I have been running this model for now 465 hours and recently for the last four days I have not received any credits. Is this a problem?

12/15/2007 10:59:41 PM|climateprediction.net|Restarting task hadsm3fub_e093_005910721_6 using hadsm3 version 506

---

Your computer has made contact recently, so network activity is allowed. If you have a look at the graphics, what do they show?

If the graphics are black and white mush, then that\'s because the model has run through a phase change. Just quit BOINC, restart it, and check the graphics again.

If the graphics show all blue for temperature, then the model is an \'ice world\' and may have slowed down: trickle intervals can extend to a week or more.

Two other models have completed in that work unit, but on different platforms (one AMD/Windows, one Intel/Darwin - yours is Intel/Windows), so if it is an ice world I would happily abort it, given that it\'s only 7/24 through the third phase.

If the graphics look OK, we\'ll have to think again ...

not an ice wrold as I see reds and oranges for temperatures, can se the continents and greens. What next, let it run for awhile longer?

Your computer has made contact recently, so network activity is allowed. If you have a look at the graphics, what do they show?

If the graphics are black and white mush, then that\'s because the model has run through a phase change. Just quit BOINC, restart it, and check the graphics again.

If the graphics show all blue for temperature, then the model is an \'ice world\' and may have slowed down: trickle intervals can extend to a week or more.

Two other models have completed in that work unit, but on different platforms (one AMD/Windows, one Intel/Darwin - yours is Intel/Windows), so if it is an ice world I would happily abort it, given that it\'s only 7/24 through the third phase.

If the graphics look OK, we\'ll have to think again ...

No credits because no trickles.
Last contact with server: 14 Dec 2007 20:22:24

The 14th is about the time of the monthly Windows update. Check that your firewall is still letting BOINC access the server. You should have messages about it if it can\'t get through.
And if the trickles have been created but not sent, they will be in the climateprediction.net folder, and will start with: trickle_up_.

---

Backups: Here

Well I aborted the task , upgrade to 5.10.30 restrted my computer, downloaded another file and now I am getting http://internal server error.

This is really strange becasue malariacontrol.net is running fine and all other previuos climate control projects ran fine. In checking folders I do not see one starting with trickle_up.

Any thoughts...thanks in advance

No credits because no trickles.
Last contact with server: 14 Dec 2007 20:22:24

The 14th is about the time of the monthly Windows update. Check that your firewall is still letting BOINC access the server. You should have messages about it if it can\'t get through.
And if the trickles have been created but not sent, they will be in the climateprediction.net folder, and will start with: trickle_up_.

The \'internal server error\' is most likely due to the project servers being overloaded (a lot of people have been seeing this recently). Hopefully it\'ll resolve itself after a while.

Sometimes it can also be due to firewall/proxy servers, so if things haven\'t sorted themselves out after a day or so, post again and we\'ll make some suggestions.

I see a trickle on the server dated about 30 mins prior to your post: http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result.php?resultid=7053134 - is this the right PC?

---

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

Disavowed
Unless you have Network access suspended, you\'re not going to see the trickle_up files. They\'re very small, and are uploaded in an instant. AND they don\'t appear in the transfers tab.

I have been seeing similar errors popping up over the past week. I\'m hoping that the servers haven\'t been suffering from an excess of holiday party going. Haven\'t seen any comment at all regarding the underlying problem though.....

The \'internal server error\' is most likely due to the project servers being overloaded (a lot of people have been seeing this recently). Hopefully it\'ll resolve itself after a while.

---

Re: errors

It\'s being discussed on the php board a bit.
The team are running a lot of database scripts, searching for data the physicists want for a paper. This will continue for some time.

Note that this is not related to credit problems people may be having, which is a different matter.

Note that this is not related to credit problems people may be having, which is a different matter.

And what causes those? I have noticed, that in recent days I only got credit for SM trickles and not for CM. Here is one missing credits for second trickle, while this one had its trickle (dated 2007/12/21) credited today.

