Server Status - No Work Available
log in

Advanced search

Message boards : Number crunching : Server Status - No Work Available

1 · 2 · Next
Author Message
Profile Bill F
Send message
Joined: 3 Apr 17
Posts: 16
Credit: 156,298
RAC: 377
United States
Message 1215 - Posted: 6 Dec 2017, 23:59:31 UTC

The server status page is showing no work units available for download as of
Task data as of 6 Dec 2017, 23:46:47 UTC

I am assuming that would reflect all types of WU's for both Microsoft and Linux

WU Generator shows "running" ??

Thanks
Bill F
____________

Profile [B@P] Daniel
Volunteer developer
Send message
Joined: 19 Oct 16
Posts: 80
Credit: 2,204,809
RAC: 0
Poland
Message 1216 - Posted: 7 Dec 2017, 6:40:52 UTC - in response to Message 1215.
Last modified: 7 Dec 2017, 6:41:07 UTC

The server status page is showing no work units available for download as of
Task data as of 6 Dec 2017, 23:46:47 UTC

I am assuming that would reflect all types of WU's for both Microsoft and Linux

WU Generator shows "running" ??

Thanks
Bill F

I have checked server status page and saw interesting thing:

Users registered in past 24 hours: 1
Computers registered in past 24 hours: 1717

Someone with big computing cloud has joined. No wonder that work queue is empty.
____________

Profile valterc
Project administrator
Project tester
Send message
Joined: 30 Oct 13
Posts: 360
Credit: 17,702,998
RAC: 5,731
Italy
Message 1217 - Posted: 7 Dec 2017, 10:23:46 UTC
Last modified: 7 Dec 2017, 10:34:15 UTC

Our work generator, for the current experiment, is capable to prepare a batch every 4.5 minutes, any batch is made by 97 workunits, times two for redundancy. The calculation is therefore (60/4.5)*97*2*24 = 62080, this is the maximum number of tasks that we can distribute daily. Other experiments has different numbers.

Another point is that any workunit, with the dataset and parameters of the current experiment, is relatively fast, ie. results come back fast. This got "worse" with the recent application update.

Col323
Send message
Joined: 23 Nov 16
Posts: 7
Credit: 1,311,980
RAC: 451
Angola
Message 1218 - Posted: 7 Dec 2017, 18:00:38 UTC - in response to Message 1217.

Thanks for pushing out the application optimization! I know it makes the work generation situation "worse", but I'm always happy to do more science with less effort.

Any luck with the work generation optimization?

Yavanius
Avatar
Send message
Joined: 2 Dec 16
Posts: 2
Credit: 77,381
RAC: 0
Antarctica
Message 1239 - Posted: 17 Jan 2018, 5:41:03 UTC - in response to Message 1216.

What project is the laurel leaves for?

I found out before, forgot, and I couldn't track down what project it was again... :/

Profile Bill F
Send message
Joined: 3 Apr 17
Posts: 16
Credit: 156,298
RAC: 377
United States
Message 1240 - Posted: 17 Jan 2018, 15:39:48 UTC - in response to Message 1239.

Green Laurel Leaves with a number in the Middle ?

That is where the user has been named User of the Day and the number is how many projects.
____________

Profile [VENETO] boboviz
Send message
Joined: 12 Dec 13
Posts: 147
Credit: 1,126,004
RAC: 368
Italy
Message 1259 - Posted: 9 Feb 2018, 19:20:54 UTC

I've a lot of "transient http error"

StyM
Send message
Joined: 1 Nov 16
Posts: 1
Credit: 16,161,078
RAC: 4
Singapore
Message 1260 - Posted: 14 Feb 2018, 16:29:59 UTC

unable to get work...

Profile Bill F
Send message
Joined: 3 Apr 17
Posts: 16
Credit: 156,298
RAC: 377
United States
Message 1261 - Posted: 15 Feb 2018, 0:49:39 UTC

Yep the bucket is empty again.... no WU's available for distribution. All services show as running.

Bill F
Dallas TX
____________

Profile valterc
Project administrator
Project tester
Send message
Joined: 30 Oct 13
Posts: 360
Credit: 17,702,998
RAC: 5,731
Italy
Message 1262 - Posted: 16 Feb 2018, 14:10:01 UTC - in response to Message 1261.

We have a lot of experiments in the queue. The computational times are, in this case, very fast and the work generator is not able to fulfill all the users requests.

Profile Bill F
Send message
Joined: 3 Apr 17
Posts: 16
Credit: 156,298
RAC: 377
United States
Message 1263 - Posted: 17 Feb 2018, 3:11:27 UTC

Is there away to optimize the work unit generator to increase the amount of WU's per Hour / Day ?

