Q: Why doesn't BOINC switch projects every 60 mins like I asked it to?

A: In your General Preferences, you can set a value for "Switch between applications every... minutes". It would perhaps be better worded if it was called "Reconsider which tasks to run every... minutes".

The setting in BOINC tells it how often to reconsider which work to be crunching. So, each hour (the default value), it looks at the deadlines and the relative debt between the projects and factors that in with how often you connect to the internet, how many days of work you keep on-hand and all the rest.

Bottom line, it is deciding to continue on the work it has started. This might be because these tasks are nearing a deadline, or that other projects had deadlines previously and incurred a time debt to Rosetta which is now being repaid.

Again, expect your respective resource shares to be enforced over the course of 100 hours, not 100 minutes.

If you really want to force it to stop working on a specific task, you can suspend a task from the tasks tab, or a project from the projects tab. But you will further complicate the debt between multiple projects when you do that. BOINC learns over time how much to expect your computer to be crunching each day. If you know better then BOINC about your plans for the next few days, you might hold a task with a longer deadline so that tasks with an upcoming deadline can complete before you turn off your computer for a few days. Otherwise, BOINC automatically sees an approaching deadline and works to assure you have it done in time.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

Q: What do all the colors in the graphic mean?

A: On the home page there is a section titled "about" which has a lot of good information about the project in general. With respect to graphics there is a page titled Quick guide to Rosetta and its graphics.

Basically the colors are just the color spectrum from one end of the protein to the other. Their primary purpose is when two protein predictions (or comparing a prediction with the true structure) are overlaid on the same screen. It helps to compare the two and see where along the protein they differ.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

My computer runs 24/7 and appears to be crunching away for rosetta but I have noticed that my user average is dropping like a stone what am I doing wrong

---

Q: Why won't BOINC download more Rosetta Work units?

A: It is possible the servers are down for some reason (quite rare). You should check the server status.

But it is more likely you are seeing a message in the BOINC Manager saying "...not requesting new work". If so, then your PC is actually not asking for more work.

If you are running multiple BOINC projects, the BOINC manager plans out what will be crunched next, and balances goals like the deadlines of the work units, and your preferences as to resource share to devote to each project. In this balancing act, it sometimes crunches a project more than it's normal share in order to meet a deadline. When this occurs, the project being crunched incurs a "debt" to the other projects. BOINC will take this debt in to account when requesting new work. It is not uncommon, especially when running 3 or more projects, for BOINC to decide to crunch other projects for a period of a day or two in order to pay back such debts. And so during that time it will not order more work for projects that it does not plan to be crunching for a while.

In the end, if you leave things alone and resist the temptation to try and FORCE BOINC to do work for a specific project, it will work itself back in to balance and again download work for each project.

If you want to increase the resource share for Rosetta, you can increase the value in your Rosetta preferences, and then update to the project. You can view the present resource share in the Projects tab of the BOINC manager. The BOINC Wiki has more discussion of how resource share works.

---

Q: Why won't BOINC download more Rosetta Work units?

A: It is possible the servers are down for some reason (quite rare). You should check the server status.

But it is more likely you are seeing a message in the BOINC Manager saying "...not requesting new work". If so, then your PC is actually not asking for more work.

If you are running multiple BOINC projects, the BOINC manager plans out what will be crunched next, and balances goals like the deadlines of the work units, and your preferences as to resource share to devote to each project. In this balancing act, it sometimes crunches a project more than it's normal share in order to meet a deadline. When this occurs, the project being crunched incurs a "debt" to the other projects. BOINC will take this debt in to account when requesting new work. It is not uncommon, especially when running 3 or more projects, for BOINC to decide to crunch other projects for a period of a day or two in order to pay back such debts. And so during that time it will not order more work for projects that it does not plan to be crunching for a while.

In the end, if you leave things alone and resist the temptation to try and FORCE BOINC to do work for a specific project, it will work itself back in to balance and again download work for each project.

If you want to increase the resource share for Rosetta, you can increase the value in your Rosetta preferences, and then update to the project. You can view the present resource share in the Projects tab of the BOINC manager. The BOINC Wiki has more discussion of how resource share works.

How can I check my ranking/standings compared to other members?

---

Bullseye, just click the Statistics link on top of this message board page.

---

Rosetta Moderator: Mod.Sense

I started crunching Rosetta earlier this week on 3 of my computers. This morning I woke up and none of them were crunching. The server does not seem to want to send me any more work. What's going on? Is there a problem with the server today?

