Help Take the Fight to COVID-19 with BOINC and Folding@home


Recommended Posts

2 hours ago, bastl said:

@jeffreywhunter Did you missed the point to create an account first? Top right on your screenshot shows you're not logged in.

Yes, I had created the account, but didn't see the error.  Just tried again, both on Edge and Chrome and getting a 'can't create account'.   Evidently they are having a problem with Captcha?

 

http://my.jetscreenshot.com/12412/20200318-v7xb-70kb.jpg

 

I called David Baker at the Molecular Engineering Lab to try to get some account support and left a message.  Guess I have to leave it at that for now...

Edited by jeffreywhunter
Link to comment
9 minutes ago, jeffreywhunter said:

Yes, I had created the account, but didn't see the error.  Just tried again, both on Edge and Chrome and getting a 'can't create account'.   Evidently they are having a problem with Captcha?

 

http://my.jetscreenshot.com/12412/20200318-v7xb-70kb.jpg

Try signing up from within BOINC.

 

Or use a BOINC account manager such as BAM! (yes that's what it's called) to create account for you from their website.

If you run multiple projects and/or on multiple clients BAM! helps simplify the add project process.

Link to comment

So while F@H will outright prioritize covid-19 work units (when they are available - they get snatched up pretty fast), boinc (rosetta@home) does not outright prioritize those covid projects.

 

Rather, it appears that any covid related projects wind up getting submitted via robetta@home.  Your server may (will) also work on other project which may or may not be related.  Although all of the projects in rosetta are probably of vital importance, if you do wish to prioritize the covid ones, you can go to advanced view, then tasks, and you can suspend tasks that are not given out from robetta.  While this still isn't a guarantee that you'll be getting something directly related to covid, at least you can reasonably guess that you're not going to be doing the math for someone who is researching foot fungus.

 

https://boinc.bakerlab.org/rosetta/forum_thread.php?id=13554&postid=91904#91904

Quote

Robetta WU's start with

rb_Month_Day_TaskID_DomainID_*

For example:

rb_03_07_17775_17595_*

 

That being said, I'm sure that ALL protein research could quite possibly be of use (if only in tertiary means)  for covid research.

 

Myself, I'm running both F@H and boinc simultaneously.  It should also be mentioned that boinc and F@H are NOT competitors, but rather complimentary and each is researching a different thing.  As I understand it, F@H researches the mechanism of how the proteins fold, whereas boinc researches the final result after folding.  Both researches are necessary in order to completely understand the mechanism behind nCovid-19 and an eventual cure / vaccine.

 

Unfortunately while F@H will automatically prioritize, boinc requires you to manually do it.

  • Thanks 1
Link to comment
9 minutes ago, martinf said:

I followed the guide. Installed Boinc, signed up at Rosetta, joined the unraid team. However, all I have is _Console...

Any thoughts, suggestions?

Cheers

Martin

The "regular" BOINC doesn't have a webUI. You have to manage it from another computer with the BOINC manager. RDP BOINC includes a virtual desktop with the BOINC manager in it.

Link to comment

Hey guys, just created the template for the new linuxserver boinc image, it should show up in CA soon. Here's the support thread: 

 

 

As I said before, this one is new. It contains both the client and the manager, so no need to install anything on a different machine.

 

It supports gpu (although rosetta@home is currently not handing out gpu jobs, some of the other projects do).

 

Also, this one is multi-arch, so feel free to put any raspberry pis or other iot devices to use as well 😉

 

@squid when this is listed in CA, feel free to deprecate my old RDP Boinc one

  • Like 2
  • Thanks 1
Link to comment
11 hours ago, aptalca said:

@squid when this is listed in CA, feel free to deprecate my old RDP Boinc one

i still use your docker app to control the other boinc images ;)

and its working fine, its does not doing boinc work itself, only controlling the other.

