What's your biggest pain point that can't be solved by software alone?

Howdy all,

For the past several months, as we’ve been working on Virtualmin 6, my focus has been on removing pain points: Making more things more automatic more of the time. I think we’ve done some good things in that direction during this development cycle, with more automation during installation, more reliable installations (with a few new bugs that crept in but have mostly been fixed now), and generally making managing a web server more manageable for more people. (That’s a lot of “more” in one paragraph!)

But, there’s a few recurring pain points that we can’t really resolve by software improvements alone. I’ve identified what I think are two of the biggest ones that we could solve with a low-cost hosted service that integrates automatically into Virtualmin for those folks who want it:

  1. Spam/Antivirus scanning relay. In the past couple of years it’s become a major pain point to host antivirus, in particular, locally. ClamAV has grown to over 500MB and there’s no reasonable way to shrink it and it is growing faster every day. Spam also, is better when you have a large set of spam/ham data to train your filters on. More users==better spam filtering. And, if we had a reasonable revenue stream and there were demand, we could consider adding higher end AV solutions to the mix. Users who have a lot of Virtualmin servers can solve this problem for themselves with Cloudmin and Cloudmin Services, but folks with one or two servers might be better served by sending us a few bucks a month and not having to worry about it.
  2. Hosted DNS. We get lots of DNS questions, and while some of them wouldn’t be solved by us hosting the DNS, others definitely would. Secondary DNS is challenging for folks with a single system. Dealing with redundancy, in general, is challenging. Managing DNS for a bunch of domains across multiple servers is difficult, too. Also there are advanced techniques, like AnyCast, failover, GeoIP-based balancing, and load balancing, that might be more easily solved by someone operating at a scale of thousands or hundreds of thousands of zones than by someone managing a few dozen. (And, some, like AnyCast, require significant capital to implement and maintain.)

The idea with either service would be you’d be able to click a button, setup billing info, and be done with that problem forever; it’d just work. We’d have to implement it in both Virtualmin and as an API on the server-side, which would be complex, but what users see would be a black box where they put their problems and we’d solve the problem.

I’ve researched the market for both of these types of services, and we could do it at competitive rates with existing solutions and make a modest profit, so it’s feasible to implement either one (though it’d take months to implement each, so we need to prioritize what would make the most people happy). Either would cost a few bucks a month, depending on how many users (for mail) or zones (for DNS) you have, and somewhat based on the volume of usage.

So, what are your pain points? Note that I don’t mean bugs. If you’ve got a bug, tells us about it in a new topic or issue and we’ll fix the bug. I mean big problems that merely making Virtualmin more powerful or easier to use won’t solve. Things that require more resources than a small company can marshal but we might be able to if we amortize it out across many users.

And, related, but not quite the same: What are you considering moving to the cloud, but don’t really want to deal with learning new interfaces and APIs and new processes to implement? What would you move into the cloud immediately if it looked and acted just like what you’re used to with Virtualmin and Webmin?

I’ve just floating some ideas here. I don’t know where I’m going with this, but after all these years we still get a lot of issues in a few areas; and I’d like to solve them. So, let’s focus in on what those are and how we can solve them.

Brainstorm away!



Been thinking over your post for a couple of weeks.
I Only have about 5 yrs experience with the mins software and I only have a few select clients, but I like the above ideas.

My thoughts on each.

Spam: Never had an issue with spam/virus, turned on at install and haven’t touched it since. I would pay to offload though, not so much because of memory, more because of the larger pool to draw info from.

DNS: IS a royal pain in general! Which it shouldn’t be. I have a domain name, point it to an IP. Thats all I should have deal with. If glue records point to correct name server, it should work. This not a virtualmin or webmin issue, just DNS in general is way over complicated in my opinion. Why are there 20 dns entries for one domain name! It should have the domain name and IP, thats all. Betting 90% of that complication is because of trying to prevent spammers/hijackers.

Unrelated to that rant on DNS, at a minimum a hosted secondary dns option would be great. Most people probably already spend $5/month on another VPS somewhere to set it up as their secondary. Why not you collect that 5$. Small admins like me have ns1 on the server, a VM as slave, then some sort of free dns secondary elsewhere, which has to be setup manually for each domain. Real pain. But a hosted secondary(with multiple server locations) would solve both issues if all I had to was check a box saying send to Joe’s secondary DNS . This would especially help those with just a virtualmin install and no way of running their own secondary.

Setting up the secondary and telling virtualmin to use it isn’t difficult. Bet the interface/API for who is allowed to connect the secondary would be the complicated part on your side.

Thanks for the great software and all the time you spend answering questions.

A *min wiki where we can update the documentation. So, the pain is lack of comprehensive documentation, the solution is users, and the conduit is software/policy.

We have a wiki (doxfer.webmin.com), but due to huge and completely unmanageable amounts of spam, I had to disable user account creation, so there’s only a few real users who actually do any editing (though a couple of the users who do contribute are incredible and have done tons of work to make it much nicer than it was when I first launched it). Having to lock it down really sucks, as we desperately need help with the docs (there’s just too much to cover, when you consider all of Webmin in addition to Virtualmin and Cloudmin) but it had become a real chore to clean it up. It had gotten to where I was spending an hour every day on spam clean up in the wiki. There was just no way to justify continuing to do that on the off chance someone might edit a doc in a positive way.

That said, I will be (super) happy to setup an editor account on the wiki for anyone who asks. Just open a ticket in the issue tracker asking for a Webmin wiki account, and I’ll ask Eric to ping me about any tickets about the wiki (I’m lately not able to spend much time monitoring the issue tracker, due to some personal issues, but Eric and Jamie are).