Validate error - What this really means!

Gary Roberts
Gary Roberts
Moderator
Joined: 9 Feb 05
Posts: 5851
Credit: 110605781231
RAC: 32867177

RE: My machine has one

Quote:
My machine has one validate error,
task 550879678
All of the rest seem fine so far.


Hi Jill,

Welcome to the Einstein@Home project. That's a very nice looking machine you are using. Thank you for your contribution.

The data for the BRP4G run that you are crunching on your GPU comes from the Arecibo radio telescope and occasionally some parts of it contain RFI (radio frequency interference) which can cause these problems. An attempt is made to filter out any data containing RFI but sometimes a bit gets through. The other data also crunching on the GPU is BRP6 data that comes from the Parkes radio telescope in Australia. It never seems to contain RFI.

At the moment you have two BRP4 tasks showing validate errors. We can tell that the problem is not to do with your machine because everybody is having the same problem with these two tasks. If you click on the work unit link for each task in the above list, you will see the other copies of each task being crunched by other computers and you will see everybody is having the same problem.

The Devs should notice this shortly and remove the offending work units so that they don't keep getting sent out and failing. The limit is 20 failures like this before the system does an automatic cancel.

Cheers,
Gary.

Greg
Greg
Joined: 10 Mar 05
Posts: 9
Credit: 116663922
RAC: 0

RE: I'm getting a lot of

Quote:

I'm getting a lot of validate errors with BRP4G-Beta-opencl-ati on two GPUs that I know are stable.

https://einsteinathome.org/host/4280688
https://einsteinathome.org/host/12214325

On the same GPUs, all BRP6-opencl-ati wus validate ok. I have been using a GPU utilization factor of 0.5. I will change this to 1 and see if it helps.

To update this, there's definitely an issue with BRP4G-Beta-opencl-ati version 1.52 and GPU utilization of 0.5. Increasing utilization to 1 so that only one instance is running per GPU gives no errors. Likewise, both BRP4G-opencl-ati 1.39 and BRP6-opencl-ati work fine with utilization of 0.5 or 1.

Peanuckle
Peanuckle
Joined: 28 Feb 07
Posts: 1
Credit: 2224800
RAC: 0

I noticed a handful of

I noticed a handful of validation errors and errors while computing on *I think* Gravitational Wave search O1 all-sky F but those have all since be removed(and I stopped accepting tasks from that project). Now Gamma-ray pulsar binary search #1 https://einsteinathome.org/account/tasks&offset=0&show_names=0&state=0&appid=32 is starting to error a bit.

Is it my computer or is this a WU problem? Should I stop accepting these tasks? Should I start accepting Gravitational Wave search O1 all-sky F again? Thanks.

Holmis
Joined: 4 Jan 05
Posts: 1118
Credit: 1055935564
RAC: 0

Hi Peanuckle and welcome to

Hi Peanuckle and welcome to the message boards!

The link you posted is only accessible to you so here's one that works for everyone else: https://einsteinathome.org/host/12006005/tasks.

That computer has reported 1 validate error and 4 error while computing.
As others have managed to complete the tasks then it's safe to say that it's a problem with your computer.

I would start by checking the ventilation and cooling and clear out any dust.
If that doesn't help I would continue with backing of any overclocking done and then start examining the power supply and test the RAM.

If you want more help troubleshooting your problem you should start a new thread here in "Problems and Bug Reports" as this thread is only for "Validate errors".

Alexander Favorsky
Alexander Favorsky
Joined: 18 Jun 16
Posts: 36
Credit: 160321615
RAC: 66055

Hi everyone!These 2 WUs

Hi everyone!

These 2 WUs have 4 validate errors, 1 in progress and 1 completed:

https://einsteinathome.org/workunit/252681864

https://einsteinathome.org/workunit/252686261

Is there a problem with these WUs or are we just the 4 lucky ones having validate errors?

Holmis
Joined: 4 Jan 05
Posts: 1118
Credit: 1055935564
RAC: 0

When more than 1 or 2 tasks

When more than 1 or 2 tasks in a WU end up as validate error it's usually a sign that something is wrong with the data for that WU. The admins usually cancels these WU when they notice them and I assume that there are automatic warning systems in place to notify the admins when a WU reaches a certain number of validate errors.

Christian Beer
Christian Beer
Joined: 9 Feb 05
Posts: 595
Credit: 131890402
RAC: 276895

Holmis' explanation is

Holmis' explanation is correct. We received the "bad beams" notice on Friday and I just canceled those.

Betreger
Betreger
Joined: 25 Feb 05
Posts: 987
Credit: 1447628358
RAC: 674358

Lots of validate errors today

Lots of validate errors today on BRP4G, sadly they run full time so much crunching wasted.

Christian Beer
Christian Beer
Joined: 9 Feb 05
Posts: 595
Credit: 131890402
RAC: 276895

Betreger wrote:Lots of

Betreger wrote:
Lots of validate errors today on BRP4G, sadly they run full time so much crunching wasted.

There are a lot of bad beams recently. I'm shortly looking at them to cancel the most problematic ones.

northcup
northcup
Joined: 23 Feb 05
Posts: 3
Credit: 50842601
RAC: 21581

[03:29:40][23252][INFO ] Data

[03:29:40][23252][INFO ] Data processing finished successfully!
03:29:40 (23252): called boinc_finish(0)

see below with many other wu - all failed when coding finished. Please check - thankyou

264840326

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.