1. 26 Jan, 2010 1 commit
  2. 13 Nov, 2009 1 commit
    • Mike Hibler's avatar
      Tweaks to get NTP working again. · f34f7abc
      Mike Hibler authored
      Somewhere around fedora10, our ntp.conf file was causing ntp1.emulab.net
      to no longer be able to update the clock on nodes.  Needed to un-restrict
      access from that server.
      
      Also, we were not going through our ntpd startup for f10 nodes.
      f34f7abc
  3. 22 Oct, 2009 1 commit
  4. 18 Aug, 2009 1 commit
    • Ryan Jackson's avatar
      Lots of linux "MFS" changes · 93a3a065
      Ryan Jackson authored
      * linuxmfs: (27 commits)
        Use LookupByIP in the local area case
        Don't check server cert in MFS environment
        Add wipe-all-disks-on-reloading code
        Generate random UUID for linux root fs
        Cleanup of mdev.conf; sane device permissions
        New root password
        add kmem group
        Make sure ETHERTYPE_LOOPBACK is defined
        Add teachswitch to the build
        Add support for old GNU tar format
        New root passwd
        Make 'permissions' target work again
        Fixes to device permissions
        Add osconfig to the build
        Add the file command to build
        Make osconfig work on linux MFS
        Detach any loopback devices left over after mounting initrds.
        Fix for handling ide devices as swap partitions
        Make sure F10's dhclient state gets wiped
        Make sure to reread partition table after writing image
        ...
      93a3a065
  5. 06 Nov, 2007 1 commit
  6. 18 Oct, 2007 1 commit
    • Mike Hibler's avatar
      Preliminary support for Ubuntu Linux. · 38bc8fa1
      Mike Hibler authored
       * added new tmcd directory with Ubuntu (really, Debian) specifics
       * fixed up GNUmakefiles to not do "-g wheel" when creating directories
       * other, relatively minor tweaks
      38bc8fa1
  7. 25 Apr, 2007 1 commit
  8. 16 Jan, 2007 1 commit
  9. 11 Jan, 2007 1 commit
  10. 13 Mar, 2006 1 commit
    • Leigh B. Stoller's avatar
      A set of changes to run "prepare" on a node just prior to an image · d8f8f9b4
      Leigh B. Stoller authored
      being taken.
      
      The basic strategy is to have node_reboot (when -p option supplied)
      invoke a special command on the node that will cause the shutdown
      procedure to run prepare as it goes single user, but before the
      network is turned off and the machine rebooted. The output of the
      prepare run is capture and send back via the tmcd BOOTLOG command and
      stored in the DB, so that create_image can dump that to the logfile
      (so that the person taking the image can know for certain that the
      prepare ran and finished okay).
      
      On linux this is pretty easy to arrange since reboot is actually
      shutdown and shutdown runs the K scripts in /etc/rc.d/rc6.d, and at
      the end the node is basically single user mode. I just added a new
      script to run prepare and send back the output.
      
      On FreeBSD this is a lot harder since there are no decent hooks.
      Instead, I had to hack up init (see tmcd/freebsd/init/{4,5,6}) with
      some simple code that looks for a command to run instead of going to a
      single user shell. The command (script) runs prepare, sends the output
      back to tmcd, and then does a real reboot.
      
      Okay, so how to get -p passed to node_reboot? I hacked up the
      libadminmfs code slightly to do that, with new 'prepare' argument
      option. This may not be the best approach; might have to do this as a
      real state transition if problems develop. I will wait and see.
      
      Also, I changed www/loadimage.php3 to spew the output of the
      create_image to the browser.
      d8f8f9b4
  11. 13 Dec, 2005 1 commit
  12. 02 Dec, 2005 1 commit
    • Mike Hibler's avatar
      Remove /var/lib/rpm/__db* on linux. · 7e5633a6
      Mike Hibler authored
      These files get regenerated on every boot anyway
      (i.e., rc.sysinit removes them and they keep coming back!)
      
      Note: I don't spend all my time analyzing the RHL boot sequence, I noticed
      this as a consequence of looking at what changed in a "delta" disk image
      created with the new imagezip.
      7e5633a6
  13. 17 Jun, 2005 1 commit
  14. 19 May, 2005 1 commit
  15. 14 Jul, 2004 1 commit
    • Mike Hibler's avatar
      1. Better control net detection. · ff2335dd
      Mike Hibler authored
         No longer rely on looking at kernel boot time messages and extracting
         a hardware signature to determine the nodetype to then determine the
         control net.  Now we just DHCP on all interfaces and decree that the
         interface that answers is our control net interface.
      
         An extraordinary number of sleezy tricks were needed to get FBSD4,
         FBSD5, and RHL to DHCP on all interfaces without changing any standard
         scripts.
      
         For now, the nodetype/cpuspeed/chipset scripts still exist for the
         benefit of healthd, which uses the output of nodetype to determine
         what kernel module to load.  We should fix this.
      
         Side-effect: pump, the old RHL DHCP client, is history!  For older
         RHL releases, you will need a version of dhclient.
      
         Side-effect: in Linux, all non-control net interfaces are left up
         but without a legit IP address.  This is a consequence of dhclient.
         In FBSD, it was trivial to clean this up, RHL will take a little
         more work.  Up or down, it shouldn't matter.
      
