Problems and Technical Issues with Rosetta@home

Message boards : Number crunching : Problems and Technical Issues with Rosetta@home

To post messages, you must log in.

Previous · 1 . . . 154 · 155 · 156 · 157 · 158 · 159 · 160 . . . 257 · Next

AuthorMessage
rkv_2401

Send message
Joined: 11 Dec 19
Posts: 5
Credit: 1,232,418
RAC: 0
Message 104125 - Posted: 10 Jan 2022, 11:57:01 UTC - in response to Message 104114.  

Hi, yes, I would like to get it working on my i5 machine and the error I keep getting is 'VirtualBox jobs require hardware acceleration support - your processor does not support the required instruction set.'

The machine in question is an i5 10400 on a GIGABYTE B460M-DS3h V2.0, 32GB RAM, 512GB nVME SSD, Windows 11 64-bit. Virtualization is enabled in the BIOS, supported by the hardware, and Windows reports it as being on too.

I would be happy to run memory intensive python tasks on this machine - if they worked!
ID: 104125 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
tullio

Send message
Joined: 10 May 20
Posts: 63
Credit: 630,125
RAC: 32
Message 104127 - Posted: 10 Jan 2022, 13:40:02 UTC

I have an Intel i5 9400F, Windows 11 home edition, VBox 6.1.30 and it works perfectly on rosetta python.
Tullio
ID: 104127 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104139 - Posted: 10 Jan 2022, 14:58:31 UTC - in response to Message 104109.  

I have been trying to run Rosetta@home on 3 different computers for at least the last 3 months now - didn't get a single task no matter what I tried until yesterday, now I've finally had a 'breakthrough' on one of my PCs getting some tasks, but all of them just fail immediately!

Nov 2021 - I try to run Rosetta@home on my Surface Pro 2 (Win 10, i5-4300U w/ 8GB RAM, CPU supports hardware virtualization, virtualization support enabled according to task manager) - 'VirtualBox requires hardware support - which your CPU does not provide' error, or something to that effect. I try restarting the system, reinstalling VBox 6.x, uninstalling Vbox 6.x and installing Vbox 5.2.44 according to some suggestions on this thread, nothing works, I wait for regular tasks, nothing happens, I eventually gave up on it.

Dec 2021 - I try to run Rosetta@home on a new desktop (Win 11, i5-10400 w/ 32GB RAM, CPU supports hardware virtualization, virtualization support enabled according to task manager) - same error as above, tried the same steps, nothing happens, I let it wait for tasks - so far, it hasn't gotten a single one in 14+ days.

Jan 2022 - I half-heartedly try to run Rosetta on my previous crunching machine (Win 10, Ryzen 5 3600 w/ 24GB RAM, CPU supports hardware virtualization, virtualization support DISABLED according to task manager, Hyper-V enabled) - now I start getting tasks on this machine (but not on the 10th gen i5 one that's also plugged in and running BOINC?) And all of them fail anyway - v4.20, vBox 1.03 jobs, all of them fail instantly. *shrugs*

All 3 machines are/were on VBox 5.2.44 as suggested earlier in this thread.
Are you sure virtualization is on? Check in the BIOS, don't believe the task manager, maybe it's talking about software virtualization.
ID: 104139 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104140 - Posted: 10 Jan 2022, 14:59:38 UTC - in response to Message 104113.  

Two of you with identical errors. That sounds like a code problem by the team not your machines.
And 12 hrs for 3 minutes of CPU time, definitely something wrong.
Are you running RAH on a dedicated core and don't have much of a system load?
I've run stuff almost like that in the past, but never 12hrs for 3 minutes!
In my case loads of RAM, tried to run one at a time, nothing overloaded. If this is a programming error, can we report it somewhere? I doubt any of them are reading this.
ID: 104140 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104141 - Posted: 10 Jan 2022, 15:22:46 UTC - in response to Message 104095.  
Last modified: 10 Jan 2022, 15:24:35 UTC

I am getting identical times for running time and CPU time.
Tullio
I'm getting slightly more for CPU time than running time, I thought maybe it was multicore! E.g. 3h31m CPU time in 3h15m running time.

