Ruby gems are still not safe to use

http://cristianobetta.com/blog/2013/02/02/ruby-gems-are-not-safe-to-use/

by cbetta • 13 years ago

58 43 13 years ago

helloamar• 13 years ago

Then y many are giving lot of hype?

tr4656• 13 years ago

Because Ruby is something they use and they want to fix the problem. If they don't highlight the problems, nothing will get solved.

taf2• 13 years ago

"Stop running code on gem install."

- this is a real issue. I've used rpm shell execution to modify sshd as well as other system components in order "install" additional software. http://web.archive.org/web/20090211040821/http://www.idle-ha...

as you can see from that archived post, it's very important to have trust of what you are installing. especially when you have to install with root permissions....

Seeing how many references exist to "sudo gem install blah"... this is very serious as it's a high reward if you're able to get your remote code executing with root privileges (assuming as most would not limit sudo access e.g. user ALL=(ALL) ALL )...

Xylakant4• 13 years ago

I don't see the big gain in stopping to run code on install. By definition, we install gems to run code. If we don't trust the gem author not to mess with our system on install, how can we trust him not to mess with our system when we use the gem? Granted, there might be some people that install gems as root and run them as unprivileged user only, but even as a non-root user it's a problem to run code you don't trust.

FooBarWidget3• 13 years ago

You're right, I totally didn't think about that. Securing the C compilation system could give a false sense of security.

What we should have instead is a good signing infrastructure to detect when trusted gems have been tampered by a third party.

h_r• 13 years ago

Well... you would have to trust both the gem author and those who might have compromised an author's credentials. And my understanding is that the install can easily be running under a different set of credentials than normal use. (not a Ruby or Rails user here)

FooBarWidget3• 13 years ago

You don't have to install with root permissions. If you run as normal user it will install to the home directory instead. But many people want to install gems system-wide so they install with sudo.

The real problem is executable code. Building C extensions typically require invoking arbitrary commands. The problem is also not unique to RubyGems: RPMs and DEB packages have preinstall and postinstall scripts, and they require root privileges.

I think a good solution would be to to run C extension compilation code as a sandboxed non-root user. If a RubyGem is being installed as a normal user, the compilation code should still be run as a separate, sandboxed user, to prevent it from messing with the user's home directory. Any build products that the compilation process generates will be copied over the destination directory. The sandbox user's home directory would be wiped after every installation.

This would severely limit the C extension building system's power (they can't generate files outside the gem directory etc without being wiped) but I think that's acceptable. Use cases that require more power can rely on external user-invoked commands, e.g. passenger-install-apache2-module.

wildchild• 13 years ago

There is also Gemfile that can FileUtils#rm_rf

qrush• 13 years ago

The call to action in this post is not strong enough - RubyGems and RubyGems.org are completely volunteer-run, open source projects. If you want to fix these problems, please get involved and stick around.

PommeDeTerre3• 13 years ago

Another approach is to completely move away from using Ruby, Ruby on Rails and related software.

I think that the recent security issues are evidence of many systemic problems within the Ruby community, and with their approach and attitude toward software development.

Security should be inherent and considered from the very start, rather than brought on over time by an endless stream of patches and updates.

Furthermore, the focus should not be on cranking out libraries and code as quickly as possible, especially when said code is rife with security holes.

There are many other programming languages, libraries and communities that take a far more sensible approach to software development. We see far fewer of these kinds of issues arise when things are not done the "Ruby" way.

phillc• 13 years ago

Your philosophies are sound for banking software.

There are cases where startups, social impact organizations, or any fast moving team would pick rails for its fast movement, accessibility, and support, even if they thought that there were even more security issues than that have happened.

BSousa• 13 years ago

Do you suggest an alternative? That works on Mac/Linux.

Is Python/Django that much secure or just not targeted enough?

I'm evaluating languages/frameworks for a project and I really want to use Haskell and yesod or happstack, but after starting my project in them, I always end up going back to Rails for the documentation/ease. I may try and stick to it this time but any suggestions would be great.

knowtheory4• 13 years ago

I'm loathe to engage in more negativity, but dude, you're just engaging in software-bigotry and trolling now.