Or can a second process be be created to have two work unit generators feeding the bucket if you have a lot or experiments queued ?

Thanks
Bill F
____________

Profile [VENETO] boboviz
Send message
Joined: 12 Dec 13
Posts: 147
Credit: 1,126,004
RAC: 368
Italy
Message 1265 - Posted: 22 Feb 2018, 17:22:48 UTC

Now i'm crunching a mix of wus: Escherichia, Vitis and Pseudomonas

:-)

Profile Coleslaw
Send message
Joined: 11 May 14
Posts: 5
Credit: 1,416,925
RAC: 195
United States
Message 1266 - Posted: 27 Feb 2018, 22:11:25 UTC

valterc, is the limitation a personnel issue or is it a limitation of the hardware that is preventing more work being created? Is this something the donors/users can assist with in some way?
____________

Profile valterc
Project administrator
Project tester
Send message
Joined: 30 Oct 13
Posts: 360
Credit: 17,702,998
RAC: 5,731
Italy
Message 1267 - Posted: 1 Mar 2018, 19:41:38 UTC - in response to Message 1266.

valterc, is the limitation a personnel issue or is it a limitation of the hardware that is preventing more work being created? Is this something the donors/users can assist with in some way?

The work generator is the main problem, it creates workunits in a time that doesn't depend on the execution time of them. So, if the running time of the organism/dataset/parameters combination that we need to run is 'fast' it is not able to fulfill all the users' requests. More powerful hardware will surely help but I guess that first we need to either completely rewrite the work generator or make a parallel version of it(not with the current hardware). BTW, we are doing some offline tests preparing an experiment with a new dataset/organism and we surely know that it will take a long time for it to complete, running times seem 'very slow'. In this case we will have some time (many months) before having this kind of shortage again (and hopefully to really solve this problem).

Profile Bill F
Send message
Joined: 3 Apr 17
Posts: 16
Credit: 156,298
RAC: 377
United States
Message 1268 - Posted: 12 Mar 2018, 16:17:35 UTC

We are back to zero work units available for down load. Since the Admin has not responded on this thread I will try to send a PM to him.

Bill F
Dallas TX
____________

Buro87 [Lombardia]
Send message
Joined: 23 Nov 16
Posts: 21
Credit: 1,215,875
RAC: 175
Italy
Message 1269 - Posted: 13 Mar 2018, 12:36:51 UTC - in response to Message 1268.

We are back to zero work units available for down load. Since the Admin has not responded on this thread I will try to send a PM to him.

Bill F
Dallas TX


he has answered in the post before your's

Profile valterc
Project administrator
Project tester
Send message
Joined: 30 Oct 13
Posts: 360
Credit: 17,702,998
RAC: 5,731
Italy
Message 1270 - Posted: 15 Mar 2018, 10:27:16 UTC - in response to Message 1269.

This afternoon we will have a meeting to discuss various issues, among them the new big experiment and the efficiency of the system. I will update you all after that.

Profile valterc
Project administrator
Project tester
Send message
Joined: 30 Oct 13
Posts: 360
Credit: 17,702,998
RAC: 5,731
Italy
Message 1271 - Posted: 20 Mar 2018, 16:57:59 UTC - in response to Message 1270.
Last modified: 20 Mar 2018, 16:58:24 UTC

This afternoon we will have a meeting to discuss various issues, among them the new big experiment and the efficiency of the system. I will update you all after that.

I have just generated a few workunits related to the new experiment, just a test, for the moment, checking if everything will work well.

Buro87 [Lombardia]
Send message
Joined: 23 Nov 16
Posts: 21
Credit: 1,215,875
RAC: 175
Italy
Message 1272 - Posted: 20 Mar 2018, 21:52:12 UTC - in response to Message 1271.

This afternoon we will have a meeting to discuss various issues, among them the new big experiment and the efficiency of the system. I will update you all after that.

I have just generated a few workunits related to the new experiment, just a test, for the moment, checking if everything will work well.


i have done 4 of them. around 2h 45min on ryzen 5-1600

Buro87 [Lombardia]
Send message
Joined: 23 Nov 16
Posts: 21
Credit: 1,215,875
RAC: 175
Italy
Message 1275 - Posted: 3 Apr 2018, 7:53:45 UTC - in response to Message 1271.

This afternoon we will have a meeting to discuss various issues, among them the new big experiment and the efficiency of the system. I will update you all after that.

I have just generated a few workunits related to the new experiment, just a test, for the moment, checking if everything will work well.


Hi Valter, any updates from the new project?

1 · 2 · Next
Post to thread

Message boards : Number crunching : Server Status - No Work Available


Main page · Your account · Message boards


Copyright © 2018 CNR-TN & UniTN