Only seeing this in Boinctasks though, Boinc Manager doesn't seem to show CPU time.
ID: 104141 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104145 - Posted: 10 Jan 2022, 17:10:02 UTC - in response to Message 104140.  
Last modified: 10 Jan 2022, 17:19:25 UTC

Two of you with identical errors. That sounds like a code problem by the team not your machines.
And 12 hrs for 3 minutes of CPU time, definitely something wrong.
Are you running RAH on a dedicated core and don't have much of a system load?
I've run stuff almost like that in the past, but never 12hrs for 3 minutes!
In my case loads of RAM, tried to run one at a time, nothing overloaded. If this is a programming error, can we report it somewhere? I doubt any of them are reading this.
<-- yep. no one reads no one cares.

The only way they notice anything is when the results of 30+ tasks error out and skew the data then they come back for a look. But a few here and there on single machines...nah..not interested.

And I have no way to look at any of your tasks on any of your machines.
It's all 0's and your first machine has not made "contact" since Jan 1.
Everything else is november or even may.
So if they are connected to GRC, then I guess we don't see them.
Which computer ID do you refer to when having problems?
ID: 104145 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104146 - Posted: 10 Jan 2022, 17:10:58 UTC - in response to Message 104141.  
Last modified: 10 Jan 2022, 17:16:45 UTC

I am getting identical times for running time and CPU time.
Tullio
I'm getting slightly more for CPU time than running time, I thought maybe it was multicore! E.g. 3h31m CPU time in 3h15m running time.

Only seeing this in Boinctasks though, Boinc Manager doesn't seem to show CPU time.


It takes a bit to open a VM and check everything and then write everything and shut it down.
Your within the limits. I see about the same.

Run time 4 hours 49 min 55 sec
CPU time 4 hours 38 min 16 sec
Validate state Valid
Credit 218.04
Device peak FLOPS 5.41 GFLOPS
Application version rosetta python projects v1.03 (vbox64)
windows_x86_64
Peak working set size 61.11 MB
Peak swap size 58.45 MB
Peak disk usage 8,745.88 MB


You can find this at the top of the results page for whatever task you want to look at.
ID: 104146 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104148 - Posted: 10 Jan 2022, 17:39:52 UTC - in response to Message 104105.  

The 0x80070005 error code arises when the Windows access is denied. ... The 0x80070005 error relates to access permissions in Windows, sometimes encountered by users when they are trying to update their Windows or installing a new program, and they don't possess full access permission for certain files

Which is odd, because it ran ok, but when its writing the output there is where the errors are.
I have to get to bed again, but maybe tomorrow night I can look.

But it appears to be a windows permissions thing at first glance.
I see this in a TACC post and a LHC post.

If you search the gibberish:

1) aRC=E_ACCESSDENIED (0x80070005)
2) aIID={85cd948e-a71f-4289-281e-0ca7ad48cd89} aComponent={MachineWrap} aText={The object functionality is limited}

You might find the answer...add the word BOINC if you can not find what you are looking for.


I'm off to bed, but will have a look tomorrow afternoon. The only thing I found was Vbox was not installed with admin rights. Apparently being a user with admin rights isn't good enough. I'll try reinstalling it on an offending machine tomorrow using elevated privileges. Stupid Microsoft.....
No, that didn't help.
ID: 104148 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104149 - Posted: 10 Jan 2022, 17:41:50 UTC - in response to Message 104145.  
Last modified: 10 Jan 2022, 17:42:16 UTC

Two of you with identical errors. That sounds like a code problem by the team not your machines.
And 12 hrs for 3 minutes of CPU time, definitely something wrong.
Are you running RAH on a dedicated core and don't have much of a system load?
I've run stuff almost like that in the past, but never 12hrs for 3 minutes!
In my case loads of RAM, tried to run one at a time, nothing overloaded. If this is a programming error, can we report it somewhere? I doubt any of them are reading this.
<-- yep. no one reads no one cares.

The only way they notice anything is when the results of 30+ tasks error out and skew the data then they come back for a look. But a few here and there on single machines...nah..not interested.

And I have no way to look at any of your tasks on any of your machines.
It's all 0's and your first machine has not made "contact" since Jan 1.
Everything else is november or even may.
So if they are connected to GRC, then I guess we don't see them.
Which computer ID do you refer to when having problems?
Surely there's an address we can send to, David Baker for example? And it isn't just machines here and there. Judging by how little work is done on pythons, hardly anyone is doing them. And 6 of my 7 machines failed to.

