问题
lately I have been trying to customize my Plone by adding helpful addons via the eggs section of my buildout.cfg, running buildout and restarting the zeocluster plonectl restart.
I have succesfully installed several addons this way.
But some stop Plone from working. For example trying to add plone.app.ldap:
buildout works fine and also the restart of the server works initially, but when accessing Plone in the browser it doesn't load and plonectl status tells me my two clients have lost connection to the zeoserver. The events log tells the same story, but other than that I do not see which error caused the problem. When i remove the addon everything works fine again.
Is the way I am handling my addons correct? Did I miss something? Where can I find additional information what crashes my zeocluster?
回答1:
Some problems are swallowed on startup. You can see them by starting your client on the foreground with bin/zeoclient fg.
See http://docs.plone.org/manage/troubleshooting/basic.html
Note: the zeoclient script may be called client, or client1 or instance or something like that, depending on how it is named in the buildout.
This will probably show a Python traceback. Maybe the add-on is missing a dependency. That would be a problem to report to the add-on authors. Adding the missing dependency to the eggs add rerunning buildout should fix the problem.
回答2:
Faced same problem. Here is how I fixed it:
Mention clearly the version you wish to install for the Add-on in the [versions] section of buildout.cfg.
[versions]
...
'add-on' = 'version'
Open the documentation of the add-on, probably on pypi resources, check for the versions. It so happens that plone automatically gets the version for you while running buildout and the add-ons, not compatible with your version of plone are installed. So it is always better to specify your version requirements.
来源：https://stackoverflow.com/questions/39199022/plone-restart-after-buildout-fails-how-to-find-the-error