Link to comment

found i bug in the docker system.

i have 2 F@H docker images setup. the first one is setup to use the gt710, the second one is setup to use the 1050TI

but after create the second images it only sees the 710 (which is not assign to it)

then a change the config.xml:

  <slot id='0' type='CPU'/>
  <slot id='1' type='GPU'/>
  <slot id='2' type='GPU'/>

now it using the 1050ti and de 2nd image has NO work assign to the gt710, only the 1st one...

pretty weird that the 2nd image is seeing the GPU which is NOT setup to using it...

first image:

642117940_Schermafdrukvan2020-03-1911-37-07.png.d026a62e07970a4cca753d1f0fdf302e.png

 

second image:

573685818_Schermafdrukvan2020-03-1911-29-35.png.636024907847c13849a66872f460f3fe.png

 

at least 2 GPU's (of the 3 gpu installed) are working :D

Link to comment
9 minutes ago, jude said:

I set up RDP-Boinc following @SpaceInvaderOne youtube video and have had to shut it off as it was causing disks to spin up, and stay spun up. I have had it off now for 24 hours and zero disk spin ups.

 

Anyone else noticed this?

Where is your appdata located?

If it's on the array or if your cache is on a HDD or if you run out of cache space with Prefer setting etc. then it would certainly spin up your drives since it would require storage IO, thus spinning up drives.

 

My appdata is on SSD cache (cache = only) and no spin up issues,

Link to comment
6 minutes ago, testdasi said:

Where is your appdata located?

If it's on the array or if your cache is on a HDD or if you run out of cache space with Prefer setting etc. then it would certainly spin up your drives since it would require storage IO, thus spinning up drives.

 

My appdata is on SSD cache (cache = only) and no spin up issues,

appdata is located in /mnt/cache/appdata/RDP-Boinc and appdata is cache only. 

 

cache is an SSD has 120+GB free

 

Boinc was allocated 10GB of space on the cache drive

 

Still can't figure out what the issue is ........ 

Link to comment
4 hours ago, sjaak said:

i still use your docker app to control the other boinc images ;)

and its working fine, its does not doing boinc work itself, only controlling the other.

oh yeah, sure, you can use that manager to connect to other clients. I hadn't thought of that.

Link to comment
53 minutes ago, jude said:

appdata is located in /mnt/cache/appdata/RDP-Boinc and appdata is cache only. 

 

cache is an SSD has 120+GB free

 

Boinc was allocated 10GB of space on the cache drive

 

Still can't figure out what the issue is ........ 

Another possibility is your docker image is on the array. Post your diagnostics if you need someone else to take a look.

Link to comment
On 3/18/2020 at 6:20 AM, testdasi said:

That sounds to be like an innate instability somewhere in the system. BOINC and f@h put unusually high load on the hardware so any instability is more likely to manifest itself.

I resolved my lockup issue with BOINC.  It has now been running 15+ hours with no problems.

 

I noticed that right before the last crash, the MB temp had risen from 33C to 78C.  Cooling was obviously an issue.  Under normal loads I had never seen it higher than 34C.

 

Through the IPMI Support plugin, I had set my side intake fans (which blow air from side to side through the drive cages) to react to HDD temps.  In my brilliance, I thought, if all the HDDs are spun down, I certainly don't need these fans running at full RPM so I had them dialed down to 200 RPM with no active hard drives.  If one hit 35C they ramped up to 700 RPM and at 40C they kicked in on full at 1400 RPM.  Problem is that this air flow once passed through the drive cages, flows back to the motherboard.  I had essentially cut off most cooling to the MB when the HDDs were spun down.

 

As long as something else was going on on the server that spun up an HDD or two, the system was happy.  As soon as they all spun down, it was just a matter of time before a lockup under a sustained heavy load (such as BOINC) because the MB cooling (except for a bottom intake fan) had been shut off.

Edited by Hoopster
Link to comment
24 minutes ago, Hoopster said:

I resolved my lockup issue with BOINC.  It has now been running 15+ hours with no problems.

