The Dangers of SSL Certificates
52 points
10 hours ago
| 18 comments
| surfingcomplexity.blog
| HN
tialaramex
22 minutes ago
[-]
The monitoring is upside down, which is the case almost everywhere I've ever worked.

You want an upside down pyramid, in which every checked subsystem contributes an OK or some failure, and failure of these checks is the most serious failure, so the output from the bottom of your pyramid is in theory a single green OK. In practice, systems have always failed or are operating in some degraded state.

In this design the alternatives are: 1. Monitor says the Geese are Transmogrified correctly or 2. Monitoring detected a Goose Transmogrifier problem, or 3. Goose Transmogrifier Monitor failed. The absence of any overall result is a sign that the bottom of the pyramid failed, there is a major disaster, we need to urgently get monitoring working.

What I tend to see is instead a pyramid where the alternatives 1 and 2 work but 3 is silent, and in a summarisation layer, that can fail silently too, and in subsequent layers the same. In this system you always have an unknown amount of silently failed systems. You are flying blind.

reply
dextercd
9 hours ago
[-]
You need external monitoring of certificate validity. Your ACME client might not be sending failure notifications properly (like happened to Bazel here). The client could also think everything is OK because it acquired a new cert, meanwhile the certificate isn't installed properly (e.g., not reloading a service so it keeps using the old cert).

I have a simple Python script that runs every day and checks the certificates of multiple sites.

One time this script signaled that a cert was close to expiring even though I saw a newer cert in my browser. It turned out that I had accidentally launched another reverse proxy instance which was stuck on the old cert. Requests were randomly passed to either instance. The script helped me correct this mistake before it caused issues.

reply
weddpros
9 minutes ago
[-]
The scalable way (up to thousands of certificates) is https://sslboard.com. Give it one apex domain, it will find all your in-use certificates, then set alerts (email or webhook). Fully external monitoring and inventory.
reply
compumike
8 hours ago
[-]
100%, I've run into this too. I wrote some minimal scripts in Bash, Python, Ruby, Node.js (JavaScript), Go, and Powershell to send a request and alert if the expiration is less than 14 days from now: https://heyoncall.com/blog/barebone-scripts-to-check-ssl-cer... because anyone who's operating a TLS-secured website (which is... basically anyone with a website) should have at least that level of automated sanity check. We're talking about ~10 lines of Python!
reply
firesteelrain
9 hours ago
[-]
There is a Prometheus plugin called ssl_exporter that will provide the ability for Grafana to display a dashboard of all of your certs and their expirations. But, the trick is that you need to know where all your certs are located. We were using Venafi to do auto discovery but a simple script to basically nmap your network provides the same functionality.
reply
machinationu
1 hour ago
[-]
relevant certificates could be located by scanning the certificate transparency logs
reply
tialaramex
12 minutes ago
[-]
What you're monitoring is "Did my system request a renewed cert?" but what most people's customers care about is instead, "Did our HTTPS endpoint use an in-date certificate?"

For example say you've got an internal test endpoint, two US endpoints and a rest-of-world endpoint, physically located in four places. Maybe your renewal process works with a month left - but the code to replace working certificates in a running instance is bugged. So, maybe Monday that renewal happens, your "CT log monitor" approach is green, but nobody gets new certs.

On Wednesday engineers ship a new test release to the test endpoint, restarting and thus grabbing the renewed cert, for them everything seems great. Then on Friday afternoon a weird glitch happens for some US customers, restarting both US servers seems to fix the glitch and now US customers also see a renewed cert. But a month later the Asian customers complain everything is broken - because their endpoint is still using the old certificate.

reply
philippta
1 hour ago
[-]
When I connect my server over SSH, I don't have to rotate anything, yet my connection is always secure.

I manually approve the authenticity of the server on the first connection.

From then, the only time I'd be prompted again would be, if either the server changed or if there's a risk of MITM.

Why can't we have this for the web?

reply
jsiepkes
3 minutes ago
[-]
> Why can't we have this for the web?

