1. 23 Apr, 2019 7 commits
  2. 22 Apr, 2019 1 commit
  3. 19 Apr, 2019 2 commits
    • David Johnson's avatar
      Better handle systemd-networkd chattiness in control net search. · 126ef78e
      David Johnson authored
      systemd-networkd and friends have become very chatty.  This commit
      is about turning down the noise.  It also removes PreferredLiftime=forever
      because it is no longer valid where it used to be, and cannot be used
      apparently in the DHCP case.  Seems that the default is now "forever"
      anyway, so it's now irrelevant to us.  (Older systemd-networkds would
      set the address lifetime to the advertised lease.)
      We also only mark an iface with CriticalConnection=yes once that iface
      has been chosen as the control net.  We used to just mark them all
      in the udev helper so that we didn't have to modify the generated
      config after successful detection, but now systemd-networkd complains
      about bringing down a searched-but-not-control-net interface if
      it is critical.  So, avoid that.
      Finally, I added `-q` to our invocation of systemd-networkd-wait-online,
      and increased the timeout with which we call it.  Timeout increase is
      because we would get spurious event loop disconnect messages without it;
      and q to quiet it in other ways.  Ugh.
    • David Johnson's avatar
      Remove m2crypto from sslxmlrpc_server; enable ssl handshake timeout by default. · fccfee60
      David Johnson authored
      Now we rely on the builtin SocketServer and ssl modules.  This combination is
      basically feature-equivalent to m2crypto, for our purposes.  The hack that
      sets a socket timeout to prevent non-ssl clients tying up the server's main
      thread (see commit 381e67a3) remains, but is significantly easier.  The
      problem is that the ssl.SSLSocket.accept method combines both the accept()
      on the server socket, and the ssl handshake "accept", into one function call,
      so we don't get an opportunity to propagate the finite timeout from the
      server socket to the client.  Thus, we override SSLSocket.accept in our own
      derived class, and avoid using the ssl module's various wrap_socket helpers.
      It seems reasonable to enable this by default, especially now that we're
      no longer vulnerable to m2crypto version variance.
      I also added real argument processing to make it easier to run devel versions.
  4. 18 Apr, 2019 2 commits
  5. 15 Apr, 2019 5 commits
    • Mike Hibler's avatar
      Avoid a null pointer deref. · 64feff35
      Mike Hibler authored
    • Mike Hibler's avatar
      Tweaks to allow 40Gb bandwidth. · ca3f5497
      Mike Hibler authored
      Actually should allow up to 100Gb.
    • Mike Hibler's avatar
      Initial steps to enable jumbo frames on experiment interfaces. · 33beb373
      Mike Hibler authored
      This is just mods to the tmcd "ifconfig" command to include an MTU= arg.
      Right now we don't have anything in the DB for MTU, so tmcd is just returning
      "MTU=" which says to not explicitly set the MTU.
      It also includes the basic client-side support which I have tested on a
      physical interface with MTU=1500. Further changes will be needed to DTRT
      on virtual interfaces and their physical carrier interface.
      But the hope is to get the client-side part nailed down before the next
      set of images are rolled, so that we will be ready when support for the
      front-side (UI and DB state) get added.
    • David Johnson's avatar
    • David Johnson's avatar
      Fix breakage to raw xmlrpc mode in 13ee8406. · 535c8d7a
      David Johnson authored
      (The hack to get "raw" xml mode from xmlrpclib is quite different than
      for m2crypto.  Basically, the response is parsed in the Transport, so
      not only do we need a special raw input method on the ServerProxy, but
      also a custom "raw" transport that skips the parser.)
  6. 09 Apr, 2019 4 commits
  7. 05 Apr, 2019 1 commit
  8. 04 Apr, 2019 3 commits
    • Leigh Stoller's avatar
      Fix some annoying behaviour: · 5b30304c
      Leigh Stoller authored
      * Do not overwrite the cluster selection(s) when the reservation info
        comes in; that takes 5-10 seconds, and if the user has already made
        the selections, leave them alone.
      * More generally, we draw the cluster selectors multiple times, and we
        were losing the previous selection. Sure, if the user selects a
        cluster and the constraint checker says it is not allowed, then its
        okay to change it (might change my mind on this). But trying too hard
        is annoying.
      Get rid of a bunch of debugging console prints.
    • Leigh Stoller's avatar
      Repo based profile fix for bug David noticed; do not update the master · 66469988
      Leigh Stoller authored
      when the repo changes, unless its the manage profile page and its the
      profile creator (the same code is shared in different places).
    • Leigh Stoller's avatar
      Fix for using Mothership system images with the wrong domain. We used to · 36a048ec
      Leigh Stoller authored
      handle this, but I lost it in the recent image cleanup. Note that this
      will need adjusting if/when we decide to make system images non
      global (and thus require a credential).
  9. 03 Apr, 2019 5 commits
  10. 02 Apr, 2019 10 commits