The machine I've been referring to most recently is https://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=6169682
ID: 104149 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
tullio

Send message
Joined: 10 May 20
Posts: 63
Credit: 630,125
RAC: 32
Message 104150 - Posted: 10 Jan 2022, 18:05:02 UTC

In Atlas@home from LHC CPU time is double than running time. I allow it to run on two processors of my six. rosetta pythom is running two tasks, one on each processor. The third one is waiting for more memory of my 12 GB, the fourth one is ready to run.
Tullio
ID: 104150 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104151 - Posted: 10 Jan 2022, 19:12:42 UTC - in response to Message 104149.  

Two of you with identical errors. That sounds like a code problem by the team not your machines.
And 12 hrs for 3 minutes of CPU time, definitely something wrong.
Are you running RAH on a dedicated core and don't have much of a system load?
I've run stuff almost like that in the past, but never 12hrs for 3 minutes!
In my case loads of RAM, tried to run one at a time, nothing overloaded. If this is a programming error, can we report it somewhere? I doubt any of them are reading this.
<-- yep. no one reads no one cares.

The only way they notice anything is when the results of 30+ tasks error out and skew the data then they come back for a look. But a few here and there on single machines...nah..not interested.

And I have no way to look at any of your tasks on any of your machines.
It's all 0's and your first machine has not made "contact" since Jan 1.
Everything else is november or even may.
So if they are connected to GRC, then I guess we don't see them.
Which computer ID do you refer to when having problems?
Surely there's an address we can send to, David Baker for example? And it isn't just machines here and there. Judging by how little work is done on pythons, hardly anyone is doing them. And 6 of my 7 machines failed to.

The machine I've been referring to most recently is https://boinc.bakerlab.org/rosetta/show_host_detail.php?hostid=6169682


------------------------------------------



Re: VBOX_E_OBJECT_NOT_FOUND (0x80BB0001)

Postby mpack » 7. Jan 2017, 15:37
Usually it means that you are using VirtualBox on a Windows host while logged in as a different user, i.e. not the user who installed the software. To install the software you should use "Run as administrator", not log in as the administrator.

Also, VirtualBox needs to install certain drivers on the host which do redirection of network and USB devices. You need to allow that to happen.

This is what all that gibberish boils down to. Something is not configured right on your system.
In one instance it was running and then something changed and it could not find the path anymore which lead to this 0x error.
And in another you got the 0x error and this message: {The object functionality is limited} and this error: {The object is not ready}

All the ones that ran have this error.
It is best you try and reinstall Vbox and see if the configuration will return to normal.
If it doesn't then you need to go over to their forum and ask what to do.
This is not a RAH issue. This is a configuration issue on your system.
ID: 104151 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104152 - Posted: 10 Jan 2022, 19:23:26 UTC - in response to Message 104150.  

In Atlas@home from LHC CPU time is double than running time. I allow it to run on two processors of my six. rosetta pythom is running two tasks, one on each processor. The third one is waiting for more memory of my 12 GB, the fourth one is ready to run.
Tullio


15 cores with the new memory installed running this group

1 Python
3 WCG (2 pandemic 1 cancer)
1 Einstein (GPU and 1 core CPU)
1 Prime grid (GPU and 1 core CPU)
5 SiDock
1 ATLAS
1 QuChem (had to restart it after first start up)

Only using 60% of my memory now 49 Gigs now installed.
ID: 104152 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104153 - Posted: 10 Jan 2022, 19:35:44 UTC - in response to Message 104151.  

Re: VBOX_E_OBJECT_NOT_FOUND (0x80BB0001)

Postby mpack » 7. Jan 2017, 15:37
Usually it means that you are using VirtualBox on a Windows host while logged in as a different user, i.e. not the user who installed the software. To install the software you should use "Run as administrator", not log in as the administrator.

Also, VirtualBox needs to install certain drivers on the host which do redirection of network and USB devices. You need to allow that to happen.

