Are trickles still necessary? I know why they were created. I was there, back in the old days, when slow, single core processors took 6 to 9 months to finished a single 160 year model. Trickles were need back then because it was not reasonable to expect that Crunchers would wait to till a model finished to get their credits.

Things have changed. Models are shorter, processors faster. Today, models take 21 days or less to crunch most models. Many finish in only 3 or 4 days. That is not an unreasonable amount of time to wait for credits. So do we really need them anymore, or have they outlived their usefulness and should be retired.

---

It seems there is an underlying assumption in your post that tasks will continue to successful completion.

If you have tasks which crash after days or weeks of processing and all credit is lost, how would you react? Some of the research examines the margins of possible atmospheres and might generate impossible conditions in that search. Trickles provide a way to salvage some/most of the credit for work done. (How do we address the furor arising from those situations?)

---

"We have met the enemy and he is us." -- Pogo
Greetings from coastal Washington state, the scenic US Pacific Northwest.

Trickles provide a way to salvage some/most of the credit for work done. (How do we address the furor arising from those situations?)

Do the trickles still give the scientists any information or is that just the zips these days? If the latter, could the receipt of the zips have the same function or would that get complicated due to how them not carrying the credit information in the same way as the trickles?

Credit is there and it seems to be correct. Gooooood job.

Credit is there and it seems to be correct. Gooooood job.

I pay little enough attention to mine that I have no idea whether it is correct or not, just that my total has gone up and my recent average down. The latter is no great surprise after so long without any work.I haven't tried to investigate yet whether the trickles not showing on the task pages means the credit is a bit out of date or not.

Of my four tasks running under Linux, all are showing the first trickle including one that uploaded yesterday. The others should have 3, 5 and 6 trickles showing yet only the first appears on the task page. I wonder if this is an issue specific to this batch of hadcm3s tasks?

Done some checking and it seems all of batch 749 are only showing 1st trickle and only have credit for that. Project now informed. The relatively small numbers of these tasks seems to have let me report it before the complaints appear.

David has replied and informed me that this will be looked at in the next few days.

I do not see any trickles after 11.09 for any batch/machines I have.

I also have one sam25 batch 742 received 09.09 that hasn't recorded a trickle on the web.

I do not see any trickles after 11.09 for any batch/machines I have.

I also have one sam25 batch 742 received 09.09 that hasn't recorded a trickle on the web.

This has been communicated to the project people.

All missing trickles are there, but credit is actually missing. This is the next job to follow.
Well done....thanks

---

I do not see any trickles after 11.09 for any batch/machines I have.

I also have one sam25 batch 742 received 09.09 that hasn't recorded a trickle on the web.

Just until one hour ago I also was missing all trickles for my current task , that is running for the last 5+ days. But hodiho, trickles have been inserted during the last hour and are showing up now.

Good work!

All missing trickles are there, but credit is actually missing. This is the next job to follow.

The credit won't show till next time the credit script is run on Saturday. Not sure what happened to the new script that would be able to run daily.

Though batch 749, only the first trickles seem to show. This is true both for completed tasks and those still in progress. I had a first one go through last night which is showing so it is not just trickles sent before something went wrong showing.
These are the hadcm3s models that I think only went out to Linux and Mac hosts.

From the project.
Hi Dave,

Batch 749 is HadCM3 and I think we have seen the “only a single trickle appearing” error before for this model. It is a bit confusing as it produces trickles ok in a local run. I an afraid I don’t have much time at the moment to do fault finding on this but I will add it to the to-do list.

Best wishes,
Sarah

I have given a bit more information to Sarah in that the trickles appear to upload fine according to my event log so we will have to wait and see on this one.

Looking at my tasks' web pages, the credit seems to be going up as expected despite the trickles not appearing there.

That probably explains the lack of furore here about it.

Maybe a database issue:

http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=21269808

http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=21235822

http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=21251981

All of them show no trickles. I've got some more wu without trickles.

Also this one:

http://climateapps2.oerc.ox.ac.uk/cpdnboinc/result.php?resultid=21264260

Have a nice day,

bonsai911

---

All my WUs do not show trickles on the web some time after early morning on 21st.

New trickles finally visible for new tasks.

---

But not the hadcm3s tasks. They are only showing the first trickle but on comparing the 12 and 120 month tasks it seems that once the first trickle is posted the credit for the whole task is granted. I will email project again.