I'm running Climate Prediction on my XP,along with a the other usual suspects. My problem is that Climate Prediction seems to be taking up a helluva lot of diskspace. When I look at c:\Program Files\BOINC\projects/climateprediction.net and checkout used diskspace, I see the following;

Size Directory
---- ---------
9895 2pm6_100147788
9895 2prj_100147983
9895 2pwz_100148181
9895 2q90_100148619
9895 2qfh_100148854
9895 2qh4_100148913
9895 2qw7_100149462
9895 2qxs_100149519
9896 2pyu_100148249
9896 2qfz_100148872
11270 2pvv_100148141
152111 0yx5_000065733
340608 2qyc_100149540

I am currently crunching 0yx5_000065733, so what are the others doing in there. I have a had problems with this project erroring out and downloading new units before the previous one finished. Can anyone tell me if the other directories are dead units, and can I delete them/

Slainte,

David

---

> I'm running Climate Prediction on my XP,along with a the other usual suspects.
> My problem is that Climate Prediction seems to be taking up a helluva lot of
> diskspace. When I look at c:Program FilesBOINCprojects/climateprediction.net
> and checkout used diskspace, I see the following;
>
> Size Directory
> ---- ---------
> 9895 2pm6_100147788
> 9895 2prj_100147983
> 9895 2pwz_100148181
> 9895 2q90_100148619
> 9895 2qfh_100148854
> 9895 2qh4_100148913
> 9895 2qw7_100149462
> 9895 2qxs_100149519
> 9896 2pyu_100148249
> 9896 2qfz_100148872
> 11270 2pvv_100148141
> 152111 0yx5_000065733
> 340608 2qyc_100149540
>
> I am currently crunching 0yx5_000065733, so what are the others doing in
> there. I have a had problems with this project erroring out and downloading
> new units before the previous one finished. Can anyone tell me if the other
> directories are dead units, and can I delete them/
>
> Slainte,
>
> David
>
>
Yeah, the other directories are dead units & you can delete them. they are useless.
Note that,if you can, you should keep the <b>completed</b> Wus on your HD or a DVD/CD: sometimes the science team needs to examine them closely and asks users to send them.
Hope that helps..:o)

---

Arnaud

Welcome to the board.

You don't want to delete your current WU as you have worked out. You are asked to keep the completed WU if possible. You don't need to keep such completed runs in the boinc directory, it is ok to move elsewhere or backup to CD/DVD or whatever then delete it from the BOINC directory.

The only model that that computer has completed is 2ngc_100144959 which I do not see listed. :?

340Mb might have been a complete unit but I checked and there were only 32 trickles. So you can delete all bar the 0yx5_000065733 and 2ngc_100144959 directories.

If you have a model that has completed a couple of phases, it has not been completely wasted and may be worth keeping if you have lots of spare space.

HTH

Thanks to all

---

IMHO that "feature" could turn into a Problem for Users over time, although it sure will likely take a long time to clog the HDD.

But eventually, either BOINC's HD limit or the physical limit will be reached, effectively stalling BOINC.
Would be useful to have a clearly visible message, asking the User to backup (if feasible) the data in question.

Otherwise, it will remain an "insider"...

---

Scientific Network : 44800 MHz - 77824 MB - 1970 GB

I have been backing up my completed models in zipped files. Now that I have a working CDRom rewriter, I suppose I should also be backing them up onto disks. Can I copy them to CDRom in zipped condition, or do I have to unzip them first?

---

__________________________________________________

&gt; Can I copy them to CDRom in zipped condition, or do I have to unzip them
&gt; first?
&gt;
There shouldn't be any problem copying them as zip files.

&gt; &gt; Can I copy them to CDRom in zipped condition, or do I have to unzip them
&gt; &gt; first?
&gt; &gt;
&gt; There shouldn't be any problem copying them as zip files.
&gt;

What I'd like to see is this platform sending the delete command to computers do remove files and directories that they wouldn't need anymore if the science teams determinded that they were no longer needed. I know this happens over as Enstein when the 11 MB file is no longer needed the server sends the message to the client to remove the file and if the client is does with it (no work units running it) then it removes it. Might be a nice feature to have here. Just my two pennies

You are right, Jason. However, my impression is that there are so far no model types that the experimenters know they will definitely never need again. Let's hope that when, in the future, particular experiments are superseded, we do get a message that we can delete them.