 

I noticed that right before the last crash, the MB temp had risen from 33C to 78C.  Cooling was obviously an issue.  Under normal loads I had never seen it higher than 34C.

 

Through the IPMI Support plugin, I had set my side intake fans (which blows air from side to side through the drive cages) to react to HDDs temps.  In my brilliance, I thought, if all the HDDs are spun down, I certainly don't need these fans running at full RPM so I had them dialed down to 200 RPM with no active hard drives.  If one hit 35C they ramped up to 700 RPM and at 40C they kicked in on full at 1400 RPM.  Problem is that this air flow once passed through the drive cages, flows back to the motherboard.  I had essentially cut off most cooling to the MB when the HDDs were spun down.

 

As long as something else was going on on the server that spun up an HDD or two, the system was happy.  As soon as they all spun down, it was just a matter of time before a lockup under a sustained heavy load (such as BOINC) because the MB cooling (except for a bottom intake fan) has been shut off.

Oh wow, that is very specific!

Link to comment
55 minutes ago, Hoopster said:

I resolved my lockup issue with BOINC.  It has now been running 15+ hours with no problems.

 

I noticed that right before the last crash, the MB temp had risen from 33C to 78C.  Cooling was obviously an issue.  Under normal loads I had never seen it higher than 34C.

 

Through the IPMI Support plugin, I had set my side intake fans (which blows air from side to side through the drive cages) to react to HDDs temps.  In my brilliance, I thought, if all the HDDs are spun down, I certainly don't need these fans running at full RPM so I had them dialed down to 200 RPM with no active hard drives.  If one hit 35C they ramped up to 700 RPM and at 40C they kicked in on full at 1400 RPM.  Problem is that this air flow once passed through the drive cages, flows back to the motherboard.  I had essentially cut off most cooling to the MB when the HDDs were spun down.

 

As long as something else was going on on the server that spun up an HDD or two, the system was happy.  As soon as they all spun down, it was just a matter of time before a lockup under a sustained heavy load (such as BOINC) because the MB cooling (except for a bottom intake fan) has been shut off.

Thanks for this, I have this same setup. But I have torrents running most of the time, so it's rare for drives to be spun down.

 

I just changed it for both sets of fans to depend on the CPU temperature, is this a good option? I've noticed that most of the time hard drives won't get too hot, even when I'm transferring a lot, or during a parity check. So this should be safe?

Link to comment
3 minutes ago, cherrybullet said:

I just changed it for both sets of fans to depend on the CPU temperature, is this a good option?

My CPU fan reacting to the CPU temps was enough to keep the CPU cool.  The potential problem with having case intake fans react to CPU temps is that, depending on the thresholds, your CPU temp may be relatively low while the MB temps are climbing if the motherboard is not being cooled properly; although, I can't really see a scenario where the MB temp is rising too much unless the CPU and/or PCH is getting a pretty good workout.

 

I have noticed that controlling the fans based on MB temp is also an option with the IPMI plugin.

 

image.png.db460f348fb437ff10cd41ed560a888b.png

 

I am going to experiment with the various temperature sensors that it can monitor and find a combination of temps and speeds that will work in all cases.  Of course, that will likely mean increasing the RPM of the fans to at least 50% on the low end.  My Noctua fans are very quiet so, I really did not need them dialed down to 200 RPM anyway.

 

If your HDDs don't get too hot, even during a parity check, you are probably fine to tie the fan performance to some other temperature measure.  In my system, the Parity drive can get up to 45C during a parity check, but the data drives are usually a few degrees lower than that.  They get closer to 45C in the Summer because of ambient temps.

Link to comment

I want to do this so bad but I'm having trouble with the basic settup. The RTP Boinc docker Space Invader One used in his example is not in plugins and I cant figure out how to configure the one that is and its very different and has no GUI :(. Can anyone help me get this up and running?

 

Edited by PilotReelMedia
Link to comment

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Restore formatting

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.