How to get specific Task ?

Author	Message
b3rl1go Send message Joined: 15 Apr 16 Posts: 6 Credit: 66,628,827 RAC: 0	Message 8096 - Posted: 10 Feb 2022, 9:20:10 UTC Last modified: 10 Feb 2022, 9:20:39 UTC
Hi, My Nvidia can't run "TF v0.12 (cuda111)". Ma Nvidia ne peut pas exécuter les "TF v0.12 (cuda111)". I only can crunch on the "TF v0.14 (cuda100)". J'arrive seulement à cruncher sur les "TF v0.14 (cuda100)". How can i say to Boinc to download only "TF v0.14" and ignore the other ? Comment puis-je dire à Boinc de ne télécharger que les "TF v0.14" et ignorer les autres ? Thank's Merci
ID: 8096 · Rating: 0 · rate: / Reply Quote

rebirther Volunteer moderator Project administrator Project developer Project tester Project scientist Send message Joined: 2 Jan 13 Posts: 7479 Credit: 43,657,067 RAC: 40,675	Message 8097 - Posted: 10 Feb 2022, 11:54:53 UTC - in response to Message 8096.
Hi, My Nvidia can't run "TF v0.12 (cuda111)". Ma Nvidia ne peut pas exécuter les "TF v0.12 (cuda111)". I only can crunch on the "TF v0.14 (cuda100)". J'arrive seulement à cruncher sur les "TF v0.14 (cuda100)". How can i say to Boinc to download only "TF v0.14" and ignore the other ? Comment puis-je dire à Boinc de ne télécharger que les "TF v0.14" et ignorer les autres ? Thank's Merci You must change your driver, see FAQ
ID: 8097 · Rating: 0 · rate: / Reply Quote

b3rl1go Send message Joined: 15 Apr 16 Posts: 6 Credit: 66,628,827 RAC: 0	Message 8098 - Posted: 10 Feb 2022, 11:59:27 UTC - in response to Message 8097.
You must change your driver, see FAQ I Can't. I'm on a Linux VM and i am not admin.
ID: 8098 · Rating: 0 · rate: / Reply Quote

b3rl1go Send message Joined: 15 Apr 16 Posts: 6 Credit: 66,628,827 RAC: 0	Message 8099 - Posted: 10 Feb 2022, 14:28:22 UTC - in response to Message 8098.
I just try to add an app_config.xml with the Following lines but it don't work. I'm still downloading TF v0.12 (cuda111) <app_config> <app> <name>TF</name> <max_concurrent>1</max_concurrent> <gpu_versions> <cpu_usage>0.1</cpu_usage> </gpu_versions> <app_version> <app_name>cuda100</app_name> </app_version> </app> </app_config> An idea?
ID: 8099 · Rating: 0 · rate: / Reply Quote

rebirther Volunteer moderator Project administrator Project developer Project tester Project scientist Send message Joined: 2 Jan 13 Posts: 7479 Credit: 43,657,067 RAC: 40,675	Message 8100 - Posted: 10 Feb 2022, 14:49:41 UTC - in response to Message 8099.
I just try to add an app_config.xml with the Following lines but it don't work. I'm still downloading TF v0.12 (cuda111) <app_config> <app> <name>TF</name> <max_concurrent>1</max_concurrent> <gpu_versions> <cpu_usage>0.1</cpu_usage> </gpu_versions> <app_version> <app_name>cuda100</app_name> </app_version> </app> </app_config> An idea? I also have tried this in the past, with your error you can run these failed tasks if you update glibc to v2.29
ID: 8100 · Rating: 0 · rate: / Reply Quote

b3rl1go Send message Joined: 15 Apr 16 Posts: 6 Credit: 66,628,827 RAC: 0	Message 8103 - Posted: 11 Feb 2022, 6:27:59 UTC - in response to Message 8100.
I also have tried this in the past, with your error you can run these failed tasks if you update glibc to v2.29 Again, it is impossible to make any update on this linux serveur.
ID: 8103 · Rating: 0 · rate: / Reply Quote

AnandBhat Send message Joined: 20 Mar 22 Posts: 10 Credit: 30,849,164 RAC: 19,246	Message 8159 - Posted: 28 Mar 2022, 4:58:48 UTC
I too am in the same situation where I cannot update the glibc version. I have tried to assign no devices to TF cuda111 by setting ngpus to zero with this app_config.xml but I've not had any luck. <app_config> <app_version> <app_name>TF</app_name> <version_num>12</version_num> <platform>x86_64-pc-linux-gnu</platform> <plan_class>cuda111</plan_class> <avg_ncpus>0</avg_ncpus> <ngpus>0</ngpus> </app_version> </app_config> I've also tried this without any luck. <app_config> <app_version> <app_name>TF</app_name> <plan_class>cuda111</plan_class> <avg_ncpus>99.000000</avg_ncpus> <ngpus>99.000000</ngpus> </app_version> <app_version> <app_name>TF</app_name> <plan_class>cuda100</plan_class> <avg_ncpus>1.000000</avg_ncpus> <ngpus>1.000000</ngpus> </app_version> <report_results_immediately/> </app_config> Any suggestions on how work for TF cuda111 can be avoided?
ID: 8159 · Rating: 0 · rate: / Reply Quote

rebirther Volunteer moderator Project administrator Project developer Project tester Project scientist Send message Joined: 2 Jan 13 Posts: 7479 Credit: 43,657,067 RAC: 40,675	Message 8160 - Posted: 28 Mar 2022, 5:08:22 UTC - in response to Message 8159. Last modified: 28 Mar 2022, 5:11:23 UTC
At the moment only update the driver. I have posted this in mersenne forum to get a smaller glibc2.29 version. There was a guide how to do this. Is someone is able to do this on Ubuntu 18 that would be fine.
ID: 8160 · Rating: 0 · rate: / Reply Quote

Author

Message

b3rl1go
Send message
Joined: 15 Apr 16
Posts: 6
Credit: 66,628,827
RAC: 0
1M in Sierpinski / Riesel Base credit

500k in Sierpinski / Riesel Base - short credit

5M in Sierpinski / Riesel Base - long credit

2M in Sierpinski / Riesel Base - average credit

Hi,

My Nvidia can't run "TF v0.12 (cuda111)".
Ma Nvidia ne peut pas exécuter les "TF v0.12 (cuda111)".