      2. Add an mfs-install make target, a scaled-down version of the client
         install.  Added a mandatory DESTDIR check so you don't accidentally
         install in the wrong place on boss.
      ff2335dd
  16. 24 Jun, 2004 1 commit
    • Mike Hibler's avatar
      Improve the client-side install. With these changes, it should now be · 976133e4
      Mike Hibler authored
      possible to:
      
      	gmake client
      	sudo gmake client-install
      
      on a FBSD4, FBSD5, RHL7.3, and RHL9.0 client node.
      
      There are still some dependencies that are not explicit and which would
      prevent a build/install from working on a "clean" OS.  Two that I know of are:
      you must install our version of the elvin libraries and you must install boost.
      976133e4
  17. 17 Feb, 2004 1 commit
    • Leigh B. Stoller's avatar
      Medium size cleanup of the client side code. The main impetus behind · 069dc7d3
      Leigh B. Stoller authored
      this was to add soft reconfig support so that nodes could be
      reconfigured without having to reboot them. This appears to work, and
      has been tested with jails getting moved around. I've also tested the
      new code on the MFS, but still no testing has been done on PLAB nodes.
      
      The main change is that most of the code moved out of libsetup.pm, and
      was split into constituent rc scripts, each of which does its own
      thing, including cleaning up and preparing for making an image. Most
      of that central knowledge has been moved out into the scripts. Still more
      to do but this was a good start.
      069dc7d3
  18. 20 Jan, 2004 1 commit
  19. 24 Jun, 2003 1 commit
  20. 04 Apr, 2003 1 commit
  21. 15 Jan, 2003 1 commit
  22. 06 Jan, 2003 1 commit
  23. 18 Dec, 2002 2 commits
  24. 26 Sep, 2002 1 commit
  25. 12 Sep, 2002 1 commit
  26. 27 Aug, 2002 1 commit
  27. 10 Jul, 2002 1 commit
  28. 07 Jul, 2002 1 commit
  29. 19 Jun, 2002 1 commit
    • Robert Ricci's avatar
      Changes for the new image: · d07dae4f
      Robert Ricci authored
      First, we'll no longer have seperate sup releases for each revision,
      so the supfiles use the release 'current'
      
      Second, since we no longer have seperate sup releases, the prepare
      scripts no longer need to remove checkouts files.
      d07dae4f
  30. 11 May, 2002 1 commit
  31. 08 May, 2002 1 commit
    • Leigh B. Stoller's avatar
      Support for vnodes, plus some other changes: · 386c2778
      Leigh B. Stoller authored
      1) Merge in the accounts code that I did for ron. Instead of resetting
         the password file on each reboot, look at the node status to
         determine if the password/group file should be reset. If the
         node is free, reset it. Otherwise, we track changes to the password
         and group file now, so that users can change it and not have their
         changes wiped out at each reboot. I had to do this for the ron
         nodes so that the testbed software would not alter or delete
         already existing accounts; I keep a couple of little dbm files
         listing all the accounts/groups added. The only downside right now
         is if a node is reallocated before it is wiped clean; I plan to add
         an os_teardown phase to experiment termination asap.
      
      2) Add tunnels support. New DB table (tunnels) provides information
         for running vtund to link up to remote nodes. Creates a vtund.conf
         file on the fly, and fires them off. The complication is that you
         cannot do the ifconfigs or the routes until the tunnels are
         connected, so that stuff has to be configured within the vtund.conf
         file on a per tunnel basis. vtund.conf has some sections for
         running commands when tunnels are brought up or down.
      
      3) Damage the routing configuration code that Mike did. To support
         tunnels, as noted above, rc,route is no longer a simple list of
         commands, but a program that adds/dels routes based on the netmask,
         with a special "enable" section for the other stuff. This allows me
         to call it from vtund.conf for up/down on each tunnel'ed
         interfaces, as needed. Quite gross, but no way around it.
      
      4) For remote nodes, add a vnodesetup script, invoked from boss when
         experiments are setup/torndown. This gets the tunnel/route/trafgen
         configuration and runs them. It then goes into the background
         waiting for a death signal, at which time it brings them down and
         cleans out the vnode state.
      386c2778
  32. 22 Mar, 2002 1 commit
  33. 24 Sep, 2001 1 commit
    • Leigh B. Stoller's avatar
      The brave new world of client side configuration. Major highlights are · cedf74ed
      Leigh B. Stoller authored
      a "cleaner" separation of the OS independent and OS dependent parts.
      Lots of stuff moved to common libsetup.pm, shared by linux and
      freebsd. OS dependent stuff is in liblocsetup.pm in the subdirs.
      Other major change was to add "update" ability so that mounts and
      accounts can be updated on the fly, when directed to do so from boss.
      This is in support of shared experiments, so that as projects are
      added to the share list, the mounts and accounts can be updated
      without having to reboot.
      cedf74ed
  34. 23 Jul, 2001 1 commit
  35. 20 Jul, 2001 1 commit
  36. 06 Jul, 2001 1 commit
  37. 05 Jul, 2001 1 commit
  38. 16 May, 2001 1 commit
  39. 09 Apr, 2001 1 commit