climateprediction.net home page
Posts by geophi
Posts by geophi
log in
1) Message boards : Number crunching : Upload failures (Message 60613)
Posted 12 days ago by Profile geophi
Thanks for the history on the cam25's. I found the one in question, and it has returned 18 zips.
https://www.cpdn.org/cpdnboinc/result.php?resultid=21709022

However, the ones that are stuck are #12 and #13. So it looks like they got lost in the shuffle.
If they have not uploaded by the time my other work has finished tomorrow, I will just can them
(as in trash can; I just realized that may not be clear to non-native English speakers).

What I've done with previous CAM25s that have persistently stuck uploads is just abort the transfers that are obviously stuck. After aborting the transfers, it will report the task, possibly as a success, and the scientists can determine whether the output is useful without the missing zips. I've only done this with the CAM25 tasks however since those are the ones that seem to occasionally have the rogue stuck uploads.
2) Message boards : Number crunching : Upload failures (Message 60518)
Posted 18 days ago by Profile geophi
I've got version 7.14.2, surely that's newer then version 6.12.xx mentioned above, so doesn't sound like that particular problem. This has been going on for days now...


My reply that mentioned the boinc version was in reponse to Chris L's and Dave's wondering about whether there is still a 14 day limit on upload tries before the task aborts...and whether cpdn now requires newer versions of boinc that may not have such limits.

The client version has nothing to do with a full disk on a project server.


Absolutely.
3) Message boards : Number crunching : Long runs and low credits (Message 60508)
Posted 18 days ago by Profile geophi
The PNW ones have gotten inflated credits relative to the other models types for a long time (when looking at credit for time crunched) so they are not the ones to compare other models to. The HADAM4 are in line with the other models such as SAM50, SAFR, SAS, ANZ and GLOBAL.

It's the SAM25, CAM25 and EU25 that are getting too low credits relative to the other models and those are being looked at for possible retroactive adjustment upward.
4) Message boards : Number crunching : Upload failures (Message 60483)
Posted 21 days ago by Profile geophi
2. I think that this may only apply to VERY old versions of BOINC.


Not entirely sure if the current models will run on versions of BOINC that old though I can't remember what minimum version we need here is now?

The version of SSL/TLS that cpdn servers use (since https) requires the client to also have upgraded capabilities in that area. As I remember from when WCG upgraded, the minimum version of boinc that should work with the newer SSL/TLS is 6.12.xx. I remember because on some Linux PCs I was running 6.10.58 before the WCG changeover and I had to change.
5) Questions and Answers : Getting started : "Communications deferred" ... continuously (Message 60460)
Posted 22 days ago by Profile geophi
Yes. It looks like you are having the same problem the one that mmonnin linked to. I was wondering if it was the https problem, but it's just the one associated with the server being full for that type of model. I hope they free up some space soon.
6) Questions and Answers : Getting started : "Communications deferred" ... continuously (Message 60450)
Posted 23 days ago by Profile geophi
For the last couple of weeks, CPDN has stopped accepting my uploads. It keeps sending me new work units, but with all the uploads waiting, I am running out of disk space. What is happening?

Can you copy and paste the messages in the message log when the file uploads fail into a post here?

In boinc manager, in the Projects tab, click on climateprediction.net and click on the Properties button the the left. Does the first line have https or http in it?
7) Questions and Answers : Windows : Tasks abandoned after computer crash (Message 60441)
Posted 23 days ago by Profile geophi
There is no way for you to receive those tasks again. The tasks in those work units will be reissued to other computers.

It's frustrating to be sure, but it's happened to all of us at one time or another.
8) Message boards : Number crunching : Free-DC reports negative credits today for CPDN (Message 60428)
Posted 24 days ago by Profile geophi
Looking at boincstats, everything looks reasonable, but perhaps the stats were only good through yesterday and didn't take into account any oddities due to the server failover?
9) Message boards : Number crunching : Upload failures (Message 60414)
Posted 26 days ago by Profile geophi
SAM and SAFR go to Jasmin. Full again.

The project IT people were notified of the SAM/SAFR problems early Saturday morning. Haven't heard anything back yet.
10) Message boards : Number crunching : Upload failures (Message 60408)
Posted 26 days ago by Profile geophi
upload3 for the hadam4 uploads is remote to oxford (at RAL?) so Sarah is trying to contact them.
11) Message boards : Number crunching : Upload failures (Message 60405)
Posted 26 days ago by Profile geophi
Same here, hadam4 uploads timing out, can't access upload3 server.
12) Questions and Answers : Unix/Linux : Benchmarks and other problems (Message 60383)
Posted 28 days ago by Profile geophi
They seem to be running about twice as fast as the expected completion time. It was expected that they would take about 1050 hours to complete, but one is at 18.455% complete after running 72.5 hours.

The initial estimate for time to completion is partially based on the boinc floating point benchmark. For some reason the 7.2.33 version that comes with Redhat 6 type installations has unrealistically low benchmarks. Some thing for my Phenom II 945 on CentOS 6. 7.2.33 gives a FP benchmark of about 1600 whereas the later versions of boinc on Ubuntu have about 3000 for the same CPU.
13) Message boards : Number crunching : 215 work units with credit pending. (Message 60353)
Posted 18 Jun 2019 by Profile geophi
I have no idea why some tasks are labeled as "validation pending" and others are not. There is no true boinc validation in cpdn (other PCs returning basically identical results for the same work unit) and there never has been. Credits are based on trickles returned times the credits/trickle for each model type. I have nearly 3000 tasks under the "validation pending" boinc tag but have been given credit for all of them.
14) Message boards : climateprediction.net Science : Project announcements discussion (Message 60350)
Posted 18 Jun 2019 by Profile geophi
I think we can say that these are 64bit because of the individual task memory requirements.
15) Message boards : Number crunching : New work (Message 60341)
Posted 17 Jun 2019 by Profile geophi
I hope the problem with shutting down boinc client when these are running crashing the v8.08 work-units is fixed. I will try to keep them running. All boinc work-units have permission to keep them in memory even when suspended

All tests on the beta/dev site showed this newer version of the app allows for shutting down the tasks and then restarting them without a problem.
16) Questions and Answers : Unix/Linux : 8.09 UK Met Office HadAM4 at N144 resolution (Message 60336)
Posted 17 Jun 2019 by Profile geophi
Yeah, the memory requirements of these are not too much different than the higher res WAH2 Windows tasks. There was a small batch of these run on this site back in early February (batch 785), but they had the problem of not being able to recover from restarts. That problem has been fixed.

Tests have been run on the beta site of a higher resolution version of hadam4 (N216) which took a little over twice the RAM per task, and a workstation version of the ECMWF called openIFS which has required 4 to 6 or more times the RAM of these hadam4 N144 tasks. Who knows when one or the other of these two higher requirements models will make it over to the main cpdn site.
17) Message boards : Cafe CPDN : Op Ed Piece in the NY Times Regarding Climate Model Computing (Message 60298)
Posted 13 Jun 2019 by Profile geophi
requirements will probably change before release. And that may well be for the worse.


OK time to order that new cpu and motherboard. Will go for 32GB memory initially but with space to add more when I can afford to increase it enough to get more cores running.


What CPU are you going for Dave?
18) Questions and Answers : Unix/Linux : *** Running 32bit CPDN from 64bit Linux - IMPORTANT *** (Message 60253)
Posted 8 Jun 2019 by Profile geophi
How on earth am I supposed to know if my machine can deal with 32-bit apps?

If you have Ubuntu 18.04, go to the command line and type

sudo apt-get install lib32ncurses5 lib32z1 lib32stdc++-6-dev

and that should take care of it. A small amount of hadam4 n144 tasks were released Friday. A lot more should be on the way next week unless there are problems. There hasn't been much in the way of Linux tasks for the last couple years. That will change shortly.
19) Message boards : Number crunching : Project going off line (Message 60249)
Posted 6 Jun 2019 by Profile geophi
I see that you are back and my trickles have uploaded so I assume that all went well?

I think so. No problems posted when Andy announced the site was back up.
20) Questions and Answers : Macintosh : New MAC Boinc Client 7.8.6 (Message 60246)
Posted 30 May 2019 by Profile geophi
Has Climate Prediction just written off Macs entirely ?

For now it appears so.

WCG has just launched a beta for a climate model. How long between beta and production though is anyone's guess. I don't know if the beta includes Macs, but for all other ongoing WCG projects, there is a Mac boinc app.


Next 20

Main page · Your account · Message boards


Copyright © 2019 climateprediction.net