Date/Time
8am PT
Meeting Online Connections
The following is the BlueJeans Information:
BlueJeans ID: BJ 153579667
-----------------------------------
-----------------------------------
Want to test your video connection?
http://bluejeans.com/111
Meeting Recording
Not used
Attendees
- Julio Ibarra (JI)
- Unknown User (kollross) (MK)
Jeronimo Bezerra (JB)
Paul Wefel (PW)
- Jeff Kantor (JK)
Goals
- Coordination of networking activities for LSST use of ESnet
Discussion items
Time (PT) | Item | Who | Notes |
---|---|---|---|
1400 | Introduction and Context | Jeff Kantor | From the Oct 13th meeting between the LSST project and DOE the following action item emerged. Focus now is on #1. |
1415 | Discussion Items | All | Clarification of timeline Jeff Kantor ESnet participation in fy19, and LSST operations in fy2020. Brainstorm connection options: JB: PW proposed 3 options. 3 Options are described in PW’s LSST-LHN ESnet colab meeting notes, and copied here: Toward the end of October Matt, Jeronimo and Paul exchanged email brainstorming some ideas
JB: what can I do to help? PW: Referring to logical network diagram, not sure what circle means on the two paths from Atlanta to Chicago. JB: Circle means protection at the optical layer? Circle means two diverse paths. AmLight will connect to ESnet transponders. Ignore circle for now. PW: ESnet does not provide optical protection. Transport is what ESnet will provide. ESnet does not need to provide protection. JB: Create a p2p link at the optical layer. LSST NET will notify ESnet if link goes down. JB: When there’s a connection at Atlanta, can first activate at 10G. Circuit from Atlanta will come after connection from Boca Raton to Atlanta is established. It does not make sense to provision dedicated 10G for transport. PW: a 10G routed port can be provided for fy19. JK: Dedicated path diverse links not needed until 2021. JB: Will write all this down into a tech document. JK: In addition to technical document, an Implementation plan and cost for what tech doc says. Instrumenting the network paths: JK: Other network segments, instrumenting with pS and using MADDash. If we asked for this capability for the Atl to Chic segments, will that present any kinds of issues? PW: available now through ESnet portal. Will want this capability in 2021. JK: not specifically called out in the request. REquest can be revised if necessary. JB: Addressing to wrong group. From ESnet, the network service will be only optical layer. They will not be able to test. The LSST network will instrument the path, alongside the other active equipment. Collocation: JI: In Atlanta, can ESnet provide collo, or do we need to request from data center operator? PW: ESnet shares colo with I2. ESnet would not be able to provide colo. 600 W Chicago: available cabinets in the cage. JK: NCSA has multiple locations in Chicago. Do green and red paths in diagram show up in the same or different locations in Chicago? How does NCSA connect to Chicago for path diversity? MK: NCSA peers with ESnet at 600 West and 710 Lakeshore Dri. Connections are two optically diverse paths up to Chicago JB: in those locations can you provide colocation? MK: probably not. 600 space is limited. 710 might be more flexible. At 710 NCSA does not control those racks. JI: suggested PW take an action on the colo Document: Starting draft of tech document - PW volunteered to generate draft. Google Docs is preferred. DOE Process: JK: What is the DOE - ESnet process that this tech proposal becomes an actual thing? What steps will Inder and PW go thru to provide these links? PW: not sure of process at this time. Broad strokes, should involve proposal to ASCR. PW will flesh out steps. PW: it would be good to know what we will be asked to produce. Tech document and end2end plan should suffice. JK: if fy20 requirement creates a problem, can discuss. There’s flexibility. |
Wrap up and next meeting | Julio Ibarra | Wrap up and Next meeting date and agenda topics: Next meeting will be 7am PT, 10pm ET, 12pm CLT |
Action items
- Paul Wefel to create google doc for technical specification
- Julio Ibarra, Paul Wefel to create "big picture" document for DOE
- Julio Ibarra to find out colocation options in Atlanta