Skip to Content.
Sympa Menu

perfsonar-user - Re: [perfsonar-user] perfSONAR dashboards might need update

Subject: perfSONAR User Q&A and Other Discussion

List archive

Re: [perfsonar-user] perfSONAR dashboards might need update


Chronological Thread 
  • From: "Uhl, George D. (GSFC-423.0)[SGT INC]" <>
  • To: Andrew Lake <>
  • Cc: Brian Tierney <>, "" <>
  • Subject: Re: [perfsonar-user] perfSONAR dashboards might need update
  • Date: Mon, 4 May 2015 14:52:21 +0000
  • Accept-language: en-US

Andy,

You’re correct. I’m running a set of non-registered perfsonar test nodes.  Brian had mention a fix for the issue to me at the I2 Global Summit so I wanted to follow up.  MTU and connectivity are already standardized and known values for me so not having them displayed is no big deal.

It would be nice to have the traceroute path information accessible from the dashboard graph that’s run on the Central MA.  This probably requires installation of additional packages from the toolkit but I’m not exactly sure what set of packages are needed and I’d rather not guess.  Right now I get “Internal Server Error” when I go to the traceroute link on the performance graph.  My work around has been to go to the toolkit page of each individual pS test server and access the traceroute graph there but aving that path information accessible from the dashboard graphs though on the centralized MA would be a big help.

Thanks,
George

From: Andrew Lake <>
Date: Monday, May 4, 2015 at 10:13 AM
To: George Uhl <>
Cc: Brian Tierney <>, "" <>
Subject: Re: [perfsonar-user] perfSONAR dashboards might need update

Hi George,


Displaying the MU and capacity requires registration in the lookup service. I don’t remember all the names of your servers but i think i can only find one in the lookup service. If I recall correctly, I think some of your servers have limited access to outside networks, so it might just be that they can’t register.

Thanks,
Andy

On May 1, 2015, at 2:09 PM, Uhl, George D. (GSFC-423.0)[SGT INC] <> wrote:

Brian,

I did a yum update on my standalone, centralized MA/maddash server after we talked on Wednesday.  Here’s a list of the pS packages that were updated:

Apr 29 08:55:18 Updated: perl-perfSONAR_PS-serviceTest-3.4.2-4.pSPS.noarch
Apr 29 08:55:31 Updated: perl-perfSONAR_PS-Nagios-3.4.2-3.pSPS.noarch
Apr 29 08:55:31 Updated: perl-perfSONAR_PS-MeshConfig-Shared-3.4.2-6.pSPS.noarch
Apr 29 08:55:54 Updated: perl-perfSONAR_PS-MeshConfig-GUIAgent-3.4.2-6.pSPS.noarch
Apr 29 08:55:54 Updated: perl-perfSONAR_PS-MeshConfig-JSONBuilder-3.4.2-6.pSPS.noarch

The last update to the maddash server was one I did in March when I upgraded from version 1.2.1 to 1.2.4 of maddash-server.

My graphs still display Capacity and MTU unknown and when I hit the traceroute link, I get a Internal Server Error message.  This probably because I built the MA last Autumn from these directions: http://docs.perfsonar.net/multi_central_MA.html and didn’t get all the packages I need to fully integrate the dashboard with the MA installation.

Thanks,
George

From: Brian Tierney <>
Reply-To: "" <>
Date: Friday, May 1, 2015 at 1:16 PM
To: "" <>
Subject: [perfsonar-user] perfSONAR dashboards might need update

 
To Maddash admins:

You might want to run a yum update on your dashboard hosts to make sure you have the latest perfSONAR RPMs. 

We've fixed a couple minor bugs in the perfSOMAR packages that Maddash uses. For example, if your plots are missing the MTU size, a yum update will fix this.


 




Archive powered by MHonArc 2.6.16.

Top of Page