Here is my messages log:
5/12/2007 9:35:23 AM|rosetta@home|Requesting 17280 seconds of new work
5/12/2007 9:35:28 AM|rosetta@home|Scheduler RPC succeeded [server version 509]
5/12/2007 9:35:28 AM|rosetta@home|Deferring communication for 4 min 2 sec
5/12/2007 9:35:28 AM|rosetta@home|Reason: requested by project
5/12/2007 9:35:28 AM|rosetta@home|Deferring communication for 1 min 0 sec
5/12/2007 9:35:28 AM|rosetta@home|Reason: no work from project
5/12/2007 9:36:28 AM|rosetta@home|Fetching scheduler list
5/12/2007 9:36:33 AM|rosetta@home|Master file download succeeded
5/12/2007 9:36:38 AM|rosetta@home|Sending scheduler request: To fetch work
5/12/2007 9:36:38 AM|rosetta@home|Requesting 17280 seconds of new work
5/12/2007 9:36:43 AM|rosetta@home|Scheduler RPC succeeded [server version 509]
5/12/2007 9:36:43 AM|rosetta@home|Deferring communication for 4 min 2 sec
5/12/2007 9:36:43 AM|rosetta@home|Reason: requested by project

Any ideas?

you should post that question here
or
read this thread to keep up with whats happening.

basicly the scheduler has run out of work to send out due to some internal problem.

to make yourself safe against this kind of thing in the future you can make a couple of changes in your preferences to have your cpu run longer or the computer connect to the web every x days to build a queue of work for x day and x hours per work unit.

you can get more workunits in queue by clicking on participants at the top of the page here, then goto general preferences -> goto the bottom of the page and click on edit preferences -> then goto connect to network and change that value to x days other than 1 day. This will give you x days of work for your system to crunch.
The total number of work units downloaded will be determined by how long your cpu is told to crunch each workunit. If set at 8 hrs per work unit then you will get 6 work units for 2 days. Then you will have plenty of work to do and when there are errors in the system like this you still have work to process.

To set your cpu time per work unit, you can go back to the same menu as above and then click on Rosetta@home preferences. Scroll to the bottom of the page and click edit and then change your cpu time to x hrs - with x being greater than the value you have. As I said above, at 8 hrs cpu time you get 6 work units of 8 hrs crunch time per unit. Plenty of work then.

hope this helps.

I started crunching Rosetta earlier this week on 3 of my computers. This morning I woke up and none of them were crunching. The server does not seem to want to send me any more work. What's going on? Is there a problem with the server today?

Here is my messages log:
5/12/2007 9:35:23 AM|rosetta@home|Requesting 17280 seconds of new work
5/12/2007 9:35:28 AM|rosetta@home|Scheduler RPC succeeded [server version 509]
5/12/2007 9:35:28 AM|rosetta@home|Deferring communication for 4 min 2 sec
5/12/2007 9:35:28 AM|rosetta@home|Reason: requested by project
5/12/2007 9:35:28 AM|rosetta@home|Deferring communication for 1 min 0 sec
5/12/2007 9:35:28 AM|rosetta@home|Reason: no work from project
5/12/2007 9:36:28 AM|rosetta@home|Fetching scheduler list
5/12/2007 9:36:33 AM|rosetta@home|Master file download succeeded
5/12/2007 9:36:38 AM|rosetta@home|Sending scheduler request: To fetch work
5/12/2007 9:36:38 AM|rosetta@home|Requesting 17280 seconds of new work
5/12/2007 9:36:43 AM|rosetta@home|Scheduler RPC succeeded [server version 509]
5/12/2007 9:36:43 AM|rosetta@home|Deferring communication for 4 min 2 sec
5/12/2007 9:36:43 AM|rosetta@home|Reason: requested by project

Any ideas?

---

Can somebody please explain what keys on the keyboard change what about the graphics. I am interested in both a layman's and a scientific standpoint. I don't know why this info isn't in the graphics FAQ.

All I know is B and C do different stuff. Some of the numbers do the same. Please help. Thanks,

Ben

Q: How to merge together credits from two machines?

A: click the "Participants" link at the top of the page you are viewing now.
Click the "view" link beside "Computers on this account".
Click the computer ID in question and go to the bottom of the page.
Click the link "merge this computer".
Select the other machine that is identical with the one shown and click the "Merge hosts" button.

Now you have only one machine left, and it has the credits from both host IDs.

You cannot merge together data from two different user account numbers. Only two different hosts.

I recently had a minor failure in my main cruncher, which required a reinstall of BOINC to get it working again (on BOINC). This ended up creating a new set of CPIDs. I was able to merge the other four projects, and for some reason, Rosetta has identified the two instances of the same computer in a way that keeps it from allowing me to merge them: 522351 and 611519.