---

__________________________________________________

&gt;
&gt; What I'd like to see is this platform sending the delete command to computers
&gt; do remove files and directories that they wouldn't need anymore if the science
&gt; teams determinded that they were no longer needed. I know this happens over
&gt; as Enstein when the 11 MB file is no longer needed the server sends the
&gt; message to the client to remove the file and if the client is does with it (no
&gt; work units running it) then it removes it. Might be a nice feature to have
&gt; here. Just my two pennies
&gt;

There is a problem here, except for the CPDN-files that is downloaded/uploaded, the files is generated directly by CPDN and isn't listed as [file_info]-elements in client_state.xml, meaning BOINC doesn't know anything about them.

So, till CPDN starts listing all the files that is normally left when successfully finished model as [file_info]-elements including the [sticky/]-tag, they will not be automatically deleted.

I’m on my second WU.
The left behind accumulated data related to my first one amounted roughly to 400 megs. I just moved this data to another drive for the time being.
I had to look into this because I just lost a Predictor WU (stderr out: Maximum disk usage exceeded) and couldn’t understand how it could have happened.

I’m curious about this.
I’m happy to contribute computing time to various projects. However I HAVE NEVER AGREED TO GRANT FREE LONG TIME STORAGE SPACE to CPDN.

So, should CPDN be interested in this data, they can have it. I’ll upload it anywhere they want. If not , well then it’s going down the drain in the near future…

---

"Tamdiu discendum est, quamdiu vivas" Seneca

Looking at the CPDN Project Directory, I also found alot of older Client Versions.

Each is only a few MB in size, but I know SETI@Home deletes those anytime a newer Version is Downloaded, for example.

IMHO the CPDN File Management does need a small overhaul to keep track on which Files are still needed and which are not.

I'm saving my completed Models to a DVD-RAM right now to free up the Disk Space.

---

Scientific Network : 44800 MHz - 77824 MB - 1970 GB

&gt; Looking at the CPDN Project Directory, I also found alot of older Client
&gt; Versions.
&gt;
&gt; Each is only a few MB in size, but I know SETI@Home deletes those anytime a
&gt; newer Version is Downloaded, for example.
&gt;
&gt; IMHO the CPDN File Management does need a small overhaul to keep track on
&gt; which Files are still needed and which are not.
&gt;
&gt; I'm saving my completed Models to a DVD-RAM right now to free up the Disk
&gt; Space.

Not as simple as at first glance, at least not for Multi- or HT CPUs. For example, my WinXP box (P4 3.0) is currently crunching its last 4.04 Model and a 4.12 Model. An automatic purge would/could have left three 4.04 Models stranded.

An after-the-fact clean-up routine could do the job, but given a single programmer on the CPDN project currently, an Alpha underway (Sulfur) and another under development (Coupled Model), daily fires to fight, et al, I'm content to do my own infrequent after-the-fact cleanup.

Given his current workload, I doubt it would seem a "small overhaul" to Tolu.

Perhaps a user-written add-on?

---

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

&gt; Looking at the CPDN Project Directory, I also found alot of older Client
&gt; Versions.
&gt;
&gt; Each is only a few MB in size, but I know SETI@Home deletes those anytime a
&gt; newer Version is Downloaded, for example.
&gt;
&gt; IMHO the CPDN File Management does need a small overhaul to keep track on
&gt; which Files are still needed and which are not.

The problem is that BOINC only deletes the downloaded files when a client is no longer required. The CPDN hadsm3 project consists of 3 programs, with the controller downloaded as an executable and the 2 workers downloaded as compressed files. BOINC doesn't know about the uncompressed executable worker programs (because there is no record of them in client_state.xml), so they won't be automatically deleted.

---

"The ultimate test of a moral society is the kind of world that it leaves to its children." - Dietrich Bonhoeffer