LWG Minutes 2018-05-31

From OpenSFS Wiki
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Attendance

Cray: Cory Spitz, Ben Evans
ORNL: James Simmons, Dustin Leverman
HPE: Olaf Weber, Christopher Voltz
IU: Ken Rawlings
Intel: Joe Gmitter, Peter Jones, Andreas Dilger, Amir Shehata

Actions

New Actions Captured:

  • None

Existing Open Actions:

  • None

Actions Recently Closed:

  • None

Minutes

2.12.0 Release Update
Peter

  • Oleg has continued with landings and is testing more patches in master-next.
  • Any report out on any testing efforts with current master?
    • Cory: We have picked up our testing effort on master, but nothing serious found. Things are looking pretty good at the moment.

Upstream Lustre Client
James

  • We have had follow-on discussions to the recent events of being removed from staging and those conversations are going well.
  • Cory: Are we all satisfied with this approach of Neil hosting on GitHub?
    • Peter: If Neil is putting in the time, it makes sense for him to dictate some things around setup since he is part of the kernel circle.
    • James: The reason I support it is because I feel we can move much faster. The Intel branch moves slower and harder to work with for this purpose.
      • Peter: Yes, however, the priority on that branch is that we are always trying to keep it functional and working properly.
    • Andreas: The problem with this approach is that all the feature work will have to be redone on the GitHub branch. That is all of multi-rail, PFL, FLR, recent LFSCK improvements, etc... It is probably over 100K lines of code that will need to be reviewed again in this approach. It seems like resources are going to be a hurdle with that approach.
    • Andreas: Why not use the newer tree if we are not that far off?
    • James: It isn't worth the effort to do the rebase, we are not that far off and James/Neil believe velocity will be faster with this approach.
    • Andreas: One of the dangers of this approach is that if James/Neil/others spend a ton of time on the GitHub branch and it still takes a year or more to get into the kernel. You will then have to either spend a ton of time syncing back and forth or the GitHub branch becomes out of date and we are back into a similar position.


lustre.org
Ken Rawlings

  • Finalizing the old wiki migration remains a priority.
  • The next steps is to work on the guidelines page with Shawn Hall.
  • Cory: John Hammond maintains a patch status page. Is that page documented as a tool on lustre.org as part of the development process? If it is not, then we should remind people of this.
    • Ken: I will review the instructions and get it in there if it is missing.


Other Business

  • None


Next meeting will be on 2018-06-28 at 11:00am Pacific