Recorders in parallel versions

This forum is for issues related to parallel processing
and OpenSees using the new interpreters OpenSeesSP and OpenSeesMP

Moderator: selimgunay

Post Reply
andreascj
Posts: 113
Joined: Mon Aug 17, 2009 5:57 am
Location: T

Recorders in parallel versions

Post by andreascj » Mon Apr 07, 2014 9:41 am

Hi all,

I noticed that when OpenSeesSP is used, node reaction recorder does not record reactions at all the nodes using the following command:

recorder Node -file $dataDir/BaseRxnsDOF1.out -time -node 1 6 11 16 21 26 31 36 41 46 51 56 61 66 71 76 -dof 1 reaction

Reactions for some of the nodes are zero using above for np > 1. It is fine for np = 1. Similarly, the OpenSees manual states that drift recorder also does not work for np >1. So I was wondering which recorders do and do not work in parallel versions. Is there any other way to record reactions in OpenSeesSP with np>1?

Thanks.

fmk
Site Admin
Posts: 5883
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Recorders in parallel versions

Post by fmk » Tue Apr 08, 2014 7:40 am

that's a bug that needs to be fixed. i will need the script as it works for the few examples i have.

andreascj
Posts: 113
Joined: Mon Aug 17, 2009 5:57 am
Location: T

Re: Recorders in parallel versions

Post by andreascj » Tue Apr 08, 2014 6:31 pm

Thanks fmk. I sent you an email with the scripts.

andreascj
Posts: 113
Joined: Mon Aug 17, 2009 5:57 am
Location: T

Re: Recorders in parallel versions

Post by andreascj » Mon Apr 14, 2014 3:40 pm

Any thoughts fmk?

fmk
Site Admin
Posts: 5883
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Recorders in parallel versions

Post by fmk » Tue Apr 15, 2014 12:56 pm

not a quick fix .. basically the reaction recorder is the 1 recorder where i need to add duplicates instead of throwing them away .. it means changing a few interfaces .. it is one of the last things i need to finish before i officially release the 2.4.4 version (a 5730 revision is being tested now, the revision with your fix will have a greater revision number)

andreascj
Posts: 113
Joined: Mon Aug 17, 2009 5:57 am
Location: T

Re: Recorders in parallel versions

Post by andreascj » Wed Apr 16, 2014 6:05 am

Thanks fmk. Do you then think that adding up the bottom story element forces will give correct reaction for any number of processors (until the reaction recorder is fixed)?

fmk
Site Admin
Posts: 5883
Joined: Fri Jun 11, 2004 2:33 pm
Location: UC Berkeley
Contact:

Re: Recorders in parallel versions

Post by fmk » Wed Apr 16, 2014 8:21 am

yes.

andreascj
Posts: 113
Joined: Mon Aug 17, 2009 5:57 am
Location: T

Re: Recorders in parallel versions

Post by andreascj » Wed Apr 16, 2014 4:43 pm

Thanks.

Post Reply