A central point of this article seems to be claiming that adding a device in AD is an administrative bottleneck compared to cloud.
If that is the claim, it is false.
By default Authenticated Users have an "Add workstations to domain" capability (AKA ms-DS-MachineAccountQuota attribute on their account) that allows a regular user to create up to 10 machine accounts with zero admin intervention.
Machine accounts have limited rights compared to regular users.
AD provides fine grained distributed policy and cert managment.
After decades of emperical refinement, AD is often superior to what the cloud offers including Entra.
In environments with many PCs that require centrally managed policy, AD is still relavant and probably required.
I think it is also true that there are and always will be many companies that simply cannot expose these PCs to the cloud either because they're working with data that cannot be in the cloud like legal files, finanical documents, schematics or they cannot risk being tethered to the Internet.
You know that little micro PC attached to the cart in every exam room in a doctors office?
There is a 99% chance it is joined to AD.
What Linux admins don’t “get” about Active Directory was the superpowers granted by its Group Policy feature.
Everyone loves to harp on about how terrible the Registry is, but that’s missing the forest for a tree.
The registry + GPO engine is a strongly typed and layered settings engine with:
Machine, Machine Policy, User, and User Policy layered on top of each other, consistently, through pretty much every native app!
More importantly, policy would apply or unapply dynamically and cleanly.
You could move a laptop to a different office and soon as it connected the wifi its set of configured printers would change, its outbound proxy could be reconfigured, etc.
More importantly: Microsoft engineers of the era put in the work of writing up thousands of settings in template files that shipped with Windows, with simple checkbox UI controls and paragraphs of descriptions in every language!
If you think your terrible “settings automation” tool is a viable alternative just ask yourself this question: does it “put the work in” or does its user have to?
Group Policy is so immensely handy. It was a game-changer feature for provisioning and maintaining Windows clients and servers from its introduction in Windows 2000. When Active Directory (AD) was brand new I became immediately enamored with Group Policy. "Superpowers" is definitely the right way to put it.
I had a hard time getting excited about Ansible, Puppet, Desired State Configuration, etc, because I was provisioning machines with Group Policy in 2001. Load Windows, join the domain, and the machine gets configured (software loaded, settings locked-in, defaults overridden, etc).
Today I'm provisioning tons of Windows VMs this way: PXE boot the VM to an unattended Windows setup that ends up joining the domain and getting all its configuration. It makes disposable VM instances and horizontal scaling Windows VMs a breeze.
The now-defunct Likewise Open product (https://github.com/vmware-archive/likewise-open) extended Group Policy to Linux machines. I'd lost touch with the product but it looks like it got GPL'd. It was handy, albeit I don't have the experience with it like I had with the native Windows Group Policy client. I only encountered it in one environment and, at the time, it was commercially licensed.
I'm not in love w/ the CIFS-based portion of Group Policy, but there's nothing stopping anybody from extending the AD schema and using the AD LDAP server to deliver policy settings (or any other mechanism). Using the filesystem to deliver settings always seemed like a hack born of expedience.
Given that this discussion is all past-tense, I assume they're talking about how the product was 25 years ago, when machine translation barely existed.
While the Microsoft of that era got a lot of things wrong, they did get some things right.
Have you ever read Microsoft documentation in a language other than English?
Microsoft's alphabetical indices are even sorted after the original English name instead the translated version. To find a word in the list, you need to search for the English word.
As a random example: https://learn.microsoft.com/fr-fr/windows/win32/adschema/a-u... is "Contrôle de compte d’utilisateur" in the list on the left and sorted between "UPN-Suffixes" and "User-Cert" because "Co" sorts between "UP" and "Us" for Microsoft engineers.
When it comes to friction in corporations it's typically policy based
It starts off easy to get stuff done, then corporate IT starts to lock down (for various reasons), then those on the edge move to new pastures where they can do work (provision a VM on prem? 50 page form passed through 12 departments. Provision the exact same VM in ec2? 3 clicks)
Eventually corporate IT decides that there needs to be governance and the cycle continues.
