial drop in credits, going from around 3000 per day to 2200 or less.

---

www.metta.org.uk
The Holistic Health Information Service

Never mind the credits, are the same number of trickles being uploaded per day?
Trickles = credits, and trickle uploads are recorded in the messages. (Also in stdoutdae.txt)

It maybe worthwhile scanning through your PCs to see if there are any with a \'stuck\' model. The most common cause is a slow-processing iceworld slab model.

For example, this slab?

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7015736

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Take a look at this model:

hadcm3istd_01h7_1920_160_05922419_0

It last trickled, according to the website. on the 18th of January. The model trickles once per model year, right? As you can see the last trickle was for 1954, I think, could be 1955. When looking at the graphics for that model on my computer it just started crunching 1960, where are the 4 or 5 trickles produced since 1954/55? I restarted BOINC not too long ago so I can\'t show you the trickle messages in the log, but the one I saw(yesterday) did not show an error message, transfer tab is empty by the way. Can this be attributed to a remark in this message of the news and anouncements thread (quote: There have also been delays recently with the appearance of our successful trickles on our models\' project web pages.)?

I think there is a lag on trickles at the moment, as you\'ve read.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

Hi Transient

Yes, the HADCM models trickle every year about 4 December.

I have a HADCM model that\'s now reached Jan 2024, but the graph on its web page still only shows 2021. The last trickle recorded there was on 17 Dec which was Thursday! The model must have trickled in Dec 2022 and 2023 since then, but these trickles don\'t show yet.

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=6705729

So yes, it\'s all very delayed. The fact that Tolu and Milo have been setting up the new CPDN beta site as a new project with a new url may have caused extra delays.

Milo thinks the CPDN servers now understand that \'weekend\' means \'take time off\'. I expect we\'ll all receive the delayed credits in a day or two. The credits don\'t get lost, they\'re just temporarily stuck inside one of the servers.

---

Cpdn news

One work unit (hadsm3fub_032r_005920398_2) 6119020 is currently at 14.07% with no trickles? Date:01/04/1817 Hrs elapsed:0043:34:11 (1.43s/TS)
Is this normal?

---

www.metta.org.uk
The Holistic Health Information Service

No, it\'s not normal, but most likely it\'s due to the current delay on trickles being shown on the website.

Are there any \'trickle_up_...\' files in the project\'s folder? (c:\\program files\\boinc\\projects\\climateprediction.net folder, or possibly in a \\hadsm3fub_032r_005920398_2\\ subfolder).

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

OK -- I have something of a BOINC collection of workstations I work with -- so what I\'m seeing may well be a case where at the moment, some small proportion (say 25%), successfully upload the trickles or show the work as credited on a given day. I had not seen that in the past, either I got the full set of credit or none, on a given day. In any event, I guess I (along with the rest of us) can hope that at some point in time, there is a \'catch up\' action which brings things back into sync.

There\'s no such think as a partial trickle as far as credit is concerned.
It may be that the credits program is running at different times now and then, as programs are stopped to move data around on the servers.
Also, the programs that move the trickles from the upload servers to the trickle server seems to be slow at present.

This would mean that sometimes people don\'t get all of thge trickles for the day included in that days credit run, and the export file run.
But these WILL get included in the next days runs.

---

I\'ve looked at CPDN on BoincStats and a small number of credits has been added this evening at the daily update. I seem to have a small number, presumably for yesterday\'s crunching. So a lot of trickles and credits are stuck in the system.

---

Cpdn news

I\'ve PMed Milo and Tolu, Milo will be back on Tuesday. It seems that the trickle job has been causing issues recently and it has been modified to try to rectify this, but may need more work.

---

I'm a volunteer and my views are my own.
News and Announcements and FAQ

OK -- that tracks with me.

I\'ve PMed Milo and Tolu, Milo will be back on Tuesday. It seems that the trickle job has been causing issues recently and it has been modified to try to rectify this, but may need more work.

---

There is now a flood of trickles, so it looks like the backlog has been cleared: the corresponding credits will no doubt be updated overnight.

There is now a flood of trickles, so it looks like the backlog has been cleared: the corresponding credits will no doubt be updated overnight.

Not cleared yet - but clearing ...

Still has a good few to catch up on my account.

F.

---

Sounds good -- I\'ll check for that this evening (US-MST)(which will be after the overnight update in the UK).

There is now a flood of trickles, so it looks like the backlog has been cleared: the corresponding credits will no doubt be updated overnight.

---