New task batches available

Message boards : News : New task batches available
Message board moderation

To post messages, you must log in.

1 · 2 · Next

AuthorMessage
e1profesor
Project administrator
Project developer
Project scientist

Send message
Joined: 8 Jan 21
Posts: 52
Credit: 179
RAC: 0
Message 23 - Posted: 13 Jan 2021, 16:35:12 UTC

We've just fixed an issue in our work generation script, and we're happy to confirm that there are new task batches on their way.
The issue originated as a direct result of unintentional cancelation of a whole lot of workunits.
We confirm that the database integrity has been preserved and no user credit, nor result data has been affected.

We offer our sincere apologies for any inconvenience this may have caused.
ID: 23 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Contact
Avatar

Send message
Joined: 10 Jan 21
Posts: 35
Credit: 36,573
RAC: 53
Message 24 - Posted: 13 Jan 2021, 17:39:31 UTC

1 task per core runs at a time so I get this status on all wu's unless I suspend all but one wu.
VM job unmanagable, restarting later

If Network activity is suspended, BOINC Manager displays
Waiting for network access
and processing of wu's is paused
ID: 24 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Sergey Kovalchuk

Send message
Joined: 11 Jan 21
Posts: 4
Credit: 8,427
RAC: 0
Message 31 - Posted: 14 Jan 2021, 8:33:18 UTC - in response to Message 23.  

there is something wrong with the server, please check the notification
Transitioner backlog (hours) 39.30 

maybe this service needs to be kicked (restarted)
ID: 31 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 32 - Posted: 14 Jan 2021, 9:27:51 UTC - in response to Message 24.  
Last modified: 14 Jan 2021, 9:35:15 UTC

1 task per core runs at a time so I get this status on all wu's unless I suspend all but one wu.
VM job unmanagable, restarting later



Getting same error and after pausing all but 1 it has persisted furthest to 23%

They made explicit statements about not pausing or aborting the WU's as order is important to the work.

I hope this isn't intended behavior.

EDIT:

Added app_config.xml to the project folder with contents:
<app_config>

<project_max_concurrent>1</project_max_concurrent>

</app_config>

then reread configuration files fOINC Manager BUT that fails to help and the postponed WU's continue.

I'm going to go to sleep and see what happens once they are all postponed.
ID: 32 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Contact
Avatar

Send message
Joined: 10 Jan 21
Posts: 35
Credit: 36,573
RAC: 53
Message 49 - Posted: 15 Jan 2021, 12:40:53 UTC - in response to Message 32.  
Last modified: 15 Jan 2021, 12:43:46 UTC

You can also lower CPU usage %. Try lowering Use at most __ % of CPU's until only 1 task runs at a time.
I have been able to restart paused wu's by exiting BOINC Manager (Stop running tasks) and then re-starting the Manager.
ID: 49 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Sergey Kovalchuk

Send message
Joined: 11 Jan 21
Posts: 4
Credit: 8,427
RAC: 0
Message 53 - Posted: 15 Jan 2021, 14:16:15 UTC - in response to Message 49.  

After restarting the Manager, the postponed tasks resume processing. But result is always invalid.
You can immediately delete the postponed task and not waste time

After past problems with uploading, I have one task stuck waiting for validation - more than 36 hours
ID: 53 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 54 - Posted: 15 Jan 2021, 14:17:03 UTC

These errors appear related to VBox version 6.1.xx.

However, when I reduced the VBoxSVC,exe of ver 6.1.16 priority to IDLE using Process Hacker; the number of postponed WU's is significantly down.

The other machine with Windows 10 and VBox ver 5.2.44 is having no issues.
ID: 54 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 55 - Posted: 15 Jan 2021, 14:19:04 UTC - in response to Message 53.  

After restarting the Manager, the postponed tasks resume processing. But result is always invalid.



Actually, 1 of my postponed are valid of a total of 7.

Though if that rate of invalids continues after 6 hours; I'll delete the remaining 90+ WU's.
ID: 55 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 66 - Posted: 15 Jan 2021, 20:01:22 UTC - in response to Message 55.  

After restarting the Manager, the postponed tasks resume processing. But result is always invalid.



