Subversion module - connection issues

Centos 7
CentOS Linux 7.9.2009

I’m trying to get Subversion running and I’m having issues making it work. I have followed the basic setup
docs here Subversion and Virtualmin | Virtualmin and installed the webdav module (required for it to not explode Apache).

I have created a repository and a user and given the user read/write access to the repo. It seems like it all should work, but for the life of me I can’t make a client connect to it.

When I hit the URL https://my-domain.com/svn in a browser it gives me the output below. I have run “fix permission” on the repo, and the folder looks like it should be readable by apache.

<D:error xmlns:D="DAV:" xmlns:m="http://apache.org/dav/xmlns" xmlns:C="svn:">
<C:error/>
<m:human-readable errcode="2"> Could not find the requested SVN filesystem </m:human-readable>
</D:error>

Trying to run a checkout via the command line svn client runs into a similar issue.

From what I see around, this all should work swimmingly. Any idea what’s missing?

I did see one post about editing the subversion.config file but nothing I do to that seems to have any effect either

# WANdisco Subversion Configuration
# For more information on HTTPD configuration options for Subversion please see:
# http://svnbook.red-bean.com/nightly/en/svn.serverconfig.httpd.html
# Please remember that when using webdav HTTPD needs read and write access your$

# Needed to do Subversion Apache server.
LoadModule dav_svn_module     modules/mod_dav_svn.so
# Only needed if you decide to do "per-directory" access control.
LoadModule authz_svn_module   modules/mod_authz_svn.so

<Location /svn>
  DAV svn
  SVNPath /opt/repo/
  AuthType Basic
  AuthName "SVN Repo"
  AuthUserFile /opt/WANdisco/apache22/conf/svn.passwd
  Require valid-user
</Location>

I recommend you not use Subversion.

Hi Joe - any particular reason? Is that module no longer maintained or something?

There’s no reason to ever use Subversion in a world with git.

1 Like

Ah. That’s something of a religious argument, methinks. I will say, though, in my world where version control needs to be used by non-technical operators I find that the tooling for Subversion makes life (much) easier to onboard a newbie. Having to clean up after said newbie finding “fun” commands on stack overflow (git rebase) is not my idea of a good time.

Horses for courses.

I am in same line with @Joe, git is way easier and much better - no special extras need it, it just works.

OK. That hasn’t been my experience (and in my other job I wrangle a dozen or so non-technical users who have to make PRs and work with git at least a little bit). But, I guess let’s try to sort it out.

I haven’t managed a Subversion repository in 15 years, so I don’t have any experience to apply to the problem, so I probably don’t know what I’m doing here, but I know we need to see the Apache error log. The client can’t ever have the full picture. That’s probably got more information about what it’s trying to read, which may give better clues about why it’s failing.