Boinc Pentathlon

[PDW]

Well 1 of them has completed and validated

But others have this error...

<core_client_version>7.0.65</core_client_version>
<![CDATA[
<message>
process exited with code 67 (0x43, -189)
</message>
<stderr_txt>
Deprecation warning: healthSystem: use of ImmediateOutcomes can be replaced by the more flexible DecisionTree5Day (optional)
Deprecation warning: healthSystem: use of ImmediateOutcomes can be replaced by the more flexible DecisionTree5Day (optional)
Checkpoint error: stream read error binary (pos -1 of -1 bytes)
Call stack:
+0x19 OM::util::checkpoint_error::checkpoint_error(std::string const&)
+0x5aa OM::Simulator::checkpoint(std::istream&, int)
+0x7a2 OM::Simulator::readCheckpoint()
+0xad OM::Simulator::start(scnXml::Monitoring const&)
+0x137 main()
+0xed __libc_start_main()
()
21:38:36 (1953): called boinc_finish

</stderr_txt>

A wild guess would have made me think it might be related to your libc libraries as they did have some problems with those last year. BUT you have had 1 validate so I don't see how it could be that

 

[MGP]

I don't know my way around linux so as it doesn't want to work "out of the box" I have no idea where to look. The client is even ignoring the standard text commands to stop, boinc manager hangs trying to connect to the client so I have great difficulty even finding out what config is thought to be ignored. Who knows how many times I've rebooted - usually a hard power off as nothing else responds.

To be honest if it wasn't for this machine being so beep heavy it would have grown wings and be in the middle of the street by now.

 

[PDW]

Additionally, the details for the machine says this...

openMalaria: A simulator of malaria epidemiology and control (Branch A) 6.73 x86_64-pc-linux-gnu
Number of tasks completed 1
Max tasks per day 50
Number of tasks today 0
Consecutive valid tasks 1
Average processing rate 1.6555743187312
Average turnaround time 0.11 days

openMalaria: A simulator of malaria epidemiology and control (Branch A) 6.73 i686-pc-linux-gnu
Number of tasks completed 0
Max tasks per day 100
Number of tasks today 58
Consecutive valid tasks 0
Average processing rate
Average turnaround time 0.00 days

Which is a bit confusing.
You've had 1 valid task for the first app version (x86_64-pc-linux-gnu) though it says you haven't had any tasks today !
You've had 58 of the second app (i686-pc-linux-gnu) but none have completed !

I don't know enough about this stuff to know why you would get both types on the same machine

 

[PDW]

To be honest if it wasn't for this machine being so beep heavy it would have grown wings and be in the middle of the street by now.

I've set up and used linux but I tend not to play with it too much.

Will it do WCG or even YoYo before you go away ?

 

[MGP]

At the moment it will run absolutely nothing as I can't get into boinc enough to even get it to stop looping on the malaria stuff. I have no idea what it is trying to do and have no choice but to give up on a bad job and consign this thing to the museum piece it became one folding gave up on BigADV WUs.

 

[PDW]

If you disconnect it from the internet will it not crash its way through the remaining tasks until it runs out of anything to do ? Or do the tasks never abort unless you power down, in which case it will be quite a few more reboots.

 

[PDW]

Meant to add that I think the i686 app is for 32 bit and the other app is for 64 bit.

 

[andy_taximan]

never having used linux all i can say is,, why not use windows? lol

 

[hanluc]

MGP, have you tried running it on WCG?

Oh I've just seen post #85 the only thing I can suggest is to update boinc see that does anything, I know the linux boinc package in the repository is usually an older version. You could do a direct download from boinc and install it manually.

 

[MGP]

Numerous hard rebooting and perseverance and I managed to corrupt sufficient malaria WUs that with about 5 left it would run. Perhaps there was something with the project that had too high a disk write access that it just couldn't cope with numerous WUs at a time. Overnight those completed. Malaria is now deleted never to darken my doors again.

I'm now trying to get wgc running on the rig. So far it seems to be happy .....

Oh and on a 4P rig linux is the only option. Unless you have a server edition Windows can't recognise / use more than two cpu cores.

I was actually using a customised boinc installation package on a custom ubunto that had been produced by the clever people over at [H] Hard Forum. They'd done a lot of work with multi cpu rigs to optimise stuff and their folding setups were great. As I said it had worked fine with the Pogs challenge, so I can only assume there was something about malaria that found a new problem.