How do you propose to scale trust on first use? SSH basically says the trusting of a key is "out of scope" for them and makes it your problem. As in: You can put on a piece of paper, tell it over the phone, whatever, but SSH isn't going to solve it for you. How is some user landing on a HTTPS site going to determine the key used is actually trustworthy?

There have actually been attempts at solving this with some thing like DANE [1]. For a brief period Chrome had DANE support but it was removed due to being too complicated and being in (security) critical components. Besides, since DNSSEC has some cracks in it (you local resolver probably doesn't check it) you can have a discussion about how secure DANE is.

[1] https://en.wikipedia.org/wiki/DNS-based_Authentication_of_Na...

reply
trvz
6 minutes ago
[-]
Cookie banners aren’t annoying enough for you?
reply
ILearnAsIGo
36 minutes ago
[-]
Would the issue not be that you would need to trust that first connection?
reply
01HNNWZ0MV43FF
31 minutes ago
[-]
reply
dvratil
9 hours ago
[-]
Happened on the first day of my first on-call rotation - a cert for one of the key services expired. Autorenew failed, because one of the subdomains on the cert no longer resolved.

The main lesson we took from this was: you absolutely need monitoring for cert expiration, with alert when (valid_to - now) becomes less than typical refresh window.

It's easy to forget this, especially when it's not strictly part of your app, but essential nonetheless.

reply
firesteelrain
8 hours ago
[-]
Operationally, the issue is rooted in simple monitoring and accurate inventory. The article is apt: “ With SSL certificates, you usually don’t have the opportunity to build up operational experience working with them, unless something goes wrong”

You can update your cert to prepare for it by appending—-NEW CERT—-

To the same file as ——-OLD CERT—-

But you also need to know where all your certificates are located. We were using Venafi for the auto discovery and email notifications. Prometheus ssl_exporter with Grafana integration and email alerts works the same. The problem is knowing where all hosts, containers and systems that have certs are located. Simple nmap style scan of all endpoints can help. But, you might also have containers with certs or you might have certs baked into VM images. Sure, there all sorts of things like storing the cert in a CICD global variable, bind mounting secrets, Vault Secret Injector, etc

But it’s all rooted in maintaining a valid, up to date TLS inventory. And that’s hard. As the article states: “ There’s no natural signal back to the operators that the SSL certificate is getting close to expiry. To make things worse, there’s no staging of the change that triggers the expiration, because the change is time, and time marches on for everyone. You can’t set the SSL certificate expiration so it kicks in at different times for different cohorts of users.”

Every time this happens you whack a mole a change. You get better at it but not before you lose some credibility

reply
renewiltord
4 hours ago
[-]
Can do with any weighted LB, right? E.g. route53 or Cloudflare LB. But even manually you just need k IPs (perhaps even 2) and have host k1 and host k2 report different (overlappingly valid) certs. Then (1/k) users will see bad cert. your usual will be near zero failures but canary will have 100% failures.

I’ve always used the calendar event before expiry and then manual renew option but I wonder why I didn’t do this. It’s trivial to roll out. With Route53 just make one canary LB and balance 1% traffic to it. Can be entirely automated.

reply
loloquwowndueo
10 hours ago
[-]
There are plenty of other technologies whose failure mode is a total outage, it’s not exclusive to a failed certificate renewal.

A certificate renewal process has several points at which failure can be detected and action taken, and it sounds like this team was relying only on a “failed to renew” alert/monitor.

A broken alerting system is mentioned “didn’t alert for whatever reason”.

If this certificate is so critical, they should also have something that alerts if you’re still serving a certificate with less than 2 weeks validity - by that time you should have already obtained and rotated in a new certificate. This gives plenty of time for someone to manually inspect and fix.

Sounds like a case of “nothing in this automated process can fail, so we only need this one trivial monitor which also can’t fail so meh” attitude.

reply
tetha
43 minutes ago
[-]
> If this certificate is so critical, they should also have something that alerts if you’re still serving a certificate with less than 2 weeks validity - by that time you should have already obtained and rotated in a new certificate. This gives plenty of time for someone to manually inspect and fix.

