Message boards : Theory Application : New Native App - Linux Only
Message board moderation

To post messages, you must log in.

Previous · 1 . . . 4 · 5 · 6 · 7 · 8 · 9 · 10 · Next

AuthorMessage
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 5961 - Posted: 20 Feb 2019, 9:31:37 UTC

I increased ncpus to 3. 3 tasks running and 1 ready to start. Limit on tasks in progress.

<max_wus_to_send>2</max_wus_to_send> makes no sense now, because the host limit = 4.
ID: 5961 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 5962 - Posted: 20 Feb 2019, 9:34:09 UTC - in response to Message 5961.  
Last modified: 20 Feb 2019, 9:34:17 UTC

I increased ncpus to 3. 3 tasks running and 1 ready to start. Limit on tasks in progress.

2 makes no sense now, because the host limit = 4.


On the server max_wus_to_send is currently set to 1.
ID: 5962 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 5964 - Posted: 20 Feb 2019, 9:43:31 UTC - in response to Message 5962.  

I increased ncpus to 3. 3 tasks running and 1 ready to start. Limit on tasks in progress.

<max_wus_to_send>2</max_wus_to_send> makes no sense now, because the host limit = 4.


On the server max_wus_to_send is currently set to 1.

Could you increase max_wus_in_progress to e.g. 16 and set max_wus_to_send to 2?
ID: 5964 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 5967 - Posted: 20 Feb 2019, 10:12:00 UTC - in response to Message 5964.  

Could you increase max_wus_in_progress to e.g. 16 and set max_wus_to_send to 2?


Done
ID: 5967 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 5969 - Posted: 20 Feb 2019, 10:21:11 UTC - in response to Message 5967.  
Last modified: 20 Feb 2019, 10:28:28 UTC

Could you increase max_wus_in_progress to e.g. 16 and set max_wus_to_send to 2?


Done

The number of tasks is not limited to 8. I've 10 now. So you seem to be right.
I don't know now, how to limit the number of wu's to the number of cores times N . . .
ID: 5969 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 5980 - Posted: 20 Feb 2019, 12:47:28 UTC - in response to Message 5969.  

The number of tasks is not limited to 8. I've 10 now. So you seem to be right.
I don't know now, how to limit the number of wu's to the number of cores times N . . .


Submitted an issue.
ID: 5980 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 5984 - Posted: 20 Feb 2019, 13:22:00 UTC - in response to Message 5980.  
Last modified: 20 Feb 2019, 13:22:50 UTC

Submitted an issue.

For the time being, you want maybe reduce max_wus_in_progress to a for you acceptable number . . . 8-) maybe
ID: 5984 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 5986 - Posted: 20 Feb 2019, 13:39:08 UTC - in response to Message 5924.  

How about the science application not obeying the suspend by the user / BOINC wrapper?


Issue submitted
ID: 5986 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 5990 - Posted: 20 Feb 2019, 14:41:10 UTC - in response to Message 5984.  

For the time being, you want maybe reduce max_wus_in_progress to a for you acceptable number . . . 8-) maybe

I have set it to 4. The focus is on getting it working rather than throughput.
ID: 5990 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
computezrmle
Volunteer moderator
Project tester
Volunteer developer
Volunteer tester
Help desk expert
Avatar

Send message
Joined: 28 Jul 16
Posts: 484
Credit: 394,839
RAC: 0
Message 6006 - Posted: 21 Feb 2019, 6:22:32 UTC

@Laurence

Did the tasks below fail due to an app error or did you simply cancel them on the server last night to get the queue drained?

https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=2753236
https://lhcathomedev.cern.ch/lhcathome-dev/result.php?resultid=2753243
ID: 6006 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 6007 - Posted: 21 Feb 2019, 6:27:20 UTC - in response to Message 6006.  

App error. I pushed out a new version.
ID: 6007 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 6009 - Posted: 21 Feb 2019, 8:47:08 UTC - in response to Message 5990.  

I have set it to 4. The focus is on getting it working rather than throughput.

On my Windows client, I have 8 in progress now and then
lhcathome-dev 21 Feb 09:45:19 This computer has reached a limit on tasks in progress
ID: 6009 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 6011 - Posted: 21 Feb 2019, 8:56:56 UTC - in response to Message 6009.  
Last modified: 21 Feb 2019, 8:57:21 UTC

I have set it to 4. The focus is on getting it working rather than throughput.