I only can crunch on the "TF v0.14 (cuda100)".
J'arrive seulement à cruncher sur les "TF v0.14 (cuda100)".

How can i say to Boinc to download only "TF v0.14" and ignore the other ?
Comment puis-je dire à Boinc de ne télécharger que les "TF v0.14" et ignorer les autres ?

Thank's
Merci

ID: 8096 · Rating: 0 · rate:

Reply Quote

rebirther
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,657,067
RAC: 40,675
2M in Sierpinski / Riesel Base credit

5M in Sierpinski / Riesel Base - short credit

1M in Sierpinski / Riesel Base - long credit

500k in Sierpinski / Riesel Base - average credit

Hi,

My Nvidia can't run "TF v0.12 (cuda111)".
Ma Nvidia ne peut pas exécuter les "TF v0.12 (cuda111)".

I only can crunch on the "TF v0.14 (cuda100)".
J'arrive seulement à cruncher sur les "TF v0.14 (cuda100)".

How can i say to Boinc to download only "TF v0.14" and ignore the other ?
Comment puis-je dire à Boinc de ne télécharger que les "TF v0.14" et ignorer les autres ?

Thank's
Merci

You must change your driver, see FAQ

ID: 8097 · Rating: 0 · rate:

Reply Quote

b3rl1go
Send message
Joined: 15 Apr 16
Posts: 6
Credit: 66,628,827
RAC: 0
1M in Sierpinski / Riesel Base credit

You must change your driver, see FAQ

I Can't. I'm on a Linux VM and i am not admin.

ID: 8098 · Rating: 0 · rate:

Reply Quote

b3rl1go
Send message
Joined: 15 Apr 16
Posts: 6
Credit: 66,628,827
RAC: 0
1M in Sierpinski / Riesel Base credit

I just try to add an app_config.xml with the Following lines but it don't work.
I'm still downloading TF v0.12 (cuda111)

<app_config> <app> <name>TF</name> <max_concurrent>1</max_concurrent> <gpu_versions> <cpu_usage>0.1</cpu_usage> </gpu_versions> <app_version> <app_name>cuda100</app_name> </app_version> </app> </app_config>

An idea?

ID: 8099 · Rating: 0 · rate:

Reply Quote

rebirther
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,657,067
RAC: 40,675
2M in Sierpinski / Riesel Base credit

I just try to add an app_config.xml with the Following lines but it don't work.
I'm still downloading TF v0.12 (cuda111)

<app_config> <app> <name>TF</name> <max_concurrent>1</max_concurrent> <gpu_versions> <cpu_usage>0.1</cpu_usage> </gpu_versions> <app_version> <app_name>cuda100</app_name> </app_version> </app> </app_config>

An idea?

I also have tried this in the past, with your error you can run these failed tasks if you update glibc to v2.29

ID: 8100 · Rating: 0 · rate:

Reply Quote

b3rl1go
Send message
Joined: 15 Apr 16
Posts: 6
Credit: 66,628,827
RAC: 0
1M in Sierpinski / Riesel Base credit

I also have tried this in the past, with your error you can run these failed tasks if you update glibc to v2.29

Again, it is impossible to make any update on this linux serveur.

ID: 8103 · Rating: 0 · rate:

Reply Quote

AnandBhat
Send message
Joined: 20 Mar 22
Posts: 10
Credit: 30,849,164
RAC: 19,246
500k in Sierpinski / Riesel Base credit

100k in Sierpinski / Riesel Base - short credit

I too am in the same situation where I cannot update the glibc version.

I have tried to assign no devices to TF cuda111 by setting ngpus to zero with this app_config.xml but I've not had any luck.

<app_config> <app_version> <app_name>TF</app_name> <version_num>12</version_num> <platform>x86_64-pc-linux-gnu</platform> <plan_class>cuda111</plan_class> <avg_ncpus>0</avg_ncpus> <ngpus>0</ngpus> </app_version> </app_config>

I've also tried this without any luck.

<app_config> <app_version> <app_name>TF</app_name> <plan_class>cuda111</plan_class> <avg_ncpus>99.000000</avg_ncpus> <ngpus>99.000000</ngpus> </app_version> <app_version> <app_name>TF</app_name> <plan_class>cuda100</plan_class> <avg_ncpus>1.000000</avg_ncpus> <ngpus>1.000000</ngpus> </app_version> <report_results_immediately/> </app_config>

Any suggestions on how work for TF cuda111 can be avoided?

ID: 8159 · Rating: 0 · rate:

Reply Quote

rebirther
Volunteer moderator
Project administrator
Project developer
Project tester
Project scientist

Send message
Joined: 2 Jan 13
Posts: 7479
Credit: 43,657,067
RAC: 40,675
2M in Sierpinski / Riesel Base credit

At the moment only update the driver. I have posted this in mersenne forum to get a smaller glibc2.29 version. There was a guide how to do this. Is someone is able to do this on Ubuntu 18 that would be fine.

ID: 8160 · Rating: 0 · rate:

Reply Quote