Meeting Notes for 5 September 2007

Present: Liz, Eric, Mihael, Bob, Tom L., Tom J.
Absent: Mike W.

Issues with www11

The www11 machine is still not working properly. Mihael has not had time to work on this. The solution should be something more permanent which pings the various machines and updates the lists for machines to be included in the load-balancing. The perl script for load balancing should not be in someone's local area.

For the record, Tibi documented what's in place now: On the file /mcs/ contains a re-write rule which directs any incoming request to run the file = /home/tiberius/public_html/direct.cgi=, which is the load balancing script.

To change this to use some other location one would simply have to copy or move direct.cgi somewhere else and then change this .htaccess file to point to it. To add www11 to the pool once it is fixed one simply needs to add it to the servers array in this script.

Wiki/Forum Issue

(Carry over from last telecon)

Bob, Eric and Tom L. discussed this offline during the past week.

Eric is going to start by reproducing the PBwiki that the Fellows are currently using (see His will be based on MediaWiki, to which he will add file upload capability. The Fellows will try this and give feedback.

One part of setting this up is to determine which machine it should live on, and where to put the file up-load area. The uploaded files should be somewhere that is backed up, but Eric doesn't completely know yet what is and isn't backed up. Do we need more disk space?

Action Item: Eric will send an email to MCS systems to ask what is being backed up now (and he'll CC to the list).

Eric suggested a page in the developer’s wiki with information on what is backed up, or more generally how would we recover the different parts of the project in case of a failure?

Action Item: Eric will start such a page.

We should think about how to name wikis and other site components so we know what we are talking about and don't confuse them. Mihael suggests that having only one wiki will avoid such confusion, but Eric points out that we'll have at least two already: the current CI developers wiki (which you are reading right now) and the new wiki for the Fellows. Then there is the "glossary" Eric set up on the LIGO prototype site. So some thought on how to avoid confusion would be worthwhile. But using fewer wikis with more internal organization is certainly one thing that may help.

Mihael's refactored code

Mihael’s code testing – Mihael put in all of Nick’s changes. There is a problem with the geometries. Liz is continuing testing with IE on the PC.

LIGO Analysis

Eric’s success! Milestone: the LIGO analysis was run on www13 from the web interface. This now provides a working example that runs from start to finish. Now they can work on putting that on the grid. Mihael is hoping to show Eric an example of somthing, and Eric wants to clean up the LIGO code before showing that to Mihael.

Action Item: Eric and Mihael will each clean up the code they want to show to the other, and then get together via phone to go over it.

CMS Update

Tom L. and Dan need advice from the two Mikes on how to proceed. The current version is with SWIFT. Requires – 1) Dan’s Change of the code and 2) get the new e-Lab stuff working on SWIFT. We probably won't be able to add this by the Sept. 30th

LIGO Teacher's workshop

TomL and Eric reported on the workhop at Hanford Aug 13-14, in which 10 teachers participated. They got an intro to LIGO, the Hanford lab, ELabs and e-Labs, and the LIGO analysis tool (running from

Work still needed to be done on LIGO e-Lab documentation. Needs references for milestones and teacher pages. Dale is working on it.

Tom L. talked about the effectiveness of remote participation to workshops. Is this feasible? Eric was involved remotely in this workshop. Eric discussed a teacher from last year's workshop who participated remotely from Philadelphia, but she did not continue. DId she fail to continue because of the remoteness, or other reasons? We need to find out more about what works with remote participation and what doesn't.

LIGO is really set up to use the discussion site. As Tom described it, it provides a way for teacher and student to interact with "give and take".

We had a long discussion of how to capture the remarks about the data in the discussion (e.g., what should go into the metadata comments on data like we have in the Cosmic e-Lab.) Where is it appropriate to use metadata? How we might be able to get access to the kind of information in the discussion groups. Should we be relying on HTML in the metadata, or should we think about using the wiki more for references and glossary items? It would be nice to have a search of the comments associated with data. Currently it is not included in the search.

Liz suggested that now we need to concentrate on implementing what already exists and then think about if there are better ways to do this.

Eric pointed out that there is currently no connection between LIGO data and the VDS metadata system, though he will try to learn more about it as he and Mihael work on putting the LIGO analysis on the grid. Eric sees a need for a database for information about LIGO channels, data segements, data quality flags, and perhaps other things, but he doesn't know if this will be provided by VDS or he'll need to set something else up separately. (Currently using flat files for channel info, and not dealing with segments at all.)

-- Main.LizQuigg - 06 Sep 2007
-- Main.EricMyers - 07 Sep 2007
Topic revision: r7 - 2019-05-22, AdminUser
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback