1. 29 Aug, 2012 1 commit
  2. 28 Aug, 2012 1 commit
  3. 27 Aug, 2012 7 commits
    • Leigh B Stoller's avatar
      Switch to AUTOLOAD instead of all the explicit access methods. · 5e4f9e7f
      Leigh B Stoller authored
      Add metadata_url and imagefile_url support in image creation
      and lookup.
      5e4f9e7f
    • Leigh B Stoller's avatar
      Add lan table locking to assist with avoiding race conditions when · de0c1000
      Leigh B Stoller authored
      operating on shared lans.
      de0c1000
    • Leigh B Stoller's avatar
      Increase length of osname to 30 chars. · c2f63e4f
      Leigh B Stoller authored
      c2f63e4f
    • Leigh B Stoller's avatar
      Add support for experimental networks on the control network. · df048b11
      Leigh B Stoller authored
      So what if your testbed has a control network but no experimental
      network? In the past you were SOL, but with this commit you can now
      create links and lans on the control network that look just like
      an experimental network link/lan.
      
      To make this work, ptopgen sports a new option (-C) that will put the
      control network links and wires and switches into the ptop file.
      
      libvtop generally operates as normal, but need to arrange for the
      physical ports to be put into dual tag/trunk mode, where the native
      vlan is the Control network. This is done with by setting attributes
      on the lan table entry that indicate dual and what to use for the
      native vlan. snmpit looks for these attributes.
      
      There are a couple of places that use the stack name (Control or
      Experiment) to determine if a vlan is control or experimental. This is
      not longer truu, and so need to use an attribute in the lan table
      entry. 
      
      The last bit of the puzzle is that snmpit has to be careful when
      disabling trunking on these ports. When this happens, all vlans are
      cleard from the ports (by the device layer), including the Control
      network itself, which would make the node unreachable. I had to add
      some special cases to watch for that, and return the ports to the
      control network.
      
      To turn this on, create a ControlNetVlans and enable it. The mapper
      looks for this and passes the -C argument to ptopgen.
      
      Nothing special in the NS file, except you have to turn on vlan
      encapsulation; tb-set-vlan-emulation vlan
      
      No delay nodes, but linkdelays work okay. Works for openvz containers
      as well.
      df048b11
    • Leigh B Stoller's avatar
    • Leigh B Stoller's avatar
      7ed40187
    • Leigh B Stoller's avatar
      c6073582
  4. 23 Aug, 2012 1 commit
  5. 03 Aug, 2012 2 commits
  6. 01 Aug, 2012 2 commits
    • Leigh B Stoller's avatar
      This commit adds some simple support for using the Infiniband on the · 997b21b5
      Leigh B Stoller authored
      Probe Cluster. The problem is that the IFB is a shared network that
      every node attaches to, which can looks like an ethernet device that
      can ifconfig'ed. In other words, one big lan.
      
      But we still want the user to be able to create a lan so that they can
      interact with it in thei NS file like any other network.
      
      The NS syntax is:
      
      	set lan2 [$ns make-lan "node1 node2 node3" * 0ms]
      	tb-set-switch-fabric $lan2 "infiniband"
      
      The switch fabric tells the backend to do IP assignment for the
      specific global network. Yes, I tried to be a little but general
      purpose. Lets see how this actually turns out.
      
      This first commit treats the fabric as a single big lan on the same
      subnet.
      
      NOTE 1: Since the unroutable IP space is kinda small, but the Probe
      Cluster is really big, we can easily run out of bits if we tried to do
      assignment on virtual topos. Instead, fabrics get their IP allocation
      at swapin time, and the allocations are deleted when the experiment is
      swapped out. The rationale is that the number of swapped in
      experiments is much much smaller then the number of possible topos
      that can be loaded into the DB. Still might run out, but less likely.
      
      The primary impact of above is that IP assignments can change from
      one swap to another, but this is easy to deal with if the user is
      scripting their experiment; the IP allocation is available via the
      XMLRPC interface.
      
      NOTE 2: The current code allocates from a single big network, which
      makes it easy for users to mess each other up if they start doing
      things by hand. Ultimately, we want each lan in each experinent to use
      their own subnet, but that is going to take more work, so lets do it
      in the second phase.
      
      The definition of "network fabrics" is in the new network_fabrics
      tables. As an example for probe:
      
      	INSERT INTO `network_fabrics` set
      		idx=NULL,
      		name='ifband',
      		created=now(),
      		ipalloc=1, ipalloc_onenet=1,
      		ipalloc_subnet='192.168.0.0',ipalloc_netmask='255.255.0.0'
      997b21b5
    • Leigh B Stoller's avatar
      9155e015
  7. 30 Jul, 2012 1 commit
  8. 25 Jul, 2012 1 commit
  9. 24 Jul, 2012 1 commit
    • Mike Hibler's avatar
      Add a 'disabled' field to the subbosses table. · e08bfeec
      Mike Hibler authored
      This allows us to more easily disable a subboss in the event of a temporary
      subboss outage (e.g., hardware failure). Previously we would have to remove
      the related rows from the DB and restore them later.
      e08bfeec
  10. 23 Jul, 2012 1 commit
    • Mike Hibler's avatar
      Minor tweak to "waitmode" in libreboot. · 735758ce
      Mike Hibler authored
      Previously ISUP and TBFAILED were the two states that signified that a
      node had rebooted to the satisfaction of libreboot (with waitmode==1).
      Add RELOAD/RELOADING to that list since the frisbee MFS never sends either
      of ISUP or TBFAILED.
      
      Required a modification to TBNodeStateWait() to allow waiting for an
      op-mode/state combo as well as just a state.
      
      I made this change in anticipation that it would be useful for more
      responsive monitoring of failure in tbswap/os_load. But now I am no so sure.
      735758ce
  11. 17 Jul, 2012 3 commits
  12. 13 Jul, 2012 1 commit
  13. 11 Jul, 2012 1 commit
  14. 03 Jul, 2012 1 commit
  15. 02 Jul, 2012 4 commits
  16. 29 Jun, 2012 1 commit
  17. 26 Jun, 2012 1 commit
  18. 21 Jun, 2012 1 commit
  19. 12 Jun, 2012 3 commits
  20. 11 Jun, 2012 1 commit
  21. 07 Jun, 2012 1 commit
    • Leigh B Stoller's avatar
      New script, clone_image to simplify create/snapshot from a node. · b01c991d
      Leigh B Stoller authored
      clone_image is a wrapper around newimageid_ez and create_image, that
      simplifies the most common operation; creating a new imageid derived
      from the image/os that is currently running in the node, and then
      taking a snapshot of the node. So for example, if node pcXXX is
      running image FREEBSD, and you want to create a custom image from that
      node, all you need to do:
      
      	boss> clone_image myfreebsd pcXXX
      
      which will create the new descriptor, deriving everything from the
      FREEBSD image on the node, and then take a snapshot from pcXXX. If
      the descriptor already exists, just take the snapshot.
      
      So what if you do:
      
      	boss> clone_image FREEBSD pcXXX
      
      well, the image is always looked up in the project the node is
      currently attached to, so in fact a new descriptor is created in that
      project, and you do not actually overwrite an image from some other
      project. 
      
      I've added some locking to images to prevent concurrent snapshots.
      This seemed like a good idea since this script is going to be used
      from the ProtoGeni interface. More on this in another commit.
      b01c991d
  22. 06 Jun, 2012 2 commits
  23. 18 May, 2012 1 commit
  24. 17 May, 2012 1 commit