Greetings,

---

If the servers are running slowly as at the moment, it can take time for trickles to get transferred from the upload server to the \'storage\' server, and then there is a delay for the credit program to run.
In theory, it runs once per day, but the exact time may be varying, as it\'s a seperate program, and it too may have been put on hold for a while, while the search scripts were being run. The credits program re-calculates credit for all models each time it runs, as BOINC sometimes allocates it\'s own credits at random to some models when they finish, giving them extra.
Because the credits script takes up a lot of server resources, it may have been stopped for a while.

But credits aren\'t a life or death situation, and will get granted, even if a few hours late.

What I\'ve been seeing is upload issues (trickles not going thru) - often with http errors. I figure these will get caught up over time. Also, when trying to access my account, or the message boards, I\'ll run into mysql error messages. I note here that the mysql error messages (too many connections mostly) seem to happen right around the time when the daily updates to user accounts are made (for me that is around 11PM local (MST).

If the servers are running slowly as at the moment, it can take time for trickles to get transferred from the upload server to the \'storage\' server, and then there is a delay for the credit program to run.
In theory, it runs once per day, but the exact time may be varying, as it\'s a seperate program, and it too may have been put on hold for a while, while the search scripts were being run. The credits program re-calculates credit for all models each time it runs, as BOINC sometimes allocates it\'s own credits at random to some models when they finish, giving them extra.
Because the credits script takes up a lot of server resources, it may have been stopped for a while.

But credits aren\'t a life or death situation, and will get granted, even if a few hours late.

---

Barry
The text you quote applies to vid.

My message before that applies to you.

Because the daily updates use a lot of server resources, if the physicist\'s database queries are running at the same time, you WILL get the problems that you mention.
All of the items that you mention are on the same server: climateapps2.oucs.ox.ac.uk

OK -- thanks for clearing up my confusion on this.

So the reason we are seeing server issues is that the physicists are USING the data -- that\'s OK by me.

Barry
The text you quote applies to vid.

My message before that applies to you.

Because the daily updates use a lot of server resources, if the physicist\'s database queries are running at the same time, you WILL get the problems that you mention.
All of the items that you mention are on the same server: climateapps2.oucs.ox.ac.uk

---

...
But credits aren\'t a life or death situation, and will get granted, even if a few hours late.

Thanks for explanation, I didn\'t mean it was a life / death situation, it was just you mentioned it and I noticed it, so I asked about details. And yes, got my trickles credited somtime today.

Greetings,

---

The following was posted in the news and announcements earlier today regarding the problem I noted in this thread:

\"When trying to upload trickles, the message that there\'s an HTTP error has occurred. It doesn\'t mean that the CPDN server is down, or that your internet connection has failed. The best idea is to suspend network activity in BOINC manager and allow it again a short time later.\"

For me, the idea to suspend network activity is suboptimal even as a workaround. A primary focus of the development work on the BOINC client was the concept of multiple project support. I\'ve got a total of six current active projects and reporting results or getting work for the other projects doesn\'t go away when I encounter the reported problem with CPDN. Since I maintain something of a large farm of workstations in various locations, this sort of manual intervention is not something I\'m keen on. Instead, I simply let the workstations continue processing (and reporting or attempting to report when CPDN is having a bad hair day) so that the other projects don\'t get adversely affected by a loading problem at CPDN.

Thinking about it, perhaps an option for further enhancement in the BOINC client would be support for having a suspend network activity be project specific. There are times I suspend a specific project on a workstation -- when an announcement of a planned outage of say 12 or more hours on a specific project is made, or a project goes temporarily vapor with no notice. Doing that takes advantage of the BOINC multi-project concept by simply shifting CPU cycles to projects that are not encountering problems at that point in time.

I\'ve noted one manual workaround that sometimes resolves an HTTP error (which I have seen on occasion with other projects) is to unload BOINC and restart it on a workstation. This results in the upload succeeding (at least some of the time).

---

I did propose a boinc enhancement to do this (it\'s in the Trac system somewhere).

---

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