This is also why you want a mix of alerts from the service users point of view, as well as internal troubleshooting alerts. The users point-of-view alerts usually give more value and can be surprisingly simple at times.

"Remaining validity of the certificates offered by the service" is a classical check from the users point of view. It may not tell you why this is going wrong, but it tells you something is going wrong. This captures a multitude of different possible errors - certs not reloading, the wrong certs being loaded, certs not being issued, DNS going to the wrong instance, new, shorter cert lifecycles, outages at the CA, and so on.

And then you can add further checks into the machinery to speed up the process of finding out why: Checks if the cert creation jobs run properly, checks if the certs on disk / in secret store are loaded or not, ...

Good alerting solutions might also allow relationships between these alerts to simplify troubleshooting as well: Don't alert for the cert expiry, if there is a failed cert renew cron job, alert for that instead.

reply
SoftTalker
5 hours ago
[-]
Wait until they start expiring 47 days from issue (coming soon). Though maybe this will actually help, because it will happen often enough that you (a) won't completely forget how to deal with it and (b) have more motivation to be proactive.
reply
yearolinuxdsktp
9 hours ago
[-]
Additionally, warnings can be built into the clients themselves. If you connect to a host with less than 2 weeks cert expiry time, print a warning in your client. That will be further incentive to not let certs be not renewed in time.
reply
1970-01-01
8 hours ago
[-]
I agree with this. Certs are designed to function as digital cliff. They will either be accepted or they won't, with no safe middle ground. Therefore all certs in a chain can only be as reliable as the least understood cert in your certificate management.
reply
0xbadcafebee
2 hours ago
[-]
> With SSL certificates, you usually don’t have the opportunity to build up operational experience working with them, unless something goes wrong. And things don’t go wrong that often with certificates

Don't worry. With 2 or 3 industry players dictating how all TLS certs work, now your certs will expire in weeks rather than years, so you will all be subject to these failures more frequently. But as a back-stop to process failures like this, use automated immutable runbooks in CI/CD. It works like this:

1) Does it need a runbook? Ask yourself, if everything was deleted tomorrow, do you (and all the other people) remember every step needed to get everything running again? If not, it needs a runbook.

2) What's a runbook? It's a document that gives step by step instructions to do a thing. The steps can be text, video recordings, code/shell snippets, etc as long as it does not assume anything and gives all necessary instructions (or links to them) so a random braindead engineer at 3am can just do what it says and it'll result in a working thing.

3) Automate the runbook over time. Put more and more of the steps into some kind of script the user can just run. Put the script into a Docker container so that everyone's laptop environment doesn't have to be identical for the steps to work.

4) Run the containerized script from CI/CD. This ensures all credentials, environment vars, networking, etc are the same when it runs which better ensures success, and that leads to:

5) Running it frequently/on a schedule. Most CI/CD systems support scheduled jobs. Run your runbooks frequently to identify unexpected failures and fix bugs. Most of you get notifications for failed builds, so you'll see failed runbooks. If you use a cron job on a random server, the server could go down, the job could get deleted, or the reports of failure could go to /dev/null; but nobody's missing their CI/CD build failures.

Running runbooks from CI/CD is a game changer. Most devs will never update a document. Some will update code they run on their laptop. But if it runs from CI/CD, now anyone can run it, and anyone can update it, so people actually do keep it up to date.

reply
flowerlad
10 hours ago
[-]
We need a way to set multiple SSL certificates with overlapping duration. So if one certificate expires the backup certificate will become active. If the overlap is a couple of months then you have plenty of time to detect and fix the issue.

Having only one SSL certificate is a single point of failure, we have eliminated single points of failure almost everywhere else.

reply
woodruffw
9 hours ago
[-]
You can do this pretty easily with Let’s Encrypt, to my knowledge. You can request resistance every 30 days, for example, which would give you a ladder of three 90 day certificates.

