- 21 Jul, 2010 7 commits
-
-
Ryan Jackson authored
The toolchain build should be dependent on an actual file to ensure we don't rebuilt it for every target that goes into the initramfs.
-
Ryan Jackson authored
-
Ryan Jackson authored
Also rename old fake tcsh to tcsh.fake, and run a new script that symlinks tcsh and csh to tcsh.fake if tcsh doesn't exist. This way tcsh can be removed from the build but tcsh/csh users can still log in.
-
Ryan Jackson authored
-
Ryan Jackson authored
-
Ryan Jackson authored
-
Ryan Jackson authored
-
- 20 Jul, 2010 2 commits
-
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
- 19 Jul, 2010 12 commits
-
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
or more then one experiment, of course.
-
Mike Hibler authored
-
Mike Hibler authored
commit.
-
Mike Hibler authored
use when (re)loading a node.
-
Robert Ricci authored
-
Leigh B Stoller authored
<disk_image name="urn:publicid:IDN+emulab.net+image+GeniSlices//UBUNTU91-LAMP" /> In other words, allow "GeniSlices" for the project, so we can add images to be used from the rspec.
-
Ryan Jackson authored
Use 'git diff-tree --no-commit-id --cc' on merge commits to determine if there are any differences between the merge commit and its parents. This is the case if any new files were hand-modified or added as part of the merge (such as when conflicts were resolved). If git-diff-tree produces no output, the merge was trivial. If the merge was trivial, the commit message is checked for an empty body. If empty, the commit is skipped when generating the email. This way if the committer documents the merge in some way (list of changes, reason for the merge, etc.) the email includes that information.
-
Leigh B Stoller authored
-
Leigh B Stoller authored
impotent, except that actually nalloc the nodes, and retry if the nodes are sucked up by someone else inthe meantime. Also, in the CM, add better ssh key checking.
-
- 18 Jul, 2010 3 commits
-
-
Kevin Atkinson authored
added to pubs.html.
-
Eric Eide authored
-
Eric Eide authored
-
- 17 Jul, 2010 2 commits
-
-
Mike Hibler authored
-
Mike Hibler authored
-
- 16 Jul, 2010 9 commits
-
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
Resolve the slice/sliver credential and use that.
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
Leigh B Stoller authored
-
- 15 Jul, 2010 5 commits
-
-
Ryan Jackson authored
-
Jonathon Duerig authored
tuntest.py now takes three optional rspec files (examples are gec8*), one for each side of the tunnel and one providing the tunnel glue.
-
Ryan Jackson authored
-
David Johnson authored
Now, if experimenters add users or groups to their nodes, and take an image, they must explicitly tell prepare to overwrite the master passwd/group/shadow files in /etc/emulab. For Linux, os_account_cleanup tried to update the masters automatically, but this is not fullproof and can result in the masters getting passwd hashes from shadow. Now, a diff of the non-shadow files is printed from prepare so that at least people can see that the masters need to be manually updated.
-
David Johnson authored
-