I've done testing in production, it's usually pretty safe as long as you have a dedicated corner for it. It really helps migrate to production too because you just assign it to the test devices instead of the production ones. Instead of having to build it all up again in a different environment.
What I also liked about it was knowing what was going on in production. These days it's a complete nightmare getting access and thus misconfigurations don't get noticed until they cause issues. There's nobody looking at it anymore except the operations guys who just follow tickets telling them what to do. Nobody to go like "hey why did we flip option X, that's going to cause problems when..."
It's terrible yes. I'm in that situation. For every engineer there's 5 "governance" bureaucrats. I hate it deeply but I can't really move right now because I need to get a mortgage and having a long-time job is a huge plus at my age. And there's never a back from that. Never anyone who goes like "hey all this governance crap makes us really slow to react, stifles ideas and burns all innovation out of us, let's tone it down a bit".
As an example, our "leadership" like most business leaders these days are demanding we do everything with AI constantly. However if we have an idea to automate something there is a 50-step process of approvals, finance, security, privacy, etc teams. Even for stuff where some of these things don't apply because they fall under a flat rate plan. It takes months to wade through all that crap and often at the end we're told it won't fly because some VP of whatever didn't feel comfortable signing off.
And these are often very simple things. Things like a trouble ticket, checking if all the requirements are mentioned and if not to point this out to the user before an actual person has to do this. Low-risk, low-cost. So what do people do? They keep their ideas to themselves and just keep working as they are. Because it's not worth the hassle.
Companies like this tend to just get stagnant because of their policies. Some of them might have been enacted for good reasons but after a while the process becomes a goal in itself. It becomes a monster so big nobody can handle it anymore.
> You know that little micro PC attached to the cart in every exam room in a doctors office? There is a 99% chance it is joined to AD.
So are the cop cars in my city. Walked by one w/two officers in it in a supermarket parking lot last year and had to explain what a trust relationship issue is, unironically.
The document you linked is for Elastic Container Service which is not Kubernetes based. It looks like gMSA is supported for Elastic Kubernetes Service (EKS) if you're using AD as an iDP but I don't think that's the same as AWS recommending it exactly. Most EKS workloads will use IAM Roles for Service Accounts, or EKS Pod Identity. I'd also wager >99% of EKS workloads aren't Windows based so I don't think this is a very widely used technology in that context.
I still don't understand the purpose of gMSA. It almost seems like an admission that password rotation policies are nonsense, so you hide the password rotation deep in Windows, or in the case of your link, a credential-fetcher.
What exactly are we doing here? Am I the only one who thinks this is nuts?
Too many times I have had services break because a service account password expired or was changed without updating the relevant services becuase the service account was reused and not documented.
gMSA accounts solve these problems becuase the password is managed automatically, no need to manually update a service after you set it to run under a gMSA.
This also prevents service account misuse, only specific servers are allowed to access the password.
Password rotation is an issue because the necessity of using a credential incentivizes its insecure storage and low entropy so that ops teams can onboard new services. This in turn necessitates rotation to limit the danger of exposure or kerberoasting (bruteforcing the TGS).
gMSAs ensure that credentials
- Are frequently rotated with no coordination or admin involvement
- Are of sufficient entropy to be non-bruteforceable
- are infeasible to store insecurely
They also robustly solve the question of tracking use: the gMSA states where it may be used and what principals may retrieve the credential, secured with a machine password (which, itself, carries the above security properties).
I remember in the past where we had users in AD, we'd sign in via ASP.NET, where we had to do something to allow that ASP.NET session to be elevated to one valid in Active Directory.
That session was then used to connect to Microsoft SQL Server where access to data was set at a company level.
It all flowed beautifully.
I remember having an issue with A user delegating to Service A which then delegated to Service B. That second hop is kinda special. Service A must be allowed to re-delegate.
(I quickly googled it) The solution was Constrained Delegation with Protocol Transition...
This was pretty much Kerberos, but the embrace-extend Microsoft version was better than the regular krb5.
AD was the only consistent implementation across deployments, because it was nearly identical across customers.
Every linux implementation would be a unique snowflake you needed to know exactly how it worked each time, while the Microsoft version had quirks but it was the same issues in every customer (RBCD).
