1. 09 Mar, 2018 2 commits
  2. 08 Mar, 2018 3 commits
  3. 28 Feb, 2018 1 commit
    • Mike Hibler's avatar
      Prevent a new VLAN from getting put into every existing trunk port. · 4a1d42b7
      Mike Hibler authored
      This is the flip-side of something the Mellanox module already handled:
      putting a port in trunk mode and having it inherit all existing VLANs.
      
      The only sure fire way to do this was to make sure, after creating a new
      VLAN, that every existing trunk port does not include the VLAN. If it
      does, we remove it.
      4a1d42b7
  4. 27 Feb, 2018 1 commit
  5. 24 Feb, 2018 1 commit
  6. 23 Feb, 2018 1 commit
    • Mike Hibler's avatar
      Make what I hope is a fix for portchannels. · 5d02e888
      Mike Hibler authored
      In the event we are called with a single port and we could not find
      an ifindex for the portchannel it is in, return the ifindex of the port
      instead (this is old cisco behavior). However, the code that did this
      was trying to lookup the ifindex of $swports[0] in the IFINDEX hash.
      But the swports array already contains the ifindex, so we don't need
      that IFINDEX hash lookup. At any rate, it works now!
      5d02e888
  7. 20 Feb, 2018 1 commit
  8. 16 Feb, 2018 1 commit
  9. 15 Feb, 2018 1 commit
  10. 05 Feb, 2018 1 commit
  11. 22 Jan, 2018 6 commits
  12. 01 Jan, 2018 1 commit
    • Leigh Stoller's avatar
      Changes to reservation system wrt classic interface: · dc90a087
      Leigh Stoller authored
      1. Reservation system now groks experiment lockdown and swappable. When
         swapping in, lockdown and swappable mean the expected end of the
         experiment is never.
      
      2. Reservation library now handles changes to lockdowm, swappable, and
         autoswap (timeout). editexp now hands these changes off to a new
         script called manage_expsettings, which can be called by hand since
         we might need to force a change (I am not changing the classic UI, if
         a change is not allowed by the res system, we have to do it by hand).
      
      3. Minor fixes to reservation library.
      dc90a087
  13. 13 Dec, 2017 3 commits
    • Kirk Webb's avatar
    • Leigh Stoller's avatar
      This started out as a simple change ... · 1d526637
      Leigh Stoller authored
      I noticed that group_root could not delete users from projects. Seems
      like we should allow that, but with the restriction that a group_root
      cannot delete another group_root. Simple enough, right? Well thats not
      how the permission system works; permission to do stuff to users is
      based on who you are in the project, not who you are doing it to.
      
      And then there are the subtle differences in permission handling between
      the Classic interface and the Portal interface. And I am fully
      unmotivated to fix anything in the Classic interface, hard to believe?
      
      Anyway, most people are not going to notice anything since the bulk of
      the changes affect sub groups. Sigh.
      1d526637
    • Leigh Stoller's avatar
      Another mysql 5.7 fixup. · 5bfec88d
      Leigh Stoller authored
      5bfec88d
  14. 22 Nov, 2017 2 commits
  15. 21 Nov, 2017 1 commit
  16. 19 Nov, 2017 1 commit
  17. 13 Nov, 2017 1 commit
    • Mike Hibler's avatar
      Speed up the (common) case where no nodes in an experiment are in the MFS. · 270878a0
      Mike Hibler authored
      On nfree, we call nfsmfs_setup will all nodes in the experiment even
      though it is unlikely that any of them are in the MFS. Even so, we were
      unconditionally doing a "zfs share -a" to make sure the exports table was
      up to date. This hurt hard in our environment where we have a zillion users
      and oodles of nodes (these are very precise thresholds!)
      
      Also, use "mount" to get the list of mounted NFS MFS filesystems rather
      than "zfs list" which likewise can be quite painful when there are a gazillion
      (1 gazillion > 1 zillion) ZFS filesystems.
      270878a0
  18. 09 Nov, 2017 1 commit
    • Mike Hibler's avatar
      Introduce a "failed" state for resource allocation. · 7e13f79b
      Mike Hibler authored
      If a background resource allocation fails, we put the lease in the "failed"
      state instead of destroying it. There were some ripple effects, specifically,
      the lease_daemon now checks for "failed" leases and send messages to us at
      the same frequency as for "unapproved" leases. The correct response here is
      almost certainly to destroy the lease, though you can put it back in the
      "unapproved" state (via modlease) and try to approve it to see what happened.
      
      Also add background mode to approvelease since it can do time consuming
      resource allocation.
      
      Nit: cleanup logfiles used in backgroud operation.
      7e13f79b
  19. 08 Nov, 2017 1 commit
  20. 12 Oct, 2017 1 commit
  21. 10 Oct, 2017 1 commit
  22. 04 Oct, 2017 1 commit
  23. 26 Sep, 2017 2 commits
  24. 22 Sep, 2017 2 commits
  25. 20 Sep, 2017 1 commit
  26. 12 Sep, 2017 2 commits