1 / 6

perfSONAR deployment

perfSONAR deployment. Simone Campana. People involved. In random order Shawn McKee Marek Zielinski Si Liu Stefan Roiser Simone Campana Mario Reale Good coverage of expertise for 3 experiment workflows and network/ perfSONAR No expertise for ALICE

kylee
Download Presentation

perfSONAR deployment

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. perfSONAR deployment Simone Campana

  2. People involved • In random order • Shawn McKee • Marek Zielinski • Si Liu • Stefan Roiser • Simone Campana • Mario Reale • Good coverage of expertise for 3 experiment workflows and network/perfSONAR • No expertise for ALICE • First meeting yesterday, no mailing list, this presentation are the minutes.

  3. Current Status • Deployment of perfSONAR • largely deployed in USATLAS and USCMS • Full mash of links tested *within* the cloud • deployed in the LHC OPN • Full mesh of links tested • deployed at many other ATLAS sites • Full mesh of links tested *within* the cloud • Some inter-cloud tests • Current methodology and configuration will not scale with more sites • Same link tested multiple times (by different experiments) • A lot of work in adding new sites to the configuration

  4. Next Steps (I) • Every experiment will define • a list of sites from where to start • groups priorities for links to be tested • Example for ATLAS: • 3 categories of sites: OPN (T0 and T1s), T2D, T2 and T3 • Priorities • Priority 1: OPN-OPN links • Priority 2: Tx-Tx links in the same cloud • Priority 3: T1-T2D links (different cloud) • Priority 4: T2D-T2D links (different cloud) • Priority 5: all other links • We would like to start from OPN sites and T2s which qualify as T2D.

  5. Next Steps (II) • We will push the deployment of the perfSONAR service at the sites • Start from the ones indicated by experiments • With a very basic configuration (like test the connectivity with “your” T1) • We will streamline the instructions (ATLAS and CMS instructions are very similar) • We will validate the new “Mesh Configuration” capability of perfSONAR • http://www.usatlas.bnl.gov/twiki/bin/view/Projects/PerfSONAR_PS_Mesh • We will define testing patterns (rates, number of packets, etc ..) for each of the priorities in previous slide • We will leverage the “Mesh Configuration” to propagate the testing patterns to all sites (starting from on cloud)

  6. Next steps (asynch) • We need to verify the compatibility between perfSONAR flavors. Will do. • We need to make perfSONAR a WLCG service • Publication in GOCDB/OIM, basic SAM tests etc. • Will do, but: • We need to do the same in other areas (xrootd, squids) • Can we have a discussion all together with GOC and OIM people? A contact name, mailing list? • We need to make sure the monitoring scales as well • Work going on for the BNL Dashboard • http://perfsonar.racf.bnl.gov:8080/exda/

More Related