Edit: but to be clear, I don’t understand why you’d want this. If you’re worried about your CA going offline, you should shorten your renewal period instead.

reply
flowerlad
9 hours ago
[-]
Do services such as K8S ingress and Azure web apps allow you to specify multiple certificates?

Update: looks like the answer is yes. So then the issue is people not taking advantage of this technique.

reply
woodruffw
9 hours ago
[-]
I don’t think there’s a ton of benefit to the technique. If you’re worried about getting too close to your certificate expiry via automation, the solution is to renew earlier rather than complicate things with a ladder of valid certs.
reply
bawolff
2 hours ago
[-]
There are reasons to do this, just not because of expiry.

The main reason to have multiple certs is so if your host (and cert prov key) is compromised, you can quickly switch to a backup, without first having to sort out getting a new cert issued.

reply
kees99
9 hours ago
[-]
Exactly. It's not like backup certificate have validity starting at a future date.
reply
flowerlad
9 hours ago
[-]
Yes the backup certificate can have validity starting at a future date. You just need to wait till that future date to create it.
reply
throw0101c
8 hours ago
[-]
> We need a way to set multiple SSL certificates with overlapping duration.

Both Apache (SSLCertificateFile) and nginx (ssl_certificate) allow for multiple files, though they cannot be of the same algorithm: you can have one RSA, one ECC, etc, but not (say) an ECC and another ECC. (This may be a limitation of OpenSSL.)

So if the RSA expires on Feb 1, you can have the ECC expire on Feb 14 or Mar 1.

reply
deIeted
7 hours ago
[-]
That's a lot of words coming from people who were against this very idea not that long ago. Before Let's Encrypt existed, 90% of you were violently against the idea. "No, that's not how it's supposed to work." That's how it was.
reply
nrhrjrjrjtntbt
4 hours ago
[-]
As always, you need a test that runs and notifies SRE or oncall. Ideally 14 or maybe 28 days before expiry.
reply
gmuslera
9 hours ago
[-]
If you think SSL certificates are dangerous, try seeing the dangers of NOT using them, specially for a service that is a central repository of artifacts meant to be automatically deployed.

It is not about encryption (that a self-signed certificate lasting till 2035 will suffice), but verification, who am I talking with, because reaching the right server can be messed up with DNS or routing, among other things. Yes, that adds complexity, but we are talking more about trust than technology.

And once you recognize that it is essential to have a trusted service, then give it the proper instrumentation to ensure that it work properly, including monitoring and expiration alerts, and documentation about it, not just "it works" and dismiss it.

May we retitle the post as "The dangers of not understanding SSL Certificates"?

reply
duufuvkfmc
8 hours ago
[-]
Debian’s apt do not use SSL as far as I know and I am not aware of any serious security disaster. Their packages are signed and content is not considered confidental.
reply
crote
8 hours ago
[-]
If I'm not mistaken, apt repositories have very similar failure modes - just using PGP certs instead of SSL certs. The repository signing key can still expire or get revoked, and you'll have an even harder time getting every client to install a new one...
reply
tuetuopay
8 hours ago
[-]
Debian 13 uses https://deb.debian.org by default. Even the upgrade docs from 12 to 13 mention the https variant. They were quite hostile for a while to https, but now it seems they bit the bullet.
reply
gmuslera
7 hours ago
[-]
Debian have multiple mirrors, and some distributions even promote to have local mirrors, the model is different, as you say the packages are signed so you know who made them, wherever you got them from.

And I said above, SSL is more than about encryption, but also knowing that you are connecting to the right party. Maybe for a repository with multiple mirrors, dns aliases and a layer of "knowing from whom this come from" is not that essential, but for most the rest, even if the information is public, knowing that it comes from the authoritative source or really from who you think it comes from is important.

reply
direwolf20
8 hours ago
[-]
The selection of packages installed on a server should be treated as confidential, but you could probably infer it from file sizes.
reply
whirlwin
2 hours ago
[-]
TLS certificates is not the only technology for which the default mode is failure. What about disks, databases or syntax errors in configuration files in general?

In technology, there are known problems and unknown problems. Expiring TLS certificates is a known problem which has an established solution.

Imagine if only some of the requests failed because a certificate is about to expire. That would be a debugging nightmare.

reply
0x073
9 hours ago
[-]
And it get worse, as they are changing the max days to until 47 in 2029.
reply
JoshTriplett
9 hours ago
[-]
On the other hand, as the time gets shorter, it'll become less likely that something will go undetected for a long time.
reply
Spivak
5 hours ago
[-]
Infra person here: you will need external monitoring at some point because checking that your site is up all over the world isn't something you want to do in house. Not because you couldn't but because their outages are likely to be uncorrelated with yours—AWS notwithstanding.

Anyway you'll have one of these things anyway and I haven't seen one yet that doesn't let you monitor your cert and send you expiration notices in advance.

reply
superkuh
9 hours ago
[-]
For corporations, institutions, and for-profits this matters and there's no real good solution.

But for human persons and personal websites HTTP+HTTPS fixes this easily and completely. You get the best of both worlds. Fragile short lifetime pseudo-privacy if you want it (HTTPS) and long term stable access no matter what via HTTP. HTTPS-only does more harm than good. HTTP+HTTPS is far better than either alone.

reply
deIeted
7 hours ago
[-]
I think your only defense would be to pretend to be a bot at this point, because what you just said was completely ridiculous and embarrassing. You realize it's not a requirement that you have to post a comment when you have no idea what to say?
reply
throw20251220
9 hours ago
[-]
TLS certificates… SSL is some old Java anachronism.

> There’s no natural signal back to the operators that the SSL certificate is getting close to expiry.

There is. The not after is right there in the certificate itself. Just look at it with openssl x509 -text and set yourself up some alerts… it’s so frustrating having to refute such random bs every time when talking to clients because some guy on the internet has no idea but blogs about their own inefficiencies.

Furthermore, their autorenew should have been failing loud and clear, everyone should know from metrics or logs… but nobody noticed anything.

reply
ronsor
8 hours ago
[-]
> TLS certificates… SSL is some old Java anachronism.

OpenSSL is still called OpenSSL. Despite "SSL" not being the proper name anymore, people are still going to use it.

By the way, TLS 1.3 is actually SSL v3.4 :)

reply
throw20251220
7 hours ago
[-]
You are so confused, it’s not funny. There is no such thing as SSL 3.4. OpenSSL is not SSL. There were 3 SSL versions: 1.0, 2.0, 3.0. Following the 3.0, the protocol has been renamed to TLS. As of 2025, all versions of SSL (1.0, 2.0, 3.0) and early versions of TLS (1.0, 1.1) are considered insecure and have been deprecated by major browsers and the IETF. Modern secure communications rely exclusively on TLS 1.2 and TLS 1.3.
reply
RijilV
7 hours ago
[-]
except of course on the wire, where it's wildly a mess.

TLS 1.3 version in the record header is 3.1 (that used by TLS 1.0), and later in the client version is 3.3 (that used by TLS 1.2). Neither is correct, they should be 3.4, or 4.0 or something incrementally larger than 3.1 and 3.3.

This number basically corresponds to the SSL 3.x branch from which TLS descended from. There's a good website which visually explains this:

https://tls13.xargs.org/#client-hello/annotated

As for if someone is correct or whatever for calling out TLS 1.x as SSL 3.(x+1) IDK how much it really matters. Maybe they're correct in some nerdy way, like I could have called Solaris 3 as SunOS6 and maybe there were some artifacts in the OS to justify my feelings about that. It's certainly more proper to call things by their marketing name, but it's also interesting to note on they behave on the wire.

