Hi Bellator

Of course you\'re not wasting your time on this project! Sometimes there have been delays with the trickles and the updating of the server web pages, but everything always catches up in the end.

We do know that cpdn is not overendowed with spare server storage capacity. I believe Milo in Oxford has been intending to move data round to try to reduce the problem. If he\'s been doing this, it could be that temporary glitches have occurred. He knows about this.

---

Cpdn news

Thanks for the various pieces of advice. I keep getting \"project servers may be temporarily down\", but this has been for a long time now (I have the same problem when trying to upload Seti and Burp). Still, all other projects perform as they should. Last contact in CPDN still shown as March 23, although there have been numerous attempts at contact. When running, the figures in Tasks do keep changing, however. Will try to be patient a little longer but I welcome any other help you can give me.

---

Here\'s the cpdn server status page:

http://climateapps2.oucs.ox.ac.uk/cpdnboinc/server_status.php

Quite honestly, as long as my BBC and cpdn models continue to crunch, I just don\'t worry about delays in the credits or whether my trickle pages are up-to-date. The difficult thing is to get a model to the end; the credits and server pages all work out correctly eventually. And we can immediately see from our boinc manager messages whether trickles were successful.

So the important thing for us all is to remember to make weekly backups.

---

Cpdn news

Thanks. I am a great believer in backups. Please tell me what directories I should back up, however, as the entire BOINC folder is rather voluminous (570Mb).

---

I\'m afraid you have to back up the complete boinc folder. If there are any subfolders for previous models you\'re no longer crunching, you could delete those first.

My sig leads to the READMEs. In the one about avoiding crashes, item #1 by Les gives an easy backup method. There are several other methods in the README specifically about backups.

For any backup to work, you have to exit from boinc before you do it.

---

Cpdn news

If you backup using WinZip or WinRAR then you\'ll be using less disk space. Once you\'ve installed WinRAR, instead of right-clicking \'copy\', \'paste\', do \'Add to Boinc.RAR\'.

Alternatively, you could copy onto a CD-RW (rewritable CD).

---

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

29/03/2007 8:20:47 AM|SETI@home|[file_xfer] Started upload of file 03no03aa.12667.9250.198574.3.42_1_0
29/03/2007 8:20:48 AM|SETI@home|[error] Error on file upload: [03no03aa.12667.9250.198574.3.42_1_0] locked by file_upload_handler PID=9136
29/03/2007 8:20:48 AM|SETI@home|[file_xfer] Temporarily failed upload of 03no03aa.12667.9250.198574.3.42_1_0: transient upload error
29/03/2007 8:20:48 AM|SETI@home|Backing off 32 min 35 sec on upload of file 03no03aa.12667.9250.198574.3.42_1_0
In addition to my climateprediction woes, I now get this - for the last several days. All other programs OK, so far. I am not sure, but I think I downloaded the latest BOINC version on March 23, the day that my computer ceased to communicate with climateprediction.
I have done a complete BOINC backup.

---

That would tend to reinforce the firewall as being the culprit - when an update occurs, the new EXE doesn\'t inherit the settings from the firewall that the older EXE had.

Try making the following changes, and see if that helps:

http://www.climateprediction.net/board/viewtopic.php?t=4272

You could also add the upload servers into the firewall\'s \'whitelist\' (different firewalls will use different names, and some don\'t have this feature) - this means that the upload servers are trusted machines and bypass a lot of the firewall\'s checks.

---

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

The merging uses the descriptions of the computer to see if they are the same. Yours aren\'t, because one is \'described\' by BOINC 5.8.8., and the other by an earlier version.

Has this problem been solved?

---

Answered on other post.

---

Backups: Here

Answered on other post.

Hm, I can´t find it - do you maybe have a link to that post?

---

Links to all of your posts are on your Account page on the server.
It\'s near the top, to the right of: \"Message boards\", and is a number followed by \"posts\".

Links to all of your posts are on your Account page on the server. It\'s near the top, to the right of: \"Message boards\", and is a number followed by \"posts\".

Hi Les, thank you. Unfortunatelly on all these links I do not find an answer to the question, if the problem has been solved.

---

There IS no solution, just a a way to fix it until the next time that you upgrade BOINC.

I think it may be to edit client_state.xml, near the start, for each of your computers, and make the computer description the same.

If so, it would have to be while BOINC is stopped, and possibly best to make all of the old descriptions the same as the new description.

Where ever the post was where someone did this, it explains the steps needed to do it.
It may be on the BOINC/dev site.

There IS no solution, just a a way to fix it until the next time that you upgrade BOINC.

I think it may be to edit client_state.xml, near the start, for each of your computers, and make the computer description the same.

If so, it would have to be while BOINC is stopped, and possibly best to make all of the old descriptions the same as the new description.

Hi Les - to do that it´s very easy - turn off boinc. In client_state.xml find the according host ID. Change it. Start boinc again. Actualise. Then the data shown in the project-stats are identical.
Unfortunatelly its not enough, merging still does not work.
My guess is, that the project saves the older descriptions and that is, why it does not find mergable computers - but it is just a wild guess, as I really do not know the reason.

Where ever the post was where someone did this, it explains the steps needed to do it. It may be on the BOINC/dev site.

---

The server also keeps track of the contact sequence number of each host. When you change things and this number is not what the server expects it can create a host ID that can not be merged even if all the descriptive fields are identical.

---

BOINC WIKI

BOINCing since 2002/12/8

The server also keeps track of the contact sequence number of each host. When you change things and this number is not what the server expects it can create a host ID that can not be merged even if all the descriptive fields are identical.

Hi John, I have never heard from the \"contact sequence number\". That would explain my problem of merging some PCs, while not being able to merge others. Is there any chance to get that straited out?

---

The best way is to set \'network activity disabled\' most of the time, and only let trickles upload at intervals (say, weekly). That way, when you restore, the sequence number will always match, and a new host ID won\'t be generated.

---

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