This is what all that gibberish boils down to. Something is not configured right on your system.
In one instance it was running and then something changed and it could not find the path anymore which lead to this 0x error.
And in another you got the 0x error and this message: {The object functionality is limited} and this error: {The object is not ready}

All the ones that ran have this error.
It is best you try and reinstall Vbox and see if the configuration will return to normal.
If it doesn't then you need to go over to their forum and ask what to do.
This is not a RAH issue. This is a configuration issue on your system.
It's not my system. I only log on as one user, which is an admin. I reinstalled elevated to admin with no improvement. Vbox works perfectly well on LHC, this is a Rosetta screwup. I've done nothing different on the machine that works, the only major difference is the CPU type. And since hardly anyone is doing Python, this suggests many are having difficulties.
ID: 104153 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
tullio

Send message
Joined: 10 May 20
Posts: 63
Credit: 630,125
RAC: 32
Message 104154 - Posted: 10 Jan 2022, 20:00:38 UTC
Last modified: 10 Jan 2022, 20:02:06 UTC

I have another PC with a AMD Ryzen CPU and 24 GB RAM, but it runs mostly World Comminity Grid projects, like OpenPandemics-Covid-19, for obvious reasons, Mapping CancerMarkers and Africa Rainfall project.The project used to be managed by IBM, now by the Krembil Research Institute, Ottawa, Canada.
Tullio
ID: 104154 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Mr P Hucker
Avatar

Send message
Joined: 12 Aug 06
Posts: 1479
Credit: 7,238,581
RAC: 12,848
Message 104156 - Posted: 10 Jan 2022, 20:41:41 UTC - in response to Message 104154.  

I have another PC with a AMD Ryzen CPU and 24 GB RAM, but it runs mostly World Comminity Grid projects, like OpenPandemics-Covid-19, for obvious reasons, Mapping CancerMarkers and Africa Rainfall project.The project used to be managed by IBM, now by the Krembil Research Institute, Ottawa, Canada.
Tullio
Yes, mine is over there most of the time since Rosetta started playing up. I can't find any GPU (AMD) biology Boinc projects though. Covid on WCG had some for a time but not any more.
ID: 104156 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
.clair.

Send message
Joined: 2 Jan 07
Posts: 202
Credit: 23,900,296
RAC: 16,691
Message 104157 - Posted: 10 Jan 2022, 21:35:17 UTC - in response to Message 104113.  

You are not alone with that error, I have some like that, many do run ok , this was another dud.
184:11:40.953516 ERROR [COM]: aRC=E_ACCESSDENIED (0x80070005) aIID={5047460a-265d-4538-b23e-ddba5fb84976} aComponent={MachineWrap} aText={The object functionality is limited}, preserve=false aResultDetail=0
184:11:40.953516 ERROR [COM]: aRC=E_ACCESSDENIED (0x80070005) aIID={5047460a-265d-4538-b23e-ddba5fb84976} aComponent={MachineWrap} aText={The object functionality is limited}, preserve=false aResultDetail=0
184:11:40.953516 ERROR [COM]: aRC=E_ACCESSDENIED (0x80070005) aIID={5047460a-265d-4538-b23e-ddba5fb84976} aComponent={MachineWrap} aText={The object functionality is limited}, preserve=false aResultDetail=0
184:11:40.953516 ERROR [COM]: aRC=E_ACCESSDENIED (0x80070005) aIID={5047460a-265d-4538-b23e-ddba5fb84976} aComponent={MachineWrap} aText={The object functionality is limited}, preserve=false aResultDetail=0
Run time 12 hours 15 min 28 sec
CPU time 3 min 40 sec
It may still be visible by the time you read this - Received 9 Jan 2022, 19:01:27 UTC
It was one of snail workunits that start but get nowhere fast
https://boinc.bakerlab.org/rosetta/result.php?resultid=1463638243

Two of you with identical errors. That sounds like a code problem by the team not your machines.
And 12 hrs for 3 minutes of CPU time, definitely something wrong.
Are you running RAH on a dedicated core and don't have much of a system load?
I've run stuff almost like that in the past, but never 12hrs for 3 minutes!