Actually, 1 of my postponed are valid of a total of 7.

Though if that rate of invalids continues after 6 hours; I'll delete the remaining 90+ WU's.


So 1 valid and 14 invalid... definitely abort them if they end up postponed.
ID: 66 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
zombie67 [MM]
Avatar

Send message
Joined: 12 Jan 21
Posts: 3
Credit: 1,035,072
RAC: 1
Message 70 - Posted: 16 Jan 2021, 3:31:50 UTC

FWIW, the "VM job unmanagable, restarting later" issue is unique to only windows machines. It doesn't happen on linux or macOS.
Reno, NV
Team: SETI.USA
ID: 70 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Frank [RKN]

Send message
Joined: 14 Jan 21
Posts: 2
Credit: 4,016
RAC: 0
Message 71 - Posted: 16 Jan 2021, 8:55:16 UTC - in response to Message 32.  


Added app_config.xml to the project folder with contents:

I had an app_config for 2 wu's, but it didn't work. After the 2 wu's stoppen with "VM job unmanagable, restarting later" the BoincManager starts the next 2 wu's, today morning all my wu's stopped.
And i couldn't restart them, maybe because my machine went out of ram (15.8 of 16 gb full).
After killing the wu's windows shows me that more than 10 gb ram are still in use, even after quit the BoincManager.
Only a reboot of this machine free the ram to normal ~ 3 gb in use, so after "VM job unmanagable, restarting later" you should check your ram in the TaskManager...

I'm looking forward to get an app except vbox ;-)
ID: 71 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 84 - Posted: 16 Jan 2021, 21:31:31 UTC - in response to Message 71.  



I'm looking forward to get an app except vbox ;-)


It needs to be repeated that you can successfully run the app now on Windows 10 using Virtual Box 5.2.44

My 2700X on Windows X (with Cortona and other spy ware ripped out) is doing fine.
ID: 84 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Frank [RKN]

Send message
Joined: 14 Jan 21
Posts: 2
Credit: 4,016
RAC: 0
Message 88 - Posted: 16 Jan 2021, 22:23:42 UTC - in response to Message 84.  


I'm looking forward to get an app except vbox ;-)

It needs to be repeated that you can successfully run the app now on Windows 10 using Virtual Box 5.2.44

If i find any bugs at my new car, i don't think about to sell it and buy a horse ;-)

I know, vbox causes a lot of issues, but downgrade is only a workaround, not the solution.
Just my opinion....
ID: 88 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 89 - Posted: 16 Jan 2021, 23:46:48 UTC - in response to Message 88.  
Last modified: 17 Jan 2021, 0:23:36 UTC


I'm looking forward to get an app except vbox ;-)

It needs to be repeated that you can successfully run the app now on Windows 10 using Virtual Box 5.2.44

If i find any bugs at my new car, i don't think about to sell it and buy a horse ;-)

I know, vbox causes a lot of issues, but downgrade is only a workaround, not the solution.
Just my opinion....



Very, very poor analogy.

A better analogy would be buying the same model car from about 2 years ago since VBox 5.2.xx was released October 2017 and version 5.2.44 was released in July 2020.

There is absolutely no issues running LHC on VBox 5.2.44 (confirmed by reported WU's over at LHC today) or on these WU's since inside these WU's are basic, streamlined, older versions of Linux doing just mathematics calculations.

We are not running games inside these VM's needing access to high end GPU's or realtime access to even the mouse.

EDIT:
The product being used inside Kryptos@Home WU's is deprecated Boot2Docker which was not designed for VBox 6.x. It was fully designed to run under VBox 5.x versions and there is a warning at GitHub:

https://github.com/boot2docker/boot2docker
"Boot2Docker is officially deprecated and unmaintained. It is recommended that users transition from Boot2Docker over to Docker Desktop instead (especially with the new WSL2 backend, which supports Windows 10 Home)."


So, knowing that Boot2Docker is the product in these WU's it makes perfect sense to use the version of VBox that it was designed to work with which is VBox 5.2.x...
ID: 89 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Peter Hucker

Send message
Joined: 13 Jan 21
Posts: 54
Credit: 1,225,444
RAC: 69
Message 95 - Posted: 17 Jan 2021, 10:28:25 UTC - in response to Message 89.  
Last modified: 17 Jan 2021, 10:46:53 UTC

If i find any bugs at my new car, i don't think about to sell it and buy a horse ;-)
Very, very poor analogy.

A better analogy would be buying the same model car from about 2 years ago since VBox 5.2.xx was released October 2017 and version 5.2.44 was released in July 2020.
No, since cars don't change so quickly as programs. A 2 year old car is compatible with every road in the country. The old Vbox is not compatible with every program.

There is absolutely no issues running LHC on VBox 5.2.44 (confirmed by reported WU's over at LHC today) or on these WU's since inside these WU's are basic, streamlined, older versions of Linux doing just mathematics calculations.
Not what they told me.

The product being used inside Kryptos@Home WU's is deprecated Boot2Docker which was not designed for VBox 6.x. It was fully designed to run under VBox 5.x versions and there is a warning at GitHub:

https://github.com/boot2docker/boot2docker
"Boot2Docker is officially deprecated and unmaintained. It is recommended that users transition from Boot2Docker over to Docker Desktop instead (especially with the new WSL2 backend, which supports Windows 10 Home)."
So this project knowingly used an out of date product?

So, knowing that Boot2Docker is the product in these WU's it makes perfect sense to use the version of VBox that it was designed to work with which is VBox 5.2.x...
Wrong. Everyone else (including LHC@home which I use) requires the later version. This project made a new app which requires something outdated. This project is at fault and needs to pay the license for the new Vbox compiler, or whatever caused them to be out of date right at the start. I will not be downgrading Vbox and breaking everything else. Until then, this project can only get 1 core from each of my machines. Awaiting re-programming at the project end....
ID: 95 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 126 - Posted: 18 Jan 2021, 12:15:49 UTC - in response to Message 95.  

There is absolutely no issues running LHC on VBox 5.2.44 (confirmed by reported WU's over at LHC today) or on these WU's since inside these WU's are basic, streamlined, older versions of Linux doing just mathematics calculations.
Not what they told me


I'm on the same forum thread over there as you.
They never said their WU's do not work on VBox 5.2.xx. They said they prefer we upgrade. That is not the same as declaring it will not work.
Indeed, several users came into the thread and said it was fine.
I ran an ATLAS on Sunday on VBox 5.2.44 with no issues on Windows 10.

The product being used inside Kryptos@Home WU's is deprecated Boot2Docker which was not designed for VBox 6.x. It was fully designed to run under VBox 5.x versions and there is a warning at GitHub:

https://github.com/boot2docker/boot2docker
"Boot2Docker is officially deprecated and unmaintained. It is recommended that users transition from Boot2Docker over to Docker Desktop instead (especially with the new WSL2 backend, which supports Windows 10 Home)."
So this project knowingly used an out of date product?


You do not know they did it "knowingly" since Boinc2Docker (the turn-key project based on Boot2Docker) doesn't mention the deprecated state.



So, knowing that Boot2Docker is the product in these WU's it makes perfect sense to use the version of VBox that it was designed to work with which is VBox 5.2.x...
Wrong. Everyone else (including LHC@home which I use) requires the later version. This project made a new app which requires something outdated. This project is at fault and needs to pay the license for the new Vbox compiler, or whatever caused them to be out of date right at the start. I will not be downgrading Vbox and breaking everything else. Until then, this project can only get 1 core from each of my machines. Awaiting re-programming at the project end....


You appear to be knowingly spreading false information
I've been on every project that uses VBox over the last 5 years (check my WUProps or Free-DC history).
Cosmology@Home uses the same product, so did TACC. Two other projects were based on Boot2Docker in the last 3 years.
LHC@Home WU's still work fine with 5.2.44 and you are lying about the downgrade effecting LHC@Home WU's

You insulted these devs enough; if you are not happy then walk and stop bitching.
ID: 126 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Peter Hucker

Send message
Joined: 13 Jan 21
Posts: 54
Credit: 1,225,444
RAC: 69
Message 131 - Posted: 18 Jan 2021, 12:40:23 UTC - in response to Message 126.  

I'm on the same forum thread over there as you.
They never said their WU's do not work on VBox 5.2.xx. They said they prefer we upgrade. That is not the same as declaring it will not work.
Indeed, several users came into the thread and said it was fine.
I ran an ATLAS on Sunday on VBox 5.2.44 with no issues on Windows 10.
Maybe it works now, but maybe it will mess up in the future, maybe there are many more things people use VB for, even outside of Boinc. This, from an LHC moderator: "Hence, it is not a nice behaviour to launch a new project now that forces already outdated software components." which you agreed with and admitted this project screwed up. Again, there is no reason to write a new program which doesn't function on the latest version of something. Requiring users to downgrade is absurd and nothing less. When this project realised it wouldn't work on the latest version, they should have gone back to the drawing board. Would you release a windows program that required everyone to go back to Windows 7? Sales would plummet.

you are lying about the downgrade effecting LHC@Home WU's
I asked LHC if they minded me downgrading and they said it was a bad idea.

You do not know they did it "knowingly" since Boinc2Docker (the turn-key project based on Boot2Docker) doesn't mention the deprecated state.
I guess they should have RTFM.

You insulted these devs enough; if you are not happy then walk and stop bitching.
I have spent hours messing about getting this project to work, I'm trying to help, despite this being the most awkward project to work with. Believe me, I have not been insulting, I have held back in many places. There will be close to 90 cores running this project now, although I've had to remove 2 of 6 from an i5 which causes slowdown of the GPU, and 12 of 24 from a Ryzen which is my main machine and slows the interface. But that's VB's/AMD's bad design.
ID: 131 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
marmot
Volunteer moderator

Send message
Joined: 13 Jan 21
Posts: 46
Credit: 233,819
RAC: 16
Message 146 - Posted: 19 Jan 2021, 14:13:14 UTC - in response to Message 131.  
Last modified: 19 Jan 2021, 14:20:37 UTC

which you agreed with and admitted this project screwed up.
You really need to stop.
I did not say 'screwed up'.
I defended their actions.


Again, there is no reason to write a new program which doesn't function on the latest version of something.
The devs here used a turn-key solution and did not write it. This project has a limited scope to solve a single cryptographic puzzle and it would be a waste of time and labor to write from scratch.


Requiring users to downgrade is absurd and nothing less.
This is science project for BOINC and not gaming environment. I've seen many requests to downgrade and even the BOINC version of VBox is locked into an older version and warns against upgrading to the newest in the instructions.


When this project realised it wouldn't work on the latest version, they should have gone back to the drawing board.
They did go back and to setup new WU's and announced them yesterday and have stopped these so... why do you suggest this?

Would you release a windows program that required everyone to go back to Windows 7? Sales would plummet.
Non-sequitor. We are in a research environment, not sales. There is plenty of processing power available for BOINC and they do not have to worry about angering a few petty individuals.

I asked LHC if they minded me downgrading and they said it was a bad idea.
They said it was not preferred but the LHC WU's work fine in 5.2.44. It is not their project. Their opinion does not matter here.

I guess they should have RTFM.
You do not seem to understand what abusive behavior is. When you tell someone to "read the fucking manual", that is vulgar and abusive behavior.

I have spent hours messing about getting this project to work, I'm trying to help, despite this being the most awkward project to work with. Believe me, I have not been insulting, I have held back in many places. There will be close to 90 cores running this project now, although I've had to remove 2 of 6 from an i5 which causes slowdown of the GPU, and 12 of 24 from a Ryzen which is my main machine and slows the interface. But that's VB's/AMD's bad design.
There are people with 128 cores to offer and you are no where near the top producer here.
I've run 18 of these WU's on a 16 thread 2700x and it barely interfered with the GUI fucntions.
You are running out of RAM and heading into a swap crisis and didn't even notice. AMD is not at fault. It's user error for not properly accounting for machine RAM.
ID: 146 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Peter Hucker

Send message
Joined: 13 Jan 21
Posts: 54
Credit: 1,225,444
RAC: 69
Message 158 - Posted: 21 Jan 2021, 13:44:16 UTC - in response to Message 146.  
Last modified: 21 Jan 2021, 14:01:07 UTC

which you agreed with and admitted this project screwed up.
You really need to stop.
I did not say 'screwed up'.
I defended their actions.
They said "Hence, it is not a nice behaviour to launch a new project now that forces already outdated software components."
You said "I can agree with that."
You admitted that this project was engaging in "not nice behaviour".

Again, there is no reason to write a new program which doesn't function on the latest version of something.
The devs here used a turn-key solution and did not write it. This project has a limited scope to solve a single cryptographic puzzle and it would be a waste of time and labor to write from scratch.
They didn't do their homework and used something that was outdated and incompatible with the version of VB which currently comes with Boinc, so anyone with up to date software could not run this project properly.

Requiring users to downgrade is absurd and nothing less.
This is science project for BOINC and not gaming environment. I've seen many requests to downgrade and even the BOINC version of VBox is locked into an older version and warns against upgrading to the newest in the instructions.
Games are nothing to do with it, we don't use VB for games, but we do use it for other Boinc projects and for work activities.

When this project realised it wouldn't work on the latest version, they should have gone back to the drawing board.
They did go back and to setup new WU's and announced them yesterday and have stopped these so... why do you suggest this?
I suggested it before that announcement, and they're still using VB for the new ones. I saw no mention of it saying it would work with new VB.

Would you release a windows program that required everyone to go back to Windows 7? Sales would plummet.
Non-sequitor. We are in a research environment, not sales. There is plenty of processing power available for BOINC and they do not have to worry about angering a few petty individuals.
I think you'll find most people have VB 6, since it's handed out with Boinc!

I asked LHC if they minded me downgrading and they said it was a bad idea.
They said it was not preferred but the LHC WU's work fine in 5.2.44. It is not their project. Their opinion does not matter here.
Their opinion matters for anyone running both projects.

I guess they should have RTFM.
You do not seem to understand what abusive behavior is. When you tell someone to "read the fucking manual", that is vulgar and abusive behavior.
You're the one using swearwords. RTFM may stand for that, but it simply means to read the manual. If I try to change the brakes on my car without reading the manual, a mechanic could say I should have RTFM, it doesn't mean he's being abusive.

I have spent hours messing about getting this project to work, I'm trying to help, despite this being the most awkward project to work with. Believe me, I have not been insulting, I have held back in many places. There will be close to 90 cores running this project now, although I've had to remove 2 of 6 from an i5 which causes slowdown of the GPU, and 12 of 24 from a Ryzen which is my main machine and slows the interface. But that's VB's/AMD's bad design.
There are people with 128 cores to offer and you are no where near the top producer here.
I'm in 13th place, two ahead of you. People in glass houses shouldn't throw stones.

I've run 18 of these WU's on a 16 thread 2700x and it barely interfered with the GUI fucntions.
You are running out of RAM and heading into a swap crisis and didn't even notice. AMD is not at fault. It's user error for not properly accounting for machine RAM.
I checked my RAM and am not running out, each task (not WU, please use the correct notation) takes 0.75GB, I have well over a GB per thread on all my machines. I don't know where you got your information from, but you haven't spied on my machines very accurately. The one in question which I said was slowing down the interface (my only AMD) has 64GB for 24 threads. Swapping RAM isn't the problem, the problem is VB not multitasking nicely. Same thing happens with multiple ATLAS or Theory tasks from LHC.
ID: 158 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Werinbert

Send message
Joined: 15 Jan 21
Posts: 4
Credit: 100,245
RAC: 0
Message 160 - Posted: 21 Jan 2021, 17:09:09 UTC

Peter, LHC (and probably all projects) say they prefer people to upgrade unless upgrading affects their project (which has happened in the past with LHC). None the less, I am running VBox 5.0.18 and it is working just fine for LHC as well as for Kryptos.
Every project has its quirks. For Kryptos it is using a slightly older version of VBox. For any project, it is easy enough to accept the quirks as they are and go on crunching. If the quirks are not compatible with your machine then simply move on to another project. However, I do see you posting on other projects complaining about quirks as if the quirks were there to insult you.
ID: 160 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
1 · 2 · Next

Message boards : News : New task batches available

©2021 Kryptos@Home