On my Windows client, I have 8 in progress now and then
lhcathome-dev 21 Feb 09:45:19 This computer has reached a limit on tasks in progress


As far as I can tell you have two quad cores CPUs, is this correct? This would mean 2*4=8. The comment in the issue suggests this parameter should be threads but it is not what we are experiencing. Will do some tests myself.
ID: 6011 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 6012 - Posted: 21 Feb 2019, 9:18:27 UTC - in response to Message 6011.  

As far as I can tell you have two quad cores CPUs, is this correct? This would mean 2*4=8

I have 1 quad-core (hostid=37) - 4 physical cores and with hyper-threading on it makes 8 threads. For BOINC 8 ncpus.
The second quad you see is a Linux VM (hostid 3717) on the same Windows host setup with 4 cores/threads.
In my preferences I've set 'No Limit' for Max # of jobs.
ID: 6012 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 6013 - Posted: 21 Feb 2019, 9:39:10 UTC - in response to Message 6012.  

As far as I can tell you have two quad cores CPUs, is this correct? This would mean 2*4=8

I have 1 quad-core (hostid=37) - 4 physical cores and with hyper-threading on it makes 8 threads. For BOINC 8 ncpus.
The second quad you see is a Linux VM (hostid 3717) on the same Windows host setup with 4 cores/threads.
In my preferences I've set 'No Limit' for Max # of jobs.

In the project db for your host I see ncpus = 8, so that is correct. In our project config we have:
max_wus_to_send = 1
max_wus_in_progress =   4

Let me increase max_wus_to_send to 2 as see if you can get 16 tasks.
ID: 6013 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 6015 - Posted: 21 Feb 2019, 9:54:43 UTC - in response to Message 6013.  

Let me increase max_wus_to_send to 2 as see if you can get 16 tasks.
Not sure whether you already made that change.
After having 8 tasks loaded I get lhcathome-dev 21 Feb 10:49:11 This computer has reached a limit on tasks in progress
We are in the same timezone, so you know, if that change was already made at 10:49 CET
Meanwhile I returned 1 job, so 7 in progress, but lhcathome-dev 21 Feb 10:52:40 No tasks are available for Theory Simulation
ID: 6015 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 6016 - Posted: 21 Feb 2019, 9:57:41 UTC - in response to Message 6015.  

Let me increase max_wus_to_send to 2 as see if you can get 16 tasks.
Not sure whether you already made that change.
After having 8 tasks loaded I get lhcathome-dev 21 Feb 10:49:11 This computer has reached a limit on tasks in progress
We are in the same timezone, so you know, if that change was already made at 10:49 CET
Meanwhile I returned 1 job, so 7 in progress, but lhcathome-dev 21 Feb 10:52:40 No tasks are available for Theory Simulation

The change has been done. Please try again.
ID: 6016 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 6017 - Posted: 21 Feb 2019, 10:13:11 UTC - in response to Message 6016.  

Let me increase max_wus_to_send to 2 as see if you can get 16 tasks.
Not sure whether you already made that change.
After having 8 tasks loaded I get lhcathome-dev 21 Feb 10:49:11 This computer has reached a limit on tasks in progress
We are in the same timezone, so you know, if that change was already made at 10:49 CET
Meanwhile I returned 1 job, so 7 in progress, but lhcathome-dev 21 Feb 10:52:40 No tasks are available for Theory Simulation

The change has been done. Please try again.
lhcathome-dev 21 Feb 11:10:43 This computer has reached a limit on tasks in progress

8 tasks loaded.
ID: 6017 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Profile Laurence
Project administrator
Project developer
Project tester
Avatar

Send message
Joined: 12 Sep 14
Posts: 1069
Credit: 334,882
RAC: 0
Message 6018 - Posted: 21 Feb 2019, 10:33:30 UTC - in response to Message 6017.  

lhcathome-dev 21 Feb 11:10:43 This computer has reached a limit on tasks in progress

8 tasks loaded.


I have submitted more tasks.
ID: 6018 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Crystal Pellet
Volunteer tester

Send message
Joined: 13 Feb 15
Posts: 1188
Credit: 862,257
RAC: 61
Message 6019 - Posted: 21 Feb 2019, 10:41:52 UTC - in response to Message 6018.  

I have submitted more tasks.
Laurence, do you remember we have had an issue with the project specific Max # of jobs?
Although you had set 'No Limit', there seems to be a limit. Could that be 8?
However yesterday, I got 16 on my Linux VM, when you raised max_wus_in_progress to 16. Weird.
ID: 6019 · Rating: 0 · rate: Rate + / Rate - Report as offensive     Reply Quote
Previous · 1 . . . 4 · 5 · 6 · 7 · 8 · 9 · 10 · Next

Message boards : Theory Application : New Native App - Linux Only


©2024 CERN