It seems that in the new drmf2017 instance, port forwarding through vagrant is not set up. When I try to access localhost:8081 through curl on drmf2017, it does not connect to anything, while if I do the same in vagrant, it works fine.

==> default: Error: Could not start Service[tomcat7]: Execution of '/etc/init.d/tomcat7 start' returned 1:
==> default: Error: /Stage[main]/Drmf::Mathosphere/Service[tomcat7]/ensure: change from stopped to running failed: Could not start Service[tomcat7]: Execution of '/etc/init.d/tomcat7 start' returned 1:

I get this error when Puppet tries to start tomcat. Everything else works perfectly for mathosphere (it builds, and when it is deploying, the error is that tomcat is not started). I think this could be improved by using the tomcat plugin for puppet, but I'm not sure how to add that.

It is also odd that when I try starting tomcat manually using the same command it seems to work perfectly fine. Also when I try to run the deploy command manually it claims missing dependencies from basex and such. It seems to provide different errors when it is running through puppet and manually.

I get this error:

==> default: Error: Could not find class drmf::mathosphere for mediawiki-vagrant.dev on node mediawiki-vagrant.dev
==> default: Error: Could not find class drmf::mathosphere for mediawiki-vagrant.dev on node mediawiki-vagrant.dev

This is likely my mistake (probably didn't remove everything to do with mathosphere). I do not know if this is worth fixing since the mathosphere pull request is basically ready to merge at this point.

After I added the instruction for running vagrant up after vagrant provision, I discovered more issues. After I initialized the DRMF submodule and reprovisioned it, I got another error:

There was an error loading a Vagrantfile. The file being loaded
and the error message are shown below. This is usually caused by
a syntax error.

Path: /srv/mediawiki-vagrant/Vagrantfile
Line number: 41
Message: RuntimeError: The mediawiki-vagrant plugin hasn't been installed yet. Please run `setup.sh`.

After I ran the setup.sh and tried it again, I got

The provider 'lxc' could not be found, but was requested to
back the machine 'default'. Please use a provider that exists.

Error: /Stage[main]/Npm::Globals/Npm::Global[grunt-cli]/Exec[npm_global_grunt-cli]/returns: change from notrun to 0 failed: /usr/bin/npm install -g grunt-cli returned 1 instead of one of [0]

This error continues to occur after multiple tries at vagrant provision. Even if the command is run manually, it fails. It seems to be complaining about some /usr/local/bin/grunt existing before the install, and it says it doesn't want to delete that so it fails here. Since this is high up the dependency tree, it seems to be the cause of everything not working (like the wikidata not working and the webpage not being served).

The odd thing is that this isn't even in code from this repo - it's part of the actual npm module for puppet. Also, this is exactly the same as it is on drmf2016 and still this problem only occurs on drmf2017. EDIT: Apparently it does occur on drmf2016 as well but it didn't when the vagrant machine was originally provisioned (so the wikidata stuff generated fine).

@physikerwelt have you encountered something like this before?

as discussed with @HowardCohl on feb 3rd

It seems that MediaWiki isn't running. There are no issues related to this when provisioning the Vagrant machine, and the MySQL databases seem to be in place for the wikidata, Apache is running (on port 8080), and yet I can't actually get a webpage at that port - not even an error page. Before, it seemed like there were dependency errors in the vagrant provision process, but that doesn't seem to be applicable now. Even Mathosphere is running (but I haven't gotten a chance to test the search engine functionality).

@physikerwelt This is where I've been stuck for some time now - it seems now that it's not a provisioning issue or a port forwarding issue but maybe an issue with some role not enabled or something. Other than that, I don't really know where to go from here. I have not yet put in the real data from drmf2016, but I don't think that's the issue since I'm not even getting an error page.

I would fix this myself but I apparently don't have permissions?

It says "puppet module to install a DRFM instance" at the top

This would probably be difficult to set up - travis would have to set up some mediawiki environment (or maybe we could keep one instance for these types of tests and do provisioning there to test over ssh)?