You're making broad generalizations about the ruby community and it's members, many of whom do not fit your stereotypes.

Has the compromise of Rubygems been an event of such massive proportion that it effects all ruby devs and those who rely upon them? Yeah. Do things need to be fixed? Yes. Can these things be fixed within the Ruby community? Yes.

So if you want to advocate that people shouldn't use Ruby or Rails, fine, your prerogative. But please, stop being an asshole while doing it.

rimantas• 13 years ago

  > We see far fewer of these kinds of issues arise when
  > things are not done the "Ruby" way.

But not because other ways are safer and products are safe. They are just not as interesting for the HN crowd.

static_typed2• 13 years ago

Finally someone else gets it.

Security is not solved by a gem install makerailsmadsecurer.

Security is a process, and it does not stop.

How many people install gems happily without really understanding what it actually permits? Especially when run as root? How many people understood the always-on, Yaml parser approach that has been responsible for some of the recent security issues in Ruby land?

Given it is possible to write secure software and frameworks, why don't we see this in Ruby land?

hopsoft• 13 years ago

Removing the ability to run code on gem install would be quite disruptive. I think that establishing a universal gem signing policy and/or some form of whitelist/blacklist strategy would be a better solution. Consumers need to be able to trust the installations of the tools they use. The same risks apply to any other installation process. Think of how we install RVM or Homebrew.

grandalf• 13 years ago

Even if 5% of the rubygems ecosystem contained malware, the biggest danger to most projects is the inclusion of gems that are sloppily maintained. Just because something is released as a gem does not mean it has good code quality or that good development practices were used to create it.

The default behavior of bundler is to grab the latest compatible gem version, and in many cases this breaks things bc of little or no QA on the part of some gem maintainers.

The top 10% of gems are well maintained but the rest should generally be avoided.

nicholasjarnold2• 13 years ago

Is it safe to install rails with something like 'gem install rails' right now? I'm totally new to Ruby and to the Rails framework, but I was going to start a side project with it this weekend (today). Any advice on how I can safely get setup while the community is figuring out how to cope with the intrusion?

oscardelben• 13 years ago

yes, all gems have been verified.

nicholasjarnold2• 13 years ago

Thank you.

curcumin• 13 years ago

This hyperbole is very silly! Weaknesses appear in everything when it becomes popular.

There needs to be something like the "app store" and I don't mean specifically apples' own.

But we need some of the big corps using ROR to step forward and provide complete support for this type of project.

mark_l_watson• 13 years ago

Sorry in advance for being off topic, but: I rely a lot on Clojure repos like clojars.org and I in addition to checking my few Rails and Sinatra apps in the last few days, I have become a little concerned about the same sort of thing happening with clojars, main mavin repos, etc.

ef4• 13 years ago

Worrying about code execution at install is silly. The whole point of installing a gem is to download code that you're going to execute.

So the whole gem (install code and runtime code) needs to be trusted, and should be verifiably signed by somebody you can trust.

Turing_Machine• 13 years ago

Right. Some of these are legitimate issues, but not that one.

Given that the Ruby code in the gem has full access to the file system with the privileges of whoever is running it, I don't see how this makes things any worse (assuming you're not installing the gems as root or whatever).

static_typed2• 13 years ago

I think the overall points raised help shape the bigger conversation about the current state and implementation of Ruby Gems.

Who built your Gem? How do you verify that still holds? You may trust developer A who released a nice Gem, but what about when he pulls in a dependency, that pulls in another dependency, and suddenly you have gems from developer B, who loves to stick a Yaml parser out there for all to compromise.

The whole design needs a rethink.

Xylakant4• 13 years ago

But isn't that a problem where no good solutions exist? We share code to reuse code. If we insist to allow only one level of dependencies, then we restrict code reuse which is bad in other ways: you promote reimplementation of functionality, more often bad than good.

Ruby gems are still not safe to use

Related Stories