> Human approval doesn’t scale to developers creating and deprecating testing databases, Kubernetes clusters, and CI runners at will.
Service (or gMSA) accounts handle this, and auth-integrated reverse proxies can move parts of the problem into a more tractable place.
I feel like the article also conveniently forgets that kerberos/ldap backends for SSO systems are extremely common, and solve this.
> In Kubernetes, each pod gets a service account token, which allows it to assert its identity within the cluster.
How is this qualitatively different than onboarding a new computer object or gMSA entitlement in AD?
The problems that the blog asserts affect AD (how to onboard and delegate access and trust) are inherent in any sufficiently robust identity implementation.
Unrelated but Infisical recently moved existing rbac features from their selfhosted version behind a paywall [1]. The discussion has had no response from the team in the past 2 days [2]
sadly, openbao is still the only full-featured and fully-opensource serious contender in this space ("sadly", because I find it a bit clunky to use and configure)
But it's open source and people have Claude so they can code an alternative or maybe just take the pre-paywalled version and add the security updates? Isn't that the point of it being OSS.
I read the reddit post:
> infisical got big off this community. this sub recommending it as the open source vault alternative, people filing bugs, sending PRs, answering questions, giving it stars. and this is how they pay that back.
It's sad it has to be said but bugs, PRs and stars don't pay the bills for the people building and maintaining software. Totally disingenuous of them to gate something in a patch release, no excuse for that at all.
>It's sad it has to be said but bugs, PRs and stars don't pay the bills for the people building and maintaining software.
That's not their argument, it's that Infisical was able to grow into the entity that it is now due to those things, and without them it may never have gotten to the point that it could justifiably try to charge for enterprise anything.
Also those people did marketing, QA, promotion, and development for the platform without compensation, your statement makes it sound like there's no monetary value in those things. So the developer/company has a right to make money off of their work but the contributors don't? Especially when the contributors were providing work and value with the expectation that they were building something the community would be able to use, and now they're not going to be getting that in the future.
This is just another company moving towards enshittification as far as I can tell.
tl;dr - AD history - AD started as the directory for Exchange email server. NT picked up the project and added necessary bits to make it work for a company intranet.
> Human approval doesn’t scale to developers creating and deprecating testing
> databases, Kubernetes clusters, and CI runners at will. Contemporary AD
> implementations deal with this by deferring authority. An admin token (sometimes
> hardcoded…) automatically joins new identities. [...]
> Even if you built all of the complex automations to reflect every testing database
> and Kubernetes pod in AD with automated admin approval, it’s security theater. The
> unified directory would exist on paper, but lack authority. [...]
> Modern cloud infrastructure automatically asserts machine identity: [...] In
> Azure, you can configure a machine identity in Entra ID and let Azure’s Instance
> Metadata Service (IMDS) automatically issue OAuth tokens when machines request
> them.
Okay, so configuring machine identities manually in AD doesn't "scale" to the container age, fine. But configuring machine identities in EntraID does? How is auto-issueing identities in AD "security theater", but auto-issueing identities in EntraID the way to go?
You don't have to restrict AD join to admins only.
There are reasons to do so, but if the claim is that velocity is more important than vetting the entity you are establishing a trust relationship with, it is an option.
Instead, we have 20 passwords for 25 different internal dev tools. We have to pay $2.99/month to some "MDM" that barely has control of the machine.
AD was good in theory, but the weight of crap implementation that had to be preserved for backwards compatibility was awful. It was like IE6 unfortunately.
Kerberos is a really good concept. That should be what we're doing instead of paying $2.99/month*mac for management.
AD might be on the outs but IDPs as a concept are not. Why are you settling for that? Especially in today's era where a simple okta integration can be oneshot by claude?
As usual, some guy with an "all-in-one platform to securely manage application secrets, certificates, and privileged access across cloud, on-prem, and AI infrastructure" trumpets a technical decision he made (maybe seconded by a friend) as The New Paradigm In Computing. Everyone must follow or you're NGMI or something.
Who is this guy? Why should I care? Why should I listen to his pronouncement when I can see plain as day that he's wrong on the facts? (For example, contra the article, AD does do machine accounts.)
What specifically about AD makes us need something new and different in AI? It's like the guy doesn't understand that there's an API to add a principal.
The real hazard in tech isn't failing to keep up with the endless parade of proclaimed New Things, but in whip-sawing your infra this way and that way every time you read an article like this.
Use old, proven technologies unless you have a specific reason to do something else. Give the old tech a chance. Respect Chesterton's Fence.
The recent DevOps Paradox podcast covered similar ground in this week's episode "Secrets Management for AI Agents." The interviewee argues that "an AI agent is a workload and should be identified the way workloads already are, using SPIFFE and its runtime, SPIRE."
AD = LDAP, Kerberos, DNS and user/group/host policy. That is not really ever going away. Particularly for on-prem large institutions with lots of machines. In BYOD environments sure but I don't think that has supplanted your traditional company issued Lenovo/HP/Dell in the majority of institutions.
sure, it is never really going away, the same way ibm mainframes and cobol never fully went away. but the era where practically every company had their own ad server running somewhere is pretty much over. even ms houses are more likely to use entra stuff or something instead of classic ad.
A lot of places are still running AD with Entra in a hybrid model. Its the legacy apps that depend on AD/LDAP that still require the setup, especially if you are global and run everything on internal networks. Eventually shops switch over so Entra becomes the "master" service and AD just inherits the data from there.
a lot of work moved away from those decices and to mobile phones and tablets.
But even on the bigger machines (avoiding the word desktoo) the world changed: Laptios are a lot nore common. There login to domain controler and fetching home directory, mijnting network shares for shared data, etc doesn't work anymore either as machines aren't on the company nerwork, even if provided by the company and under device manafment.
All that data management moved to O365 and other "cloud" syncing.
But all those things still require LDAP-like things and policies which can be managed in AD or similar.
When you can spawn local, remote or virtualized Unix machines for free everywhere AD it's obsolete. Add web clients on top and suddenly the whole NT+AD infra it's as usable as any DOS PC in 2002 for any company.
Machine accounts have limited rights compared to regular users. AD provides fine grained distributed policy and cert managment. After decades of emperical refinement, AD is often superior to what the cloud offers including Entra. In environments with many PCs that require centrally managed policy, AD is still relavant and probably required.
I think it is also true that there are and always will be many companies that simply cannot expose these PCs to the cloud either because they're working with data that cannot be in the cloud like legal files, finanical documents, schematics or they cannot risk being tethered to the Internet. You know that little micro PC attached to the cart in every exam room in a doctors office? There is a 99% chance it is joined to AD.
Everyone loves to harp on about how terrible the Registry is, but that’s missing the forest for a tree.
The registry + GPO engine is a strongly typed and layered settings engine with:
Machine, Machine Policy, User, and User Policy layered on top of each other, consistently, through pretty much every native app!
More importantly, policy would apply or unapply dynamically and cleanly.
You could move a laptop to a different office and soon as it connected the wifi its set of configured printers would change, its outbound proxy could be reconfigured, etc.
More importantly: Microsoft engineers of the era put in the work of writing up thousands of settings in template files that shipped with Windows, with simple checkbox UI controls and paragraphs of descriptions in every language!
If you think your terrible “settings automation” tool is a viable alternative just ask yourself this question: does it “put the work in” or does its user have to?
I had a hard time getting excited about Ansible, Puppet, Desired State Configuration, etc, because I was provisioning machines with Group Policy in 2001. Load Windows, join the domain, and the machine gets configured (software loaded, settings locked-in, defaults overridden, etc).
Today I'm provisioning tons of Windows VMs this way: PXE boot the VM to an unattended Windows setup that ends up joining the domain and getting all its configuration. It makes disposable VM instances and horizontal scaling Windows VMs a breeze.
The now-defunct Likewise Open product (https://github.com/vmware-archive/likewise-open) extended Group Policy to Linux machines. I'd lost touch with the product but it looks like it got GPL'd. It was handy, albeit I don't have the experience with it like I had with the native Windows Group Policy client. I only encountered it in one environment and, at the time, it was commercially licensed.
I'm not in love w/ the CIFS-based portion of Group Policy, but there's nothing stopping anybody from extending the AD schema and using the AD LDAP server to deliver policy settings (or any other mechanism). Using the filesystem to deliver settings always seemed like a hack born of expedience.
You mean a machine-generated pre-translation nobody ever checked and that is unreadable?
While the Microsoft of that era got a lot of things wrong, they did get some things right.
Microsoft's alphabetical indices are even sorted after the original English name instead the translated version. To find a word in the list, you need to search for the English word.
As a random example: https://learn.microsoft.com/fr-fr/windows/win32/adschema/a-u... is "Contrôle de compte d’utilisateur" in the list on the left and sorted between "UPN-Suffixes" and "User-Cert" because "Co" sorts between "UP" and "Us" for Microsoft engineers.
Very helpful to find anything!
It starts off easy to get stuff done, then corporate IT starts to lock down (for various reasons), then those on the edge move to new pastures where they can do work (provision a VM on prem? 50 page form passed through 12 departments. Provision the exact same VM in ec2? 3 clicks)
Eventually corporate IT decides that there needs to be governance and the cycle continues.
What I also liked about it was knowing what was going on in production. These days it's a complete nightmare getting access and thus misconfigurations don't get noticed until they cause issues. There's nobody looking at it anymore except the operations guys who just follow tickets telling them what to do. Nobody to go like "hey why did we flip option X, that's going to cause problems when..."
As an example, our "leadership" like most business leaders these days are demanding we do everything with AI constantly. However if we have an idea to automate something there is a 50-step process of approvals, finance, security, privacy, etc teams. Even for stuff where some of these things don't apply because they fall under a flat rate plan. It takes months to wade through all that crap and often at the end we're told it won't fly because some VP of whatever didn't feel comfortable signing off.
And these are often very simple things. Things like a trouble ticket, checking if all the requirements are mentioned and if not to point this out to the user before an actual person has to do this. Low-risk, low-cost. So what do people do? They keep their ideas to themselves and just keep working as they are. Because it's not worth the hassle.
Companies like this tend to just get stagnant because of their policies. Some of them might have been enacted for good reasons but after a while the process becomes a goal in itself. It becomes a monster so big nobody can handle it anymore.
So are the cop cars in my city. Walked by one w/two officers in it in a supermarket parking lot last year and had to explain what a trust relationship issue is, unironically.
https://docs.aws.amazon.com/eks/latest/best-practices/window...
Too many times I have had services break because a service account password expired or was changed without updating the relevant services becuase the service account was reused and not documented.
gMSA accounts solve these problems becuase the password is managed automatically, no need to manually update a service after you set it to run under a gMSA.
This also prevents service account misuse, only specific servers are allowed to access the password.
Password rotation is an issue because the necessity of using a credential incentivizes its insecure storage and low entropy so that ops teams can onboard new services. This in turn necessitates rotation to limit the danger of exposure or kerberoasting (bruteforcing the TGS).
gMSAs ensure that credentials
- Are frequently rotated with no coordination or admin involvement
- Are of sufficient entropy to be non-bruteforceable
- are infeasible to store insecurely
They also robustly solve the question of tracking use: the gMSA states where it may be used and what principals may retrieve the credential, secured with a machine password (which, itself, carries the above security properties).
That session was then used to connect to Microsoft SQL Server where access to data was set at a company level.
It all flowed beautifully.
I remember having an issue with A user delegating to Service A which then delegated to Service B. That second hop is kinda special. Service A must be allowed to re-delegate.
(I quickly googled it) The solution was Constrained Delegation with Protocol Transition...
This brings back memories.
This was pretty much Kerberos, but the embrace-extend Microsoft version was better than the regular krb5.
AD was the only consistent implementation across deployments, because it was nearly identical across customers.
Every linux implementation would be a unique snowflake you needed to know exactly how it worked each time, while the Microsoft version had quirks but it was the same issues in every customer (RBCD).
Service (or gMSA) accounts handle this, and auth-integrated reverse proxies can move parts of the problem into a more tractable place.
I feel like the article also conveniently forgets that kerberos/ldap backends for SSO systems are extremely common, and solve this.
> In Kubernetes, each pod gets a service account token, which allows it to assert its identity within the cluster.
How is this qualitatively different than onboarding a new computer object or gMSA entitlement in AD?
The problems that the blog asserts affect AD (how to onboard and delegate access and trust) are inherent in any sufficiently robust identity implementation.
Can't you create an Entra app/service principal and use OIDC to avoid personal identities?
[1] https://www.reddit.com/r/homelab/s/teM83aKLMa [2] https://github.com/Infisical/infisical/discussions/9499
It's what most VC backed faux open-source software companies have done for the past decade+ in order to lure in ~~victims~~ customers
I read the reddit post:
> infisical got big off this community. this sub recommending it as the open source vault alternative, people filing bugs, sending PRs, answering questions, giving it stars. and this is how they pay that back.
It's sad it has to be said but bugs, PRs and stars don't pay the bills for the people building and maintaining software. Totally disingenuous of them to gate something in a patch release, no excuse for that at all.
That's not their argument, it's that Infisical was able to grow into the entity that it is now due to those things, and without them it may never have gotten to the point that it could justifiably try to charge for enterprise anything.
Also those people did marketing, QA, promotion, and development for the platform without compensation, your statement makes it sound like there's no monetary value in those things. So the developer/company has a right to make money off of their work but the contributors don't? Especially when the contributors were providing work and value with the expectation that they were building something the community would be able to use, and now they're not going to be getting that in the future.
This is just another company moving towards enshittification as far as I can tell.
When I originally read this I thought the submitter was the author.
- group
- account
- machine
- user PIN, promoted
- machine PIN, promoted
All using with OpenSSH
The manager/creator of AD, Don Hacherl, relayed the early history of AD in a comment correcting Steve Sinofsky's incorrect high-level view of directory services evolution in WinNT - see https://hardcoresoftware.learningbyshipping.com/p/021-expand...
The following article further fleshed out the conflicting priorities of the various groups at MS concerning directory services. see https://hardcoresoftware.learningbyshipping.com/p/bonus-the-...
There are reasons to do so, but if the claim is that velocity is more important than vetting the entity you are establishing a trust relationship with, it is an option.
https://learn.microsoft.com/en-us/entra/identity/managed-ide...
Instead, we have 20 passwords for 25 different internal dev tools. We have to pay $2.99/month to some "MDM" that barely has control of the machine.
AD was good in theory, but the weight of crap implementation that had to be preserved for backwards compatibility was awful. It was like IE6 unfortunately.
Kerberos is a really good concept. That should be what we're doing instead of paying $2.99/month*mac for management.
It rarely breaks, and the featureset it provides is very nice and quite easily expandable.
its a shame integrating it with exchange makes the upgrades a nightmare, because of schema additions from exchange which get put into AD.
Who is this guy? Why should I care? Why should I listen to his pronouncement when I can see plain as day that he's wrong on the facts? (For example, contra the article, AD does do machine accounts.)
What specifically about AD makes us need something new and different in AI? It's like the guy doesn't understand that there's an API to add a principal.
The real hazard in tech isn't failing to keep up with the endless parade of proclaimed New Things, but in whip-sawing your infra this way and that way every time you read an article like this.
Use old, proven technologies unless you have a specific reason to do something else. Give the old tech a chance. Respect Chesterton's Fence.
https://www.devopsparadox.com/episodes/secrets-management-fo...
But even on the bigger machines (avoiding the word desktoo) the world changed: Laptios are a lot nore common. There login to domain controler and fetching home directory, mijnting network shares for shared data, etc doesn't work anymore either as machines aren't on the company nerwork, even if provided by the company and under device manafment.
All that data management moved to O365 and other "cloud" syncing.
But all those things still require LDAP-like things and policies which can be managed in AD or similar.
You are not tied to an user+domain anymore.
Same with what Microsoft copied Active Directory from.