Dear fellow ROLLOUT people,
Here is a summary of the discussion on the future of the SAM interface I had with David Collados and Maarten Litmaath.
I hope you have forgotten the original question I asked, if not you can skip ahead to the line with lots of equals characters.
The question: I wanted to be able to access, via some type of service, information like:
* what was the computed availability for VO "cms" (using cms's own list of tests declared as critical) for my site during some specified period ...
I can think of other good questions to ask, but the above is the main one in which I was interested.
I was planning to do this using the SAM interface, but figured this was not a great idea since SAM is on its way out.
That was the background, here is the summary of what i found out. Thanks again, David and Maarten and the SAM team!
======summary=================
The old SAM infrastructure is indeed being phased out, but there will be an equivalent API based on the Nagios based one. The API will be accessible from something called a MyEGI instance ('equivalent' to the old SAM Portal). These things are already appearing for those NGIs that have already rolled out the latest SAM-Nagios release, e.g.
NGI_CZ: https://nagios.egee.cesnet.cz/myegi
NGI_SK: https://nagmon.ui.savba.sk/myegi
ROC_Italy: https://mon-it.cnaf.infn.it/myegi
ROC_CERN: https://sam-cern-roc.cern.ch/myegi
The new interface will mimic the old SAM interface as much as possible. This old interface is documented here:
https://twiki.cern.ch/twiki/bin/view/LCG/SamDbQuery
You will note that the specific information I requested is not available via that interface, however the SAM team recognized the need for the numbers I asked for and it has been added as a feature request to their issue tracker.
The first versions of the new SAM programmatic interface for Nagios should be available sometime around January. Data will be returned in JSON or XML. For sites' availability & status, data will be available during Q1 2011, following the ACE (Availability Computation Engine) schedule presented to the WLCG MB yesterday:
http://indico.cern.ch/conferenceDisplay.py?confId=82017
=== end summary ===
J "I hope Gergo is still counting" T
|