Its a dedicated BOINC cruncher the only `system load` is BOINC , 48 cpu/threads 128GB memory
It runs rosetta on CPU and Milkyway on GPU and nothing else
It is not attached to any other projects and never has been
It was a friends workstation , he died a few years ago, he was a marine biologist ,
I got given the computer for helping clear his house witch is why it is named `Kaminski` after him [polish name] with the intention to run boinc on it till it stops working
ID: 104157 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Michael E.

Send message
Joined: 5 Apr 08
Posts: 16
Credit: 1,723,927
RAC: 323
Message 104160 - Posted: 10 Jan 2022, 22:51:59 UTC - in response to Message 104108.  

How are prospective new users gonna find this stuff out?

That is a very good question. We long-time users have heard about it on the forums, and see the one marked "vbox64" in the applications.
https://boinc.bakerlab.org/rosetta/apps.php

But that is scant information indeed, and tells you nothing about system requirements.
It is a poor way to run a railroad. They won't get much work done that way.


Thank you Jim1348! Well said.

I strongly suggest that Admins/project owners add the link to the apps page https://boinc.bakerlab.org/rosetta/apps.php that you kindly provided after a phrase like:

The Rosetta@home subprojects are described here: https://boinc.bakerlab.org/rosetta/apps.php

Now, how do I prevent Python tasks that get sent to me and only send me Rosetta tasks? That is all I have received recently on my main desktop system. One Python task ran for 3 days and others hit various resource issues. The messages related to resource issues (such as waiting for memory) are not very helpful, but maybe that is a BOINC issue? I am running BOINC 7.16.20 and VBox 6.1 on my Windows 10 PC.

On the other Windows systems where VBox is NOT installed, I do not get any ROSETTA@Home tasks at all because VBox is not installed. I get a message on those PCs that VBox is not installed (well, duh). That does not happen with Ralph@Home.

The only option in Rosetta@home Preferences seems to be the run-time length. I will try to set it to 1 day instead of 8 hours and see if that helps.

I won't uninstall BOINC on my PC as I have lots of BOINC projects.
ID: 104160 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Falconet

Send message
Joined: 9 Mar 09
Posts: 348
Credit: 989,292
RAC: 1
Message 104161 - Posted: 10 Jan 2022, 22:57:11 UTC - in response to Message 104160.  
Last modified: 10 Jan 2022, 22:59:35 UTC

To stop receiving the Pythons, go to your computer list on your Rosetta account page ( or copy-paste this link https://boinc.bakerlab.org/rosetta/hosts_user.php ), click on details and click Skip where it says VirtualBox VM jobs.

The reason your other computers aren't receiving standard Rosetta jobs isn't because you haven't installed VirtualBox on them but rather because there simply isn't any work available right now.
ID: 104161 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104163 - Posted: 10 Jan 2022, 23:37:05 UTC - in response to Message 104156.  

I have another PC with a AMD Ryzen CPU and 24 GB RAM, but it runs mostly World Comminity Grid projects, like OpenPandemics-Covid-19, for obvious reasons, Mapping CancerMarkers and Africa Rainfall project.The project used to be managed by IBM, now by the Krembil Research Institute, Ottawa, Canada.
Tullio
Yes, mine is over there most of the time since Rosetta started playing up. I can't find any GPU (AMD) biology Boinc projects though. Covid on WCG had some for a time but not any more.



Non BOINC GPU protein is at FAH

Another BOINC you can try is QuChemPediA and SiDock if your not already there.
ID: 104163 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Profile Greg_BE
Avatar

Send message
Joined: 30 May 06
Posts: 5652
Credit: 5,622,096
RAC: 48
Message 104164 - Posted: 10 Jan 2022, 23:41:18 UTC

.clair. and Peter, I can't do anything more for you.
RAH team will do nothing to help you.
So I suggest you go over to Vbox forum and post your error messages and questions there.
They might know what's going on since it is their software.
If it not in their software then its a batch of tasks that don't like your computers.
I have not had any errors that I know of in the last 48 hrs and I run 5 CPU and 2 GPU projects on 16 cores.
ID: 104164 · Rating: 0 · rate: Rate + / Rate - Report as offensive    Reply Quote
Previous · 1 . . . 154 · 155 · 156 · 157 · 158 · 159 · 160 . . . 257 · Next

Message boards : Number crunching : Problems and Technical Issues with Rosetta@home



©2022 University of Washington
https://www.bakerlab.org