 

[andy_taximan]

Typical eh
07/05/2015 20:36:22 | malariacontrol.net | No work is available for malariacontrol.net
07/05/2015 20:36:22 | malariacontrol.net | No work is available for openMalaria test version
07/05/2015 20:36:22 | malariacontrol.net | No work is available for Prediction of Malaria Prevalence
07/05/2015 20:36:22 | malariacontrol.net | No work is available for Estimation of parameters of infection dynamics (variable duration, max 4h)
07/05/2015 20:36:22 | malariacontrol.net | No work is available for openMalaria: A simulator of malaria epidemiology and control (Branch B)
07/05/2015 20:36:37 | malariacontrol.net | Sending scheduler request: To fetch work.
07/05/2015 20:36:37 | malariacontrol.net | Requesting new tasks for NVIDIA GPU
07/05/2015 20:36:38 | malariacontrol.net | Scheduler request completed: got 0 new tasks
07/05/2015 20:36:38 | malariacontrol.net | No work sent
07/05/2015 20:36:38 | malariacontrol.net | No work is available for malariacontrol.net
07/05/2015 20:36:38 | malariacontrol.net | No work is available for openMalaria test version
07/05/2015 20:36:38 | malariacontrol.net | No work is available for Prediction of Malaria Prevalence
07/05/2015 20:36:38 | malariacontrol.net | No work is available for Estimation of parameters of infection dynamics (variable duration, max 4h)
07/05/2015 20:36:38 | malariacontrol.net | No work is available for openMalaria: A simulator of malaria epidemiology and control (Branch B)

 

[Voicon]

Maybe the wrong choice for the challenge then.

 

[MGP]

Malaria is CPU yet that log suggests you are trying for GPU

But Malaria seems to be a very unstable BOINC project. After all the problems with my linux 4P rig, now running WGC quite normally, I've tried boinc on my windows 8.1 i3-4030U lappy and am already seeing computing errors and aborted WU errors (I haven't stopped anything yet).

The project is probably also only used to a certain amount of use. I suspect there has been a rush of unexpected demand so they haven't got enough WUs to go around.

 

[hanluc]

I'm getting work, perhaps just a temporary shortage. Sever status is currently showing nearly 2000 tasks ready to send.

Reminder again not to set your job cache too high, 3 days works good for me.

 

[MGP]

Our Malaria output isn't cutting it and we are on the slide

My pending validation queue is getting a bit large on both malaria and wgc so there is probably a lot of stashing going on somewhere. It's probably the same for everyone. Perhaps once Einstein is a live project people will release their queues for everything as you can't easily block just the one project. Saturday could be interesting?

But as a whole OcUK needs a lot more power than we have available. This is a challenge where there are many many big hitting teams and with only a few of us playing it's not looking good

 

[hanluc]

We have a chance to pull back a spot on malaria, as we are slowly gaining on 20th, but I agree we are too few in number to make an impact. We are however doing OK on WCG, 19th spot isn't in danger and the last 24 hours we've out produced the 3 other teams immediately above us.

 

[PDW]

We've moved up one place in Malaria

Definitely closing the gap on a couple above in WCG and that has a long way to go yet so with a bit more each day we can pass them and start chasing EVGA

 

[MGP]

So Einstein starts tomorrow. That is likely to shake things up.

Meanwhile we lost an overall place today and have slipped to #20

The good news is that for WGC over the last 24 hours we comfortably have the 15th highest output. If we can hold that then we might claim a few scalps at least for the marathon challenge, including our old foe EVGA And that could make a difference to our overall standing.

But malaria isn't going to get any better. I think we are just about safe from those behind, but it could be a close thing.

So what will Einstein do to us?

I'm not going to add anything else to my crunchers so I'll only be contributing to WGC, Malaria and Einstein. I'm going to leave it to others to worry about the YoYo thing and if anything else comes up.

As I'm away from next week most crunchers will go "not on" as some are borgs and others will not be in reach of interweb. That means, just as things get interesting I might be found wanting

At present I have:
i7 2600K with GTX970 on WGC and Einstein
AMD 4P rig (no idea what CPUs) 64 cores on WGC
i3 laptop on Malaria
i5 with basic nvida GPU borg on WGC and Einstein
i5 borg on Malaria

It's getting warm sitting here

 

[MGP]

8th place giving 54 points for Einstein That's not going to last once the effects of stashing dissipate.

 

[andy_taximan]

7th now im doing my best lol