1. 12 Jun, 2014 5 commits
  2. 11 Jun, 2014 5 commits
  3. 10 Jun, 2014 1 commit
  4. 09 Jun, 2014 2 commits
  5. 06 Jun, 2014 9 commits
    • Leigh B Stoller's avatar
      Lessons learned from my first attempt to use this to debug a problem · 508bd9e5
      Leigh B Stoller authored
      for Nick! The output format needed some, uh, tweaks.
    • Leigh B Stoller's avatar
      Add uky2 · e975057d
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
      Updates. · 5946216b
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
      Comment out the range. · 3d5a3797
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
      New script, analogous to Mike's node_traffic script. Basically, it · b885ce89
      Leigh B Stoller authored
      was driving me nuts that we do not have an easy way to see what is
      going on *inside* the fabric.
      So this one reports on traffic across trunk links and interconnects
      out of the fabric.  Basic operation is pretty simple:
      	Usage: switch_traffic [-rs] [-i seconds] [switch[:switch] ...]
      	Reports traffic across trunk links and interconnects
      	-h          This message
      	-i seconds  Show stats over a <seconds>-period interval
      So with no arguments will give portstats style output of all trunk
      links and interconnects in the database. Trunk links are aggregate
      numbers of all of the trunk wires that connect two switches.
      The -i option gives traffic over an interval, which is much more
      useful than the raw packet numbers, since on most of our switches
      those numbers have probably rolled over a few times.
      You can optionally specify specific switches and interconnects on the
      command line. For example:
      boss> wap switch_traffic -i 10 cisco3 ion
      Trunk                    InOctets      InUpkts   InNUpkts   ...
      ----------------------------------------------------------- ...
      cisco3:cisco10                128            0          1   ...
      cisco3:cisco8                2681            7          4   ...
      cisco3:cisco1                4493           25          7   ...
      cisco3:cisco9                 192            0          1   ...
      cisco3:cisco4                 128            0          2   ...
      pg-atla:ion                     0            0          0   ...
      pg-hous:ion                     0            0          0   ...
      pg-losa:ion                     0            0          0   ...
      pg-salt:ion                  2952            0         42   ...
      pg-wash:ion                     0            0          0   ...
      NOTE that the above output is abbreviated so it does not wrap in the
      git log, but you get the idea.
      Or you can specify a specific trunk link:
      	boss> wap switch_traffic -i 10 cisco3:cisco8
      Okay this is all pretty basic and eventually it would be nice to take
      these numbers and feed them into mrtg or rrdtool so we can view pretty
      graphs, but this as far as I can take it for now.
      Maybe in the short term it would be enough to record the numbers every
      5 minutes or so and put the results into a file.
    • Leigh B Stoller's avatar
    • Leigh B Stoller's avatar
      Add current_speed(). · 6787124e
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
      Add LookupByType(). · 78f183ff
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
  6. 04 Jun, 2014 11 commits
  7. 03 Jun, 2014 1 commit
  8. 02 Jun, 2014 6 commits
    • Leigh B Stoller's avatar
      Do not drop privs before the second ssh, when taking an image of a · e92c073a
      Leigh B Stoller authored
      VM. Note that the target script uses sudo, but not in all the places
      it needs to, so lets just run the script as root, and yank the sudo
      out later.
    • Leigh B Stoller's avatar
      Minor fix to clean. · 2e78f216
      Leigh B Stoller authored
    • Leigh B Stoller's avatar
    • Mike Hibler's avatar
      Support for gathering and storing Infiniband interface GUIDs in the DB. · 12a41b7e
      Mike Hibler authored
      Since GUIDs are 16 bytes and our current interface MACs are only 12 bytes,
      I agonized over whether to grow the mac column to 16 bytes and just treat
      it as a unique identifier (which is all we use that column for anyway).
      However, in the end I just added a new guid column as there were mac columns
      in a variety of other tables and it wasn't clear what the relationship was
      and what I might break.
      So, the newnode MFS will now report back a GUID for interfaces it recognizes
      as IB (FreeBSD-specific right now). The boss-side checkin code with stash
      that value in new_interfaces (and later interfaces when added). For possible
      backward compat, it will also generate a MAC address from that (possibly
      Mellanox-specific) so that all entries in the interfaces table will have
      a MAC (yes, it should really be the other way around--all interfaces should
      always have a guid).
      End of story. We don't do anything else with IB right now other than stash
      an interface GUID.
    • Mike Hibler's avatar
      Yet another fix for FBSD 10.0. · 3add3502
      Mike Hibler authored
    • Leigh B Stoller's avatar
      Minor bug fix. · 8de31238
      Leigh B Stoller authored