Custom Query (894 matches)

When running 'bcfg2-admin compare -r' on an incorrect directory produces the following traceback:

# bcfg2-admin compare -r bcfg2-0.8.7 bcfg2-0.9.2 Entry: tg-v070.uc.teragrid.org.xml Traceback (most recent call last):

File "/usr/sbin/bcfg2-admin", line 487, in ?

do_compare(sys.argv[2:])

File "/usr/sbin/bcfg2-admin", line 223, in do_compare

state = do_compare([oldd + '/' + item, newd + '/' + item])

File "/usr/sbin/bcfg2-admin", line 238, in do_compare

new = lxml.etree.parse(new).getroot()

File "etree.pyx", line 1230, in etree.parse File "etree.pyx", line 157, in etree._parseDocument File "parser.pxi", line 153, in etree.Parser.parseDocFromFile

IOError: Could not open file bcfg2-0.9.2/hostname.domainame.xml

[[email protected]:~/software/bcfg2-0.9.2/src/sbin]$ sudo bcfg2-admin Traceback (most recent call last):

File "/usr/sbin/bcfg2-admin", line 473, in ?

if sys.argv[1] == "init":

IndexError?: list index out of range

Now that the doc is moving from XML files to the Wiki, there needs to be some way of printing out the Wiki documentation, as there is 50+ pages of docs, and reading large docs on surfaces other than paper can be annoying.

The easiest solution I have found involves 2 Trac plugins and HTMLDOC:

1. apt-get install htmldoc
2. Install the ​WikiIncludePlugin
3. Install the ​PageToPdfPlugin, after patching it it to allow a TOC to be generated with the patch in ​this ticket
4. Modify the trac.ini with the [pagetopdf] options in the above ticket, as well as with these lines in the [components] section:

   pagetopdf.* = enabled
   wikiinclude.* = enabled
   

5. /etc/init.d/apache2 reload

I like doing (2) and (3) like this to maintain easy compatibility with bcfg2:

1. fakeroot python setup.py bdist_rpm
2. cd dist; fakeroot alien -d *noarch*rpm
3. install .deb via bcfg2

This solution will get you something that ​looks like this. The one downside of this is that there will not be a title page - that should be relatively easy to patch ​PageToPdfPlugin to do; the code to get the title info from Trac itself is already in the ​CombineWikiPlugin. I have an ​example of bcfg2 doc created with this; it was running from another trac wiki, and so pulled in the crap that frames the content text; that would not be the case within a wiki using the include plugin.