>>> Workunit 1379411

This work unit has been set "validate error" by all that have crunched it so far, after several days of CPU time each. Not happy about that.

---

Wave upon wave of demented avengers march cheerfully out of obscurity into the dream.

>>> Workunit 1379411

This work unit has been set "validate error" by all that have crunched it so far, after several days of CPU time each. Not happy about that.

This happens to all of us, and nobody likes it. The molecule your task was working on was probably unstable and the validate error confirms that.

It happened to me too.
https://quchempedia.univ-angers.fr/athome/workunit.php?wuid=1400377
Only once.
Probably they are isolated cases.

The validation is in two stages. Sometimes, in the first step (before comparison) an inconsistency can be detected.

All I see on today's tasks, both mine on Windows 10 and 4 or more wingmen on Linux end with validation errors.
Tullio

---

Just added a new linux rig, all 11x tasks errored out straight away. Strange things also happening with my other 2x linux rigs

Got one validated task on my Windows PC. My Linux wingman used a 40 cores CPU against my 6, yet my PC was faster.
Tullio

---

Me too. Tasks 2923275, 2923231, 2923499 and a few more, all with "validation error".
NWChem v0.11 (vbox64_t1) windows_x86_64.
Logfile looks good.

Still validation errors
4140131
4141252
etc...

I see many "validation inconclusive".
Tullio

---

I'm getting an increasing number of tasks with "Completed, can't validate". It's mainly affecting tasks with initial replication >=5 quorum 2, like these two. Is this perhaps a misconfiguration issue?

I thought a Ryzen 3900X might be better than the Intel i7-9700 I was using, in order to reduce the invalids and validation inconclusives, but it did not make much difference.
I will wait until they figure out what the problem is before jumping in again. Good luck.

Waiting for validation 40. Inconclusive validation 115.
Tullio

Waiting for validation 40. Inconclusive validation 115.
Tullio

We know that the app is in beta (see here: https://quchempedia.univ-angers.fr/athome/apps.php), but some feedback from admins will be great
Are they working on the problem? Are they working on a new app??

One additional point
The QuChem Windows Application has a problem with other vbox virtual machines running on the same host
As long as I don't have my Ubuntu Linux running in VBox on my computer the QuChem Windows App works fine
When I have the Ubuntu or Debian Linux running in VBox the QuChem Windows App stopps working with this error
QuChemPedIA@home 13.11.2020 20:04:31 Task cl9_athome_b3lyp-321gd,batch013,000139370,nwchem,1595083754_3 postponed for 86400 seconds: VM job unmanageable, restarting later.

This issue I don't see with a LHC Theory App e.g. "Theory_2390-********-57_1 using Theory version 30006 (vbox64_theory) in slot 5" and other vbox related Windows Apps and a running Ubuntu or Debian Linux VBox machine
So it looks for me like this is a QuChem related issue
And for me it looks like this is one reason for the many Validate error for the QuChem calculations (the QuChem vbox machine is stopped hard without writing a savestate file or using a normal shutdown via acpipowerbutton)
Without an other running vbox machine I get valid results, but with an other vbox machine the results are invalid

---

Matthias

I have noticed the same thing when running also LHC@home tasks alongside QuChem tasks on my Windows 10 PC with 6 processors. Usually it runs two QUChem tasks, the others waiting for memory. So I think it is a memory problem, I have 12 GB RAM on that PC. Maybe I should try running QuChem tasks on another PC which has 24 GB RAM but it is dedicated to running WCG and Rosetta@home tasks on Sars-CoVid'-2.
Tullio

---

on my computers I don't see the issue with not enough memory
My Computers have 16 GB RAM and there are always some GB RAM free
But someone mentioned in the forum that the required free RAM of 2 GB is much to high for the QuChem App

I too didn't see any RAM usage of this hight, Peak working set size is less 250 MB, but Peak swap size is up to 2 GB for my finished results in the last 20 results of my task list

@Tullio
And Boinc only uses a defined % (for me I found 80% configured in boinc is a good solution) of the installed RAM to have enough resources for the OS and other apps I use

---

Matthias

I had problems in Einstein@home when I used a GTX !060 Video board with 3 GB RAM. They scolded me and I had to buy another PC with a GTX 1650 which has 4 GB Video RAM. But QuChem does not use the GPU , so I can run Einstein GPU tasks along QuChem tasks.
Tullio

---

2020-11-24 04:42:26 (10128): Creating new snapshot for VM.
2020-11-24 04:42:32 (10128): Deleting stale snapshot.
2020-11-24 04:42:32 (10128): Checkpoint completed.
2020-11-24 04:51:56 (10128): Status Report: Trickle-Up Event[
2020-11-24 04:52:28 (10128): Creating new snapshot for VM.
2020-11-24 04:52:34 (10128): Deleting stale snapshot.

You see Trickle Up Event in side of 7 tasks at minima from mines
Inside of our universe we go to low energy system
I can not undestand inside of proteins an energy increasing system
May be a program bug or i can not understand because i have too few information

these events are pure system messages. It is not related to molecules. "Trickle-Up Event" are messages send to the server.