Re: including PHP code from another server..
| From: | Alex Pojarsky | Date: | Mon, 26 Mar 2012 19:30:43 +0000 |
| Subject: | Re: including PHP code from another server.. | ||
| References: | 1 2 3 4 5 6 | Groups: | php.general |
| Request: | Send a blank email to php-general+get-317278@lists.php.net to get a copy of this message | ||
I understand what performance issues this brings, but as for security
was just a bit curious. You have just showed me what I was thinking
about, but you wrote it much better, clear and structured.
Thank you.
2012/3/26 Stuart Dallas <stuart@3ft9.com>:
> On 26 Mar 2012, at 17:41, Alex Pojarsky wrote:
>
>> Now, as the issue adressed and script removed, can you please explain
>> what exactly are the issues of using such approach? I mean security
>> ones, not performance.
>
> It's the wrong solution to a process and organisation problem. Ultimately it's not
> really a problem IF you control every part of the infrastructure. Rene clearly doesn't so it
> has implications for everyone sharing that infrastructure, and anyone using the applications hosted
> there.
>
> * It requires the host to enable allow_url_fopen which means every single script on the server
> is then able to include/require URLs. It just needs one of them to have a related vulnerability and
> suddenly people can execute arbitrary PHP code on the server.
>
> * Rene mentioned that the code is open source. This implies that the security risk is lessened
> because the code that is being made publicly accessible is already publicly accessible, so the
> opportunity for someone to find vulnerabilities already exists. It gets an order of magnitude worse
> if other people start ignorantly using his code because they're essentially giving him the
> ability to execute arbitrary PHP code on their server. Not good no matter how much he protests that
> he won't "be evil."
>
> * You specifically wished to exclude performance from the discussion, but scalability is
> potentially a big issue here and should not be completely ignored.
>
> I think the real issue for Rene is that of perceived complexity. The idea of having to manually
> keep many copies of the same code in sync is what leads to finding solutions like this one. This
> solution leads to unnecessary network traffic and introduces potential security risks that go way
> beyond your own code, and even if it's not a big issue now it has the potential to become
> catastrophic!
>
> I'd put a fair amount of cash on my guess that Rene is not using any form of source
> control. To me that is the best solution to this problem. Curtis mentioned rsync which will also do
> the job, but in my view you're nuts if you're not using some form of source control
> already, and building a largely automated process around that is trivial and automatically audited.
>
> Rene: please read a book / website / something on PHP security. Some things are important
> whether you believe they are or not.
>
> -Stuart
>
> --
> Stuart Dallas
> 3ft9 Ltd
> http://3ft9.com/