Though it's really not a big deal, I really would like to have them merged. Is there a way to force this (e.g. do it at the server)?

Thanks.

---

This question has been asked at least half dozen times this year. I'm posting here for everyone else to read.

Q: Why are graphics are disabled in Linux?

A: "We haven't added the graphics version because it would only work on newer linux versions due to graphics library dependencies and the fact that Rosetta will not build using older versions of gcc (3.3 and older) necessary for graphics on older linux versions." We hope to have graphics in the linux application after the new release of BOINC 6.0, possibly sooner if resources permit.

Quotation from David E K

Lloyd, the version of BOINC server code that is used by Rosetta is probably not as current as your other projects. Once it is updated, you will be able to merge your hosts.

---

Rosetta Moderator: Mod.Sense

Question regarding the RMSD and Accepted Energy. When getting to the native fold of the protein, is it better for these values to be closer to 0, or lower overall.

IE is RMSD better the closer it gets to 0? And is Accepted Energy better the closer it gets to 0, or the lower it gets overall? Just a few things I can't find an answer to being a newbie and all.

And as per Ben's post above...

"Can somebody please explain what keys on the keyboard change what about the graphics. I am interested in both a layman's and a scientific standpoint. I don't know why this info isn't in the graphics FAQ.

All I know is B and C do different stuff. Some of the numbers do the same. Please help."

I'd like to know as well, though only 'C' does anything for me in the view.

B changes the base strand look, from single strands to the view with flat bands and arrow ends to show the direction of that particular strand of the protien.
you can press this 3 times to get 3 different views.

S allows the side chains to be seen and changes their look as well.

C changes the color only of the molecular view if you have that on your screen by pressing the B or S buttons.

somone will need to correct my terminolgy here, but that is what those buttons do.

Question regarding the RMSD and Accepted Energy. When getting to the native fold of the protein, is it better for these values to be closer to 0, or lower overall.

IE is RMSD better the closer it gets to 0? And is Accepted Energy better the closer it gets to 0, or the lower it gets overall? Just a few things I can't find an answer to being a newbie and all.

And as per Ben's post above...

"Can somebody please explain what keys on the keyboard change what about the graphics. I am interested in both a layman's and a scientific standpoint. I don't know why this info isn't in the graphics FAQ.

All I know is B and C do different stuff. Some of the numbers do the same. Please help."

I'd like to know as well, though only 'C' does anything for me in the view.

---

RMSD (Root Mean Square Deviation, IIRC) is used to compare the generated structure to the native structure (if the native structure is known). A score of 0 for RMSD would be an exact match. The accepted energy would be as low as possible with a score of 0 for RMSD.

So, for RMSD you would want to get as close to 0 as possible and the lower the accepted energy is, the better.

A RMSD score can only be calculated if the native structure is already known.

Greg, Transient, thank you very much for the responses, answers everything I needed!

Is there anything special I need to do to report my finished work?

Is there anything special I need to do to report my finished work?

Nope.

The reporting is done per your Connection Interval. It hopes to report more than one at a time to reduce loads on the server.

However, If you go to the "projects" page, highlight "rosetta@home" (in the right hand box), then click "update" (in the left hand box), you can manually report them/it. This is called a "project update".

But no, there's nothing special you "need" to do.

hope this helps
tony

Is it possible to change the resource share percent for projects or otherwise prioritize projects? I'm currently working on Rosetta, Einstein, SETI, and World Community Grid with each at 25%. I think Einstein and SETI are important, but Rosetta and WCG have more short term benefits so I want to increase my CPU time on them. I know that I can suspend projects and that Boinc automatically works on projects based on the time required to get them done by the deadline, but is it possible, for example, to have Boinc download and work on Rosetta more than SETI without suspending or detaching SETI?

On a less important note, how do I get my statistics to show in my signature?

---

Thanks. You accidentally left out the "C" in boinCstats.com. I changed my resource share at http://boincstats.boincstats.com/bam/resources.php, but Einstein gave an incorrect response and WCG gave no response. So I went to http://einstein.phys.uwm.edu/home.php and changed Einstein's resource as you had explained for Rosetta. It worked. I can't figure out how to change the resource number for WCG though. Maybe I have it setup wrong or have to wait until I have some work done and get credit. Any ideas? Edit: I changed the resources so that I got the percents I wanted while leaving WCG at 100.

I attached the stats GIF you provided, but I know I have about twice the work done for Einstein. How long does it take for credit to be applied? I also have a World Community Grid project running. Do projects with zero credit not appear on the stats GIF? More to the point, is it a static image or does it update?

---