Thanks for posting the link. I found this particular tidbit especially useful:
"The first is to create 'canary' accounts. Create accounts that have e-mail addresses, like "something-really-long-xyz-123@gmail.com". This account is not going to get any spam e-mail. When it does get its first spam, you'll know that it came from your database."
The problem with this advice is that it is not particularly useful.
In the event that your canary account receives a spam message, you will have gained one piece of data: that at some point in the months since you installed the account, it is likely that someone exploited a problem that exposed a table column in your database.
Now what?
You don't know what the bug was. You don't know how far-reaching the damage is. You don't know when the event occurred. There is, in fact, almost nothing actionable about this information at all.
You might be telling yourself two things:
* That, upon learning of the possible breach, you will lock down the security of your code and your site. Of course, think for two more seconds and realize this is ass-backwards; if you can reasonably lock down your site, you might as well do it before you burn your whole database.
* That, upon learning of the breach, you can get all your users to change passwords. Which, of course, you will never do, because you will be taking an enormous reputational hit over circumstantial evidence, and for the purposes of seeding your database with a couple thousand new passwords for the same attacker to steal.
People like this "canary" and "honey token" stuff because it's fun. But it's not valuable.
If I called you and said, "Hey, would you like me to tell you your site has been cracked?", would you really say "no"? I would argue that the cost of adding a canary is similar to the amount of time that phone call would take.
I'm guessing that after that call, you'd go through your logs and attempt to find out how the site was cracked. Then, if you find some anomalous behavior, you might be able to isolate the component that was being exploited and fix it. Maybe you wouldn't find it, but maybe you would.
What's your point? Sure. Set up a canary account. It won't hurt you. But the only way it will help is if your monitoring regime is already so good that it works with or without the canary.
But wouldn't the "canary" idea be more useful if the email was changed frequently? Suppose a cron job updated the email address every day to include a few non-obvious characters encoding the current date so that the spam you get tells you when that row was copied from your database.
To me it makes sense to keep good records of security events so if I were to get this "signal" I'd make sure the details were captured and the appropriate people informed. Then I'd double check the obvious things like who logged into the database server that day, who had access to backups, who may have worked on database code in that time, when that code was last (or ever) audited and so on.
What if you automatically created a new canary every day with the date in it? Something like something-really-long-xyz-yyyy-mm-dd@yourdomain.com. Then you could find out exactly when your database was stolen. If all canaries up to 2009-01-01 were spammed, and later ones weren't, you'd know they dumped your db on New Year's Day.
All that would tell you is the earliest date your database was compromised and then sold to spammers. If an attacker can harvest your database once, they can do it repeatedly --- which is why warning all your users to change their passwords in this situation is a bad idea.
You are highly unlikely to find your own bugs; teams that have invested millions of dollars in training and process (for instance, requiring detailed internal code reviews before signing off on code to ship, and passing static analysis systems) still have gameovers shown to them during third party pen tests.
Agreed -- I thought that was a good idea (useful + cheap to implement + low false positive rate).
One thing I've also debated about before is whether to use a per-site or a per-user salt for passwords. From this article, it sounds like a per-site hash is enough to defeat the rainbow table services.
Not sure I agree. It seems true that a per-site hash would be enough to defeat _just_ the rainbow table services, but I wouldn't want to have the hacker be able find one password and then be able to distinguish what the salt is for everyone. I would rather have this mitigated by knowing that even if the hacker does brute-force one password, he still won't know the salt for anyone else.
Yeah, I don't think I made the point I meant to make. Previously, I was under the (silly) assumption that folks who were trying to brute force passwords were running their own rainbow set ups, generating them as necessary based on whatever per-site hash was being used. But the fact that services exist for cracking this sort of thing implies that there are plenty of sites that aren't using salts at all, and by using a per-site salt, you are not the softest target.
None of this matters if someone is trying to crack your site in particular, but sometimes you just have to be faster than the slowest gazelle in the herd.
"The first is to create 'canary' accounts. Create accounts that have e-mail addresses, like "something-really-long-xyz-123@gmail.com". This account is not going to get any spam e-mail. When it does get its first spam, you'll know that it came from your database."