reply
throw20251220
6 hours ago
[-]
it could be called bunny17.1 on the wire and it would change nothing: https://datatracker.ietf.org/doc/rfc8446/
reply
toast0
8 hours ago
[-]
If we're being picky, they're x.509 certificates, not TLS or SSL.
reply
throw20251220
7 hours ago
[-]
Thanks for the correction.
reply
tomas789
9 hours ago
[-]
I don’t think this is as simple as it seems. For example, we have our own CA and issue several mTLS certificates, with hundreds of them currently in use across our machines. We need to check every single one (which we don’t do yet) because there is an additional distribution step that might fail selectively. And that’s not even touching on expiring CAs, which is a total nightmare.
reply
viraptor
8 hours ago
[-]
If you have your own CA, you log every certificate with the expiry details. It's easier compared to an external CA because you automatically get the full asset list as long as you care to preserve it.
reply
SoftTalker
4 hours ago
[-]
When I ran my own CA I issued certificates with 99-year expiry dates, and I never worried about them again.
reply
throw20251220
7 hours ago
[-]
Why would it be difficult? You have a single CA, so a single place where certs are issued. That means there’s a single place with the knowledge of what certs are issued for which identity, how long are those valid for, and has there been a new cert issued for that identity prior to previous cert expiration. Could not be simpler, in fact.
reply
riffic
8 hours ago
[-]
X.509 certificates
reply
themafia
5 hours ago
[-]
They specified a lot of stuff that ultimately didn't get used but ITU is still my favorite standards organization.
reply
deIeted
7 hours ago
[-]
Nobody to blame but yourselves.

How long did it take for us to get to a "letsencrypt" setup? and exactly 100ms before that existed, you (meaning 90% of you) mocked and derided that very idea

reply
thecosmicfrog
8 hours ago
[-]
> the failure mode is the opposite of graceful degradation. It’s not like there’s an increasing percentage of requests that fail as you get closer to the deadline. Instead, in one minute, everything’s working just fine, and in the next minute, every http request fails.

This has given me some interesting food for thought. I wonder how feasible it would be to create a toy webserver that did exactly this (failing an increasing percentage of requests as the deadline approaches)? My thought would be to start failing some requests as the deadline approaches a point where most would consider it "far too late" (e.g. 4 hours before `notAfter`). At this point, start responding to some percentage of requests with a custom HTTP status code (599 for the sake of example).

Probably a lot less useful than just monitoring each webserver endpoint's TLS cert using synthetics, but it's given me an idea for a fun project if nothing else.

reply
loloquwowndueo
8 hours ago
[-]
Your idea shifts monitoring to end users, which doesn’t sound awesome.

Just check expiration of the active certificate; if it’s under a threshold (say 1 week, assuming you auto-renew it when it’s 3 weeks to expiry; still serving a cert when it’s 1 week to expiration is enough signal that something went wrong) then you alert.

Then you just need to test that your alerting system is reliable. No need to use your users as canaries.

reply
thecosmicfrog
8 hours ago
[-]
Oh absolutely, I wouldn't use this for any production system. It would be a toy hobby project. I just find the notion of turning a no-degradation failure mode into a gradual-degradation one fascinating for some reason.
reply
johannes1234321
8 hours ago
[-]
For a fun project it certainly is a fun idea.

In real life, I guess there are people who don't monitor at all. For them failing requests would go unnoticed ... for the others monitoring must be easy.

But I think the core thing might be to make monitoring SSL lifetime the "obvious" default: All the grafana dashboards etc should have such an entry.

Then as soon as I setup a monitoring stack I get that reminder as well.

reply
firesteelrain
8 hours ago
[-]
This canary is a good thought. The problem the article highlights is that people don’t practice updates enough and assume someone else or something is handling it. You only get better at it the more often it happens which is partly why long expirations are not ideal. This is what the article is highlighting as the main issue.
reply
loloquwowndueo
8 hours ago
[-]
It’s not a good thought. Run a single client (uptime kuma) and ask it to alert you on expiration proximity. I.e. implement proper monitoring and alerting. No need to randomly degrade your users’ experience and hope they’ll notify you instead of shrugging and going to a site that doesn’t throw made-up http errors at them randomly.
reply
firesteelrain
8 hours ago
[-]
If a “canary” is degrading users, it’s misdesigned.

The canary narrows the blast radius and time-to-detection.

reply
loloquwowndueo
7 hours ago
[-]
Agreed. That’s exactly what the